One holds that she had a basket of eggs that miraculously turned blood red while she was waiting with other women at the tomb of Jesus as he arose from the dead.

The other holds that after Jesus’s ascension into heaven she greeted the emperor with “Christ is risen!” He then pointed to an egg on his table and said, “Christ is no more risen than this egg is red,” after which the egg immediately turned red.

While Easter eggs are often dyed many colors as if an array of spring flowers, in the East they are simply dyed red using vinegar and onion skins in commemoration of the blood Jesus shed on the cross.

Various Slavic traditions add more ornate patterns over the red foundation or use multi-colored artwork.

The Easter egg evokes various imagery associated both with the Christian story and with spring. For example, new life emerges from an egg, which is like an external womb, whereas Jesus arose with new life from the womb of the earth. Christians also gave up eggs for Lent and could finally eat them on Easter. More broadly, in temperate environments chickens stop laying eggs in winter as they recuperate their nutrient and energy stores and start laying them again in the spring, which coincides with Easter.

In most of the world, these are Paschal eggs. “Pascha” is the Latin and Greek derivation of the Hebrew Pesach and ties the blood Jesus shed on the cross to the blood of the Passover lamb.

“Easter,” according to the eighth century English monk Bede, is tied to an ancient goddess of that name, though nothing survives before Bede supporting this. It does have etymological roots with east as a direction, which itself is associated with the sun arising from nightly sleep. It has at least a curiously coincidental surface similarity to the name of the Biblical Jewish queen Ester. Roger Seheult has been collecting some striking similarities between her story and that of the Easter story.

Billions of people will be eating, gifting, or otherwise engaging with Easter eggs today.

About 250 million people will wait until next week, due to the use of an older and less frequently self-correcting approximation of the date of the Paschal full moon.

In honor of the Easter Egg, here are the top five insights about eggs from my 2+ decades of research into nutrition:

Boiled Eggs Are Supercharged with Biotin

Biotin prevents brain fog, supports healthy hair, skin, and nails, stabilizes blood sugar, normalizes cholesterol levels, and supports healthy pregnancies.

About one in thirty people have genetic mutations normalized by high doses of biotin.

All of the biotin is in the yolk.

The egg white, by contrast, ties up the biotin and makes it unusable.

The simplest way to render the egg white completely harmless in this respect is to boil it for four minutes or more.

Three Nutrients Where Eggs Really Shine

While liver is the king of all superfoods, eggs are the picky eater’s general superfood, and both of these really stand out as sources of biotin, choline, and arachidonic acid.

While we covered biotin above, choline and arachidonic acid are no less important.

A livestreamed 2010 USDA committee deliberating the then-upcoming dietary guidelines referred to “The Choline Problem.”

That is, USDA maintains that eggs are bad for you, but all of the data say people need more choline. Thus, “The Choline Problem” is the fact that everyone needs more choline and less eggs, but eggs are by far and away the top source of choline that modern Americans would actually want to eat.

Fatty liver runs rampant, and choline matters more to fatty liver than sugar, alcohol, or fat. In fact, get enough choline in your diet and nothing else will be able to give you fatty liver.

Saturated fat raises the need for choline: it is the ratio of choline to saturated fat that determines how much fat our livers fill up with if we eat foods like butter and cream.

The choline-to-saturated fat ratio is probably important to heart disease as well: recent research on “LDL aggregation” suggests that a high ratio prevents LDL particles from clumping up and contributing to atherosclerotic plaque.

Choline is incredibly important to methylation, being able to keep this process running even if you have that darn MTHFR gene. Methylation prevents us from ruminating too much and makes us mentally flexible, neutralizes our allergies, and pumps our muscles full of creatine.

Choline is used to support cell membranes and make our bile flow smoothly.

Choline is used to make the neurotransmitter acetylcholine, which cools inflammation, helps us rest and digest by activating our parasympathetic nervous system, and in our brains it supports learning, memory, and sustained, focused attention.

Arachidonic acid needs go up when we grow, get pregnant, heal from injuries, or start building muscle.

We use it to power our brains, to waterproof the skin, to prevent food allergies and intolerances, and to attack foreign invaders whenever we start getting sick. Waterproofing the skin is critical to the prevention of eczema, and helps keep you maximally moisturized.

Choline Without the TMAO

Some people are concerned that a compound made from choline in the gut, trimethylamine, and the product we make from that in our liver, trimethylamine oxide (TMAO), contributes to heart disease.

While I am not convinced of this, the form of choline found in eggs is the least likely to generate TMAO in the gut. If you’re worried about this, reach for the egg before the choline supplement.

A Folate Superfood?

The best sources of folate are liver, legumes, and leafy greens.

While we need more data, some analyses suggest that eggs from chickens raised on pasture, who eat leafy greens (grass) all day long, are incredibly high in folate.

The Incredible, Edible Egg Yolk

While boiling your egg can make the entire egg a powerhouse of nutritional value, if you have to throw something away it should be the white, not the yolk.

The yolk contains all of the arachidonic acid, choline, and biotin, as well as 100% of the carotenoids, vitamins A, E, D, and K (8 items). The white does not contain 100% of any nutrient.

The yolk contains more than 90% of the calcium, iron, phosphorus, zinc, thiamin, B6, folate, and B12, and 89% of the panthothenic acid (9 items). The white does not contain more than 90% of any nutrient, but contains over 80% of the magnesium, sodium, and niacin (3 items).

The yolk contains between 50% and 80% of the copper, manganese, and selenium, while the white contains between 50% and 80% of the potassium, riboflavin, and protein.

If you aren’t going to fully cook the white, you are probably better off throwing it out, and if you are going to throw out one or the other, your vitamins, minerals, and fatty acids will be best off throwing out the white.

But, boil that white for four minutes or longer and you’re good to go on all the nutrients.

Happy Easter!

With that, enjoy your Easter Eggs, and Happy Easter!

The idea behind these trials was that the polyunsaturated fatty acids (PUFAs) in seed oils will lower cholesterol levels in the blood and thereby prevent heart disease.

Detractors from this idea focus on PUFAs making our tissues highly vulnerable to oxidative damage — very similar to the rust that accumulates on metal objects kept outdoors over time — and maintain that these fats could worsen the risk of heart disease by making the lipoproteins that carry cholesterol in the blood, such as LDL, more vulnerable to oxidative damage.

The failure of the seed oil trials to show a convincing reduction in heart disease and their worrisome finding that they may increase the risk of cancer suggests that this vulnerability to tissue damage is the mechanism that wins out in the long run.

One of the criticisms of these trials is that they are confounded by the use of trans fats in the seed oil groups.

While there are some trans fatty acids that occur in nature, such as in beef or dairy fat, there are specific trans fatty acids that occur in massive amounts in partially hydrogenated vegetable oils that do not occur in nature.

Hydrogenation itself is done to make liquid oils harder. When carried to completion, it turns all of the monounsaturated and polyunsaturated fatty acids to saturated fatty acids.

However, the path from a polyunsaturated fatty acid to a saturated fatty acid involves a monounsaturated fatty acid as an intermediate.

The definitional chemical characteristic of a saturated fatty acid is that it contains only single bonds and no double bonds, whereas monounsaturated fatty acids contain one double bond and polyunsaturated fatty acids contain more than one. The path from a natural double bond found in an unsaturated fatty acid to the single bond of a saturated fatty acid involves the formation of an unnatural trans double bond.

Thus, partial hydrogenation causes a loss of PUFA, but an increase in both monounsaturated and saturated fats, and the unique production of unnatural trans fats that are not found in any other foods.

Partially hydrogenated oils have been effectively banned in the United States and various other countries, but were widespread during the time of the classic seed oil trials.

Were these trials truly confounded by these unnatural trans fats?

We start our analysis with the longest seed oil trial ever conducted, the eight-year LA Veterans Administration Hospital Study.

The claim that it is confounded by trans fats relates to the use of a margarine in the seed oil group. There were actually two margarines used, one which was free of trans fats and one which contained trans fats. Here, we dig into historical sources such as advertising materials, preserved boxes with ingredient labels, patent infringement lawsuits, and scientific papers of the era. We blend this with direct analyses of the margarines and foods made with them and physiological biomarkers of trans fat consumption to narrow in on the best answer.

But first, why was this trial so incredibly important?

Download This Article

Masterpass members can download this article to print it out and read it in natural lighting as recommended in What Everyone Should Be Doing For Their Health here:

What the LA Vet Trial Tested

The LA Veterans Administration Hospital Study was just over eight years long. It enrolled its subjects between 1959 and 1967, completed in 1968, and had its major publication published in 1969.

Most of the subjects were enrolled in the first two years of the trial, so most of the subjects who survived were in the trial for 6-8 years.

856 male veterans ages 55 or older staying at the Los Angeles home for disabled veterans were randomized to one of two diets that each provided 40% of calories from fat: either a control diet or a diet involving “a replacement of [two thirds of] the saturated animal fats and hydrogenated shortenings of the conventional diet by equal quantities of unsaturated fat in the form of vegetable oils.”

Apart from this change, the diets were nearly identical.

Men under 55 years old were excluded because they tended to have short stays at the home and heart disease takes a long time to develop.

The major importance of this trial lies in these three points:

1. It is one of only two double-blind trials, the other being the Minnesota Coronary Survey, that made a specific substitution of seed oils for traditional fats.

2. It is the longest of any of these trials, most of which were only five years long.

3. It is the only trial with a mean age above 65, which allows us to see the impact of seed oils on cancer.

There were two dining halls, one providing the control diet and one providing the seed oil diet.

The subjects were randomized to receive punch cards that only worked at one or the other dining hall.

The punch cards were used to track how many meals they ate, as they were allowed to leave the home and eat elsewhere.

In order to best preserve blinding, the institutional diet was changed from its historical diet and everyone was told they were on a special diet. Using this new, altered diet as a base, the two dining halls fed the exact same foods except that seed oils replaced traditional fats in one of the halls and only one egg per day was allowed in that hall.

In the first two years of the trial, over 130 people withdrew and there were 53 more withdrawals in the seed oil group than the control group. However, in the last six years of the trial there were only a few withdrawals per year and they were evenly distributed between groups.

The chief complaints in the excess of people withdrawing from the seed oil group in the first two years were disliking the flavor of the milk and complaining about the restriction of eggs. The milk had its fat removed and replaced with seed oils. The investigators improved this process after the first two years to generate a better flavor and people stopped complaining about it.

It appears the people who hated the restriction of eggs most all left during this time and were effectively weeded out. Either the same thing happened to the people who disliked the flavor of the milk or the investigators genuinely closed the gap between the taste of the seed oil-filled milk and regular milk.

Overall, this suggests there was a minor problem with the blinding of the subjects in the first two years but it was subsequently resolved.

The physicians were surveyed in the seventh year to see if they could guess which group the subjects were in. From 229 attempts, they got the control group right 49% of the time and the seed oil group 54% of the time, which is indistinguishable from a coin flip. So, the physicians were well blinded.

While the subjects were living in the home, they ate on average 80% of their meals there.

Because many of the subjects had stopped living there before they died or the study terminated, the total number of meals eaten in the halls was 56% for the control group and 49% for the seed oil group.

Since roughly half of the usual diet was replaced by the seed oil diet in the seed oil group, and since the seed oil diet replaced two-thirds of the traditional fat with seed oils, this is best described as a highly effective randomization to replace one third of the total traditional fat the person ate with seed oils.

At each meal, they took one representative tray and saved it for analysis. Once a week they pooled together the saved trays, homogenized them, and performed chemical analysis of the nutrition. This allowed exquisite characterization of the nutritional content of the diets.

They had originally planned to weigh the food left over, but they found that the amount thrown out was trivial so they stopped doing this. When people asked for second helpings, they gave them foods that were low in fat. This helped preserve the rigor of the quantitative difference in fats between the two groups.

What the LA Vet Trial Found

In the graphs below, “experimental” or “EXP.” refers to the seed oil group.

This graph shows survival from fatal atherosclerotic events over the eight years of the study:

Because the graph shows survival, the line being higher means more people survived and is good.

There are fewer fatal atherosclerotic events in the seed oil group, but the difference is not statistically significant.

The difference begins emerging after the first year, gets wider over time, and spreads apart most in the last two years of the trial.

Statistical significance could be achieved by grouping together the incidence of all “hard endpoints” including definite heart attack, sudden death due to coronary heart disease, stroke, ruptured aneurysm, and amputation due to peripheral artery disease. This brought the p value down to 0.02.

They also achieved statistical significance for fatal atherosclerotic events by separating those between the ages of 54 and 65 from those older than 65.

The benefit was largely confined to the younger group:

Statistical significance is a completely arbitrary cutoff. The p value is best understood as a continuous variable that tells us the probability we would see a difference this large or larger produced by random chance alone. A p value of 0.02 means this probability is 2%.

If we take the finding that seed oils prevent heart disease in young people but not old people at face value, this suggests one of two things are true:

• You need to intervene with seed oils early enough in the course of disease to impact it.
  
  OR
  

• As people get older, everyone will become vulnerable to heart disease and seed oils can’t do anything about it. However, they can prevent someone from getting heart disease much younger than they otherwise would.

The other double-blind trial, the Minnesota Coronary Survey, shows that seed oils make heart disease worse, not better. We will cover that trial in the next article.

The incidence of fatal cancers was higher in the seed oil group:

Because this graph shows the incidence of fatal cancer, the line going up further means more people died of cancer and is bad.

The difference didn’t even start to emerge until year two and really took off in years 5-8. It just barely missed statistical significance at p=0.06. This means the chance of observing a difference this large or larger if seed oils don’t really raise the risk of cancer is 6%.

While there were more deaths from atherosclerosis than from cancer (70 in the control group and 48 in the seed oil group), the relative difference was almost twice as large for cancer deaths: 31 in the seed oil group and 17 in the control group. That’s 46% more atherosclerotic death in the control group but 82% more cancer deaths in the seed oil group.

Taking both findings at face value, seed oils saved 22 people from dying of atherosclerosis but killed 14 people with cancer.

Critics point out that people in the seed oil group who died of cancer had lower adherence than others, but “adherence” was defined as meals eaten, and cancer decreases appetite.

Examining adherence is also an observational study nested inside the trial and leaves the realm of strong causal inferences made from randomization because people weren’t randomized to adhere more or less strongly.

The greater cancer deaths in the seed oil group is especially remarkable because randomization failed to distribute smokers evenly. The control group had twice as many heavy smokers and 60% more moderate smokers!

In fact, the excess atherosclerotic death was found exclusively among those in the control diet smoking more than ten cigarettes per day:

Smoking raises the need for vitamin E, and the control diet was remarkably deficient in vitamin E.

The only clear role of vitamin E in the body is to protect delicate polyunsaturated fats (PUFA) in our cell membranes from being damaged by metabolism or from toxic exposures such as cigarette smoke. It is best evaluated by its ratio to the PUFA it is protecting.

Seed oils have much more PUFA than other fats, so have more vitamin E. However, since the seeds that seed oils come from are not metabolically active, the vitamin E does not have to work very hard. Once these fats accumulate in human tissues, the vitamin E gets used up faster due to the damaging byproducts of human metabolism.

The control diet had barely more than a third of what is found in run-of-the-mill commercial butter and far less than what is found in grass-fed butter, while the seed oil diet was just below the optimal ratio of 0.6 milligrams alpha-tocopherol per gram of PUFA.

Smoking raises the requirement for vitamin E, so the excess of atherosclerosis is best attributed to an interaction between cigarette smoking and vitamin E deficiency.

These should have also biased toward greater cancer, yet the fatal cancer was almost doubled in the seed oil group.

There was no difference at all in total mortality:

However, it seems like the total survival might be starting to diverge after the eighth year, with a higher death rate in the seed oil group emerging.

The beginnings of divergence after the 8-year mark represents a three-month period during which everyone was on the control diet while undergoing final examinations. If this is a real divergence, it is likely a result of 6-8 years of seed oil consumption increasing tissue PUFA content and thereby increasing the vitamin E requirement, while the vitamin E intake was suddenly dropped to the deficient level of the control diet.

However, the excess of non-atherosclerotic death starts diverging in year 7, not after the trial was done:

The reason this doesn’t show up in the total survival graph is because the seed oil group’s advantage for atherosclerotic death is increasing at the same time. Thus, one for one, every person that the seed oil diet saved from an atherosclerotic death was balanced by someone who died of something else.

The most rigorous causal inference from the randomization is that seed oils decreased death from atherosclerosis, increased death from cancer, and had no impact on total mortality.

However, other important observations include that the difference in fatal cancer and non-atherosclerotic death takes years to show up and gets worse as time goes on, and that the putative beneficial effect on heart disease is probably a mirage created by the interaction between smoking and vitamin E deficiency.

The interaction between smoking and vitamin E deficiency should have biased the trial toward worse outcomes in the control group. That total mortality simply broke even suggests the PUFA load in the seed oil group killed people to the same extent as smoking and vitamin E deficiency killed people in the control group.

While not fatal, the seed oil diet also caused a 2.4-fold increase in the the risk of gallstones.

The height of bias would be to reject the cancer finding and accept the heart disease finding at face value.

The level of bias required to do this is extraordinary, for the following reasons:

• First, statistical significance was only found for atherosclerotic death by p-hacking and data mining, either by grouping “hard endpoints” that weren’t fatal with actual fatalities or by subgrouping the atherosclerosis fatalities by age. The non-hacked p value for atherosclerosis fatalities was 0.26, 4.3-fold worse than the non-hacked p value for cancer. The effect size for cancer is almost twice as big.

• Second, the increase in non-atherosclerotic death is, while not straightforwardly “statistically significant,” every bit as statistically powerful as needed to completely abolish any benefit toward atherosclerotic death.

• If the total mortality is identical, you cannot rationally claim a net benefit.

One of the major conclusions of the authors was that they had shown that these trials need to be even longer than eight years, not shorter:

This small excess of nonatherosclerotic mortality in the late years of the study raises the very important and difficult question of whether future clinical trials of diets rich in unsaturated fat must be planned for periods well in excess of eight years rather than for the five-year periods that have been the usual goal.

They considered it “unresolved” whether seed oils have “toxicity” and took the lack of an effect on total mortality to leave their usefulness for heart disease prevention a question without a definitive answer:

Total longevity was not affected favorably in any measurable or significant degree. . . . For this reason, and because of the unresolved question concerning toxicity, we consider our own trial, with or without the support of other published data, to have fallen short of providing a definitive and final answer concerning dietary prevention of heart disease.

Was the LA Veterans Administration Hospital Study Confounded by Trans Fats?

According to the authors, what they were testing was “a replacement of the saturated animal fats and hydrogenated shortenings of the conventional diet by equal quantities of unsaturated fat in the form of vegetable oils in the experimental diet.”

Taken at face value, this suggests the trans fats would be higher in the control diet due to the hydrogenated shortening.

However, there was a margarine used in the seed oil diet that may have increased trans fat consumption in this group.

The following quote from the primary 1969 paper shows the authors themselves considered both diets to be low in trans fats:

We did not analyze our diets or tissue lipids for trans fatty acids. Presumably both diets contained some trans fatty acids from partially hydrogenated fats, but such fats were only minor components of both diets. Therefore, the cholesterol-elevating⁵⁹ and triglyceride-elevating⁶⁰ effects of unnatural fatty acids cannot have been an important factor in this trial.

But is there more to the story?

As they detailed in a 1962 paper, approximately 32 grams of fat per day in the seed oil diet — 28.5% of the total fat — came from an “unsaturated margarine.” 75% of this was used as a table spread, while the rest was used to bake deserts or cook potatoes.

This margarine is the putative source of trans fats in the seed oil group.

But did it contain trans fats?

If it did, how much?

Was it enough to exceed the partially hydrogenated shortening that was explicitly used in the control group?

Let’s dig in.

What Was the “Margarine Based on Unhydrogenated Corn Oil”?

They reported about this table spread that “during the major part of this study, a margarine based on unhydrogenated corn oil has been used.”

According to the footnotes, “this term is used to describe a margarine in which the major component is unaltered vegetable oil, with a small amount of hydrogenated fat added as a hardening agent” and in “the major part” of the study up through at least 1962 the margarine used was Mazola by Corn Products Company.

The funding and gifts statement includes both Mazola Margarine provided by Corn Products Company and Emdee Margarine donated by the Pitman-Moore Company.

Corn Products Company had recently been formed by a merger between the Corn Products Refining Company and Best Foods, while Pitman-Moore was a pharmaceutical company that was a division of Allied Labs, which became part of Dow Chemical in 1961.

The authors wrote two 1962 papers, one on diet and one on “preliminary observations.” They also published a study on the vitamin E status of subjects in the trial in 1965. All three of these papers as well as the primary paper from 1969 refer to both of these margarines, but none of them clarify how much of each was used, whether they were used at the same time, or whether they transitioned from one to the other and when.

We will try to cut at this from several angles later, but first we look at how much trans fat we can expect to be in each of these margarines.

Emdee Margarine Was Free of Trans Fats

It is easiest to show that Emdee margarine did NOT contain any meaningful trans fats, so we start with Emdee.

According to an early advertisement in the Archives of Internal Medicine, the patent-pending margarine was produced by a “special process” that does “not alter the natural characteristics of corn oil.” It consisted of 80% “nonhydrogenated corn oil which has been specially processed to preserve its original content of unsaturated fatty acids” and was fortified with vitamins A and D. It was sold over-the-counter, but was only available in pharmacies. It was marketed as an “effective substitute for animal fats and hydrogenated vegetable oils.”

According to a documents from a patent lawsuit that mentioned Emdee’s composition by way of evaluating a dispute between Mazola and Fleichman’s, the other 20% of Emdee was “fully hydrogenated coconut oil.” This matches the listing of “hardened coconut oil” in the ingredients list of the last ad picture above.

Emdee was associated with US Patent 2,890,959, granted to Robert A Phillips and assigned to Allied Laboratories.

This Emdee can being sold on eBay explicitly references this patent:

Emdee ads started referencing the patent as soon as it was granted in 1959.

The patent itself does not explicitly require full hydrogenation, but it does require melting points so high that they can only be achieved with full hydrogenation, so it effectively requires full hydrogenation of the coconut oil.

While the patent prefers coconut oil, it does not require it. It explicitly gives the example of using fully hydrogenated soybean oil as an alternative and states that any other vegetable oil with a melting point in the correct range could be used.

Fully hydrogenated oil by definition contains no trans fats, because all of the unsaturated fats are converted to saturated fats. It is only partially hydrogenated oil that contains trans fats.

Clearly, the Emdee margarine used in the LA Vet trial was free of any meaningful trans fat.

There is one piece of major uncertainty here: while there is no evidence that coconut oil was ever removed from their ingredients, there is also no evidence that its use persisted past 1962. The ad showing the can with the ingredients cannot be found after 1958. No date can be safely attributed to the can above being sold on eBay, which shows coconut oil in the ingredients. The materials in the Mazola/Fleichman’s suit were submitted in 1962, despite being referenced in the final appeal decision in 1965.

The reason this is important is that coconut oil contains a highly unique fatty acid, lauric acid, that is not found in many other foods and is largely limited to coconut and tropical foods highly related to coconut. This can act as a “fingerprint” of Emdee consumption in the LA Vet trial, but until and unless evidence for the Emdee ingredients between 1963 and 1968 show up we cannot assume the fingerprint will work after 1962.

Mazola Margarine Was Consistently Marketed as Unhydrogenated

Now let’s tackle the slightly less straightforward question of whether Mazola margarine had trans fat.

Mazola consistently emphasized that lack of hydrogenation was one of the major virtues of its margarine. Listen to the woman in this 1961 Mazola Margarine TV commercial explain it:

“I used to buy a corn oil margarine which I thought gave me the same benefits as Mazola, but at a recent food demonstration I learned that the ordinary corn oil margarine I had put such faith in is hydrogenated. So most of its natural liquid corn oil nutrition is lost! But the corn oil in Mazola margarine is never hydrogenated, so you get the full nutritional benefit of Mazola’s liquid corn oil.”

Or take it from the man in this earlier 1960 TV commercial:

“Mazola Margarine — because its major ingredient, pure corn oil, has less saturated fat than the hydrogenated corn oils used in other leading margarines. Less saturated fat because its corn oil is never hydrogenated.”

In a 1960 ad in the Journal of the Medical Society of New Jersey, they had bolded “non-hydrogenated” and used all-caps to say again that its corn oil was “NOT hydrogenated.” This version was also repeated in 1960 issues of the Rhode Island Medical Journal, the Bulletin of the New York Academy of Medicine, and the Illinois Medical Journal; and in 1961 issues of the Journal of the American Osteopathic Association, Industrial Medicine and Surgery, Geriatrics, The Modern Hospital, and Hospital Management.

They used italics to emphasize that the corn oil is never hydrogenated in 1962 issues of Time, the New Yorker, Good Housekeeping, and Ladies’ Home Journal; and in 1963 issues of Better Homes and Gardens, American Home, and Ladies Home Journal.

A 1963 ad in an Ohio newspaper and a California newspaper used italics to emphasize that the “major ingredient” in Mazola margarine is never hydrogenated, and even offered a raffle to win $20,000 that required the person to use their knowledge from the first page to complete this sentence on the second page: “The corn oil in Mazola Margarine has ________ ________ ________ than the hydrogenated oil most other margarines use.”

A brief “History of the Corn Products Company” in a 1961 issue of the Journal of the American Oil Chemists Society says that “Mazola margarine” was “developed in response to demands for new foods containing unhydrogenated corn oil, as a source of the linoleic acid now known to be so important in human nutrition.”

Despite Its Marketing, Mazola Was Made with Partially Hydrogenated Seed Oils

According to a 1961 article about the upcoming “Cholesterol War on TV” in the programming magazine Sponsor, the composition of Mazola margarine was as follows:

This Corn Products spread is 47% corn oil and 53% combination cotton seed and soybean oil. The latter two are hydrogenated until they form irregular walls inside which the liquid corn oil is trapped.

This echoes an earlier 1960 blurb in the journal Food Processing:

Seeming paradox of a firm liquid-oil-containing product is result of revolutionary process whereby a matrix of partially hydrogenated oils holds sub-microscopic droplets of liquid Mazola Corn Oil, much as a honeycomb holds liquid honey.

A holder of the patent connected to Mazola Margarine, Daniel Melnick, published a paper in 1964 that described how it is made. While he didn’t name it as Mazola, he directly tied it to the Mazola patent and his description about the “major ingredient” strongly echoes the marketing of Mazola:

The first, hereafter called blended corn oil margarine fat (BCOMF) is a liquid oil margarine (27) made with non-hydrogenated corn oil as its major ingredient; other components of the blend are a lightly hydrogenated cottonseed oil and a partially hydrogenated soybean oil. The corn oil in this product provides about 90% of the total linoleic acid; the remaining 10% is contributed by lightly hydrogenated cottonseed oil.

The Main Mazola Patent Implies It Has Trans Fats

The main patent that we know is tied to Mazola Margarine, the one Melnick cited, is consistent with all of the above evidence that it contained trans fats.

A 1961 article in Advertising Age gives important background to what we know about the main patent.

Mazola Margarine was originally branded as Cornette in 1959 when it was tested on six markets but then released to the broad market in late 1960 as Mazola.

Fleichman’s beat them to the punch with a corn oil-based margarine, which they released several months before Corn Products started testing Cornette. By 1961, Fleichman’s had double the market share as Mazola.

Fleichman’s made the claim to be 100% corn oil, but all of its corn oil was partially hydrogenated, making it lower in natural polyunsaturated fatty acids.

Mazola made the claim that their corn oil was pure and unaltered, making it higher in these fatty acids.

Fleichman’s second product, however, was just as high in polyunsaturates as Mazola. Hence Mazola had lost their differentiator.

The dilemma at Corn Products is: How can they reach the consumer with the message that Mazola is a superior product, from the polyunsaturate point of view, to the salted Fleichman’s brand?

By 1962, the answer had become clear: sue them for patent infringement.

The lawsuit claimed that both Mazola Margarine and a new reformulation of the older Nucoa Margarine, also made by Corn Products, were covered by the Melnick and Luckmann Patent, No. 2,955,039, which was filed in February of 1959 and granted in October of 1960. The Fleichman’s Regular and Fleichman’s Unsalted margarines were claimed to infringe the patent.

Mazola Margarine ads began in October of 1960 emphasizing this patent:

That ad is from the October 11, 1960 Stamford, CT newspaper the Stamford Advocate.

Ads bearing the patent were repeated throughout 1961, including in a full-page ad in the Time magazine issue that featured Ancel Keys on the cover, in issues of McCall’s, the Saturday Evening Post and in a bulk mailer, and in at least 50 newspaper ads.

Mazola advertisements after 1961 never reference this patent, most likely because the litigation started with pretrial conferences in January of 1962.

The patent infringement claim was ultimately rejected by an appeals court in 1965 on the basis that the technology wasn’t actually novel and that the claims were far too broad.

Good Luck margarine had used a blend of liquid and hydrogenated oil from 1937-1942, and Emdee had made a margarine based on unhydrogenated corn oil in 1958, prior to both Fleichman’s and Cornette in 1959 and Mazola in 1960.

The court ruled that it was clear that Emdee was “the spark” that drove the subsequent development of both Fleichman’s and Mazola.

While the patent was ruled invalid, the patent itself still tells us how the margarine was made and how Melnick, Luckmann, and Corn Products Company viewed the importance of the technology they used.

According to the patent, the novelty of the covered margarine was that it was high in essential fatty acids and had a high ratio of essential fatty acids to saturated fatty acids.

It doesn’t mention anything about trans fats.

The process described involves using 30-70% unhydrogenated liquid oil with the remainder a hydrogenated oil. The hydrogenation used settings that would cause it to yield “substantially complete” hydrogenation of linoleic acid and “selectivity” that would bias it to hydrogenate polyunsaturated fats like linoleic acid rather than monounsaturated fats like oleic acid.

The fact that the linoleic acid was expected to completely convert to saturated fatty acids while the oleic acid was not suggests that the oil would be particularly high in elaidic acid, the trans isomer of oleic acid.

The hydrogenation described is indeed partial hydrogenation, which generates trans fats, consistent with early descriptions in the press and Melnick’s description in the science literature.

Mazola Ingredient Labels of the Era Consistently List “Partially Hardened” Oils

Mazola ads almost never show the ingredients and never show the ingredients clearly. Usually, the Mazola box appears as stylized art that readjusts the logo to occupy the space where the ingredients would be so they could be removed without showing any dead space.

Nevertheless, with a little craftiness we can get our hands on clear ingredients lists from that era.

This is the label of the original 1960 release printed in the journal Modern Packaging:

The second ingredient is “partially hardened soybean and cottonseed oils.”

This Instagram post shows this image, bearing the exact same ingredients list, from a set of grocery store posters from 1964. The account holder messaged me a picture of the box the posters came in, which reads “Brand Names Week May 21-31, 1964 Display Ads.”

This is almost certainly the actual box in circulation in 1964. The original 1960 box showed the corn cob facing up-and-to-the-left, not straight-up. The up-and-to-the-left position persists through 1963, but the straight-up position is found at least as early as 1965.

However, beginning in 1965, the Mazola margarine box read “New! See the Difference! Visibly Better! Definitely More Delicious” when it was “now made by a special process so that it doesn’t burn, blacken, or smoke at normal frying temperatures.”

The package remained like this through 1967.

Thus, 1964 is the only year where the corn cob could have been turned straight up before the new formula was announced on the box.

Jason Liebig (Instagram, LinkedIn) the lead storyteller for the History Channel’s “The Food That Built America,” “The Mega-Brands That Built America” and “Hazardous History with Henry Winkler,” is a collector of vintage brand memorabilia and sent me this image of a 1966 Mazola carton:

The ingredients are identical, except that instead of just using isopropyl citrate to protect the flavor, they had then started using calcium disodium EDTA to protect the flavor as well.

The use of isopropyl citrate and EDTA together to protect the flavor was submitted as a Corn Products Company/Daniel Melnick patent in 1959 and granted in 1961 as US Patent 2,983,615.

An eBay seller claims this is a 1960s-era box:

This is, however, most likely a 1970s-era box.

When I asked for a more precise year, the seller told me they had no idea except that it was from before the advent of bar codes. But the first bar code was scanned on June 26, 1974, so the absence of a bar code does not necessarily isolate it to the 60s.

The corporate office of Best Foods is listed on this box as Englewood Cliffs, NJ, whereas the earlier boxes all list it in New York, NY. An article from Long Beach, CA’s Press Telegram from November 6, 1968 refers to an open house for the new corporate office in Englewood Cliffs. This dates the box to late 1968 at earliest.

I cannot find the phrase “contains golden corn oil” in Mazola Margarine ads prior to 1973.

Another differentiator of this box from earlier ones is that it has a picture that zooms in to show a section of the corn rather than the whole cob and there is a stick of margarine sliced into pats with one of the pats melting on the corn.

Yet another is that the box is a rectangle instead of a square.

All of this imagery can be found in the 1970s-era “We Call It Maize” television campaign.

According to the November 5, 1979 issue of the Oakland Tribune, the New Mexican Apache Tenaya Torres struck a deal for this series in 1976.

All of the boxes in all of these commercials have the sliced margarine stick with one pat melting on the zoomed-in section of corn, and all have the “contains golden corn oil” droplet. I cannot find the rectangular box in any extant videos, but this Facebook post has images of Torres using a rectangular one in some of the ads.

Thus, this box almost certainly comes from the 1970s, not the 1960s.

And yet, its ingredients are exactly the same as they were in 1966, which were the same as from 1960-1965 with the sole exception of added EDTA.

This confirms that throughout the entire LA Veterans Administration Hospital Study, Mazola Margarine listed “partially hardened soybean and cottonseed oils” as its second ingredient.

While “partially hardened” is not a chemically specific term and would no longer be allowed on ingredient labels, FDA guidance at the time was wishy washy. In 1971 FDA increased the specificity required on labels above the regime that had prevailed since 1940 by making the industry be specific about what oil was used and always expressing some type of processing word for hydrogenated oils rather than simply calling them “shortening.” Nevertheless, it stated that the choice between “hydrogenated,” “partially hydrogenated” and “hardened” could be made based on whatever was “factual and desirable” without defining those qualities in any way.

“Partially hardened” should be interpreted in the light of the evidence from the patent, early press, and scientific literature, combined with common sense, to mean “partially hydrogenated.” This suggests that Mazola margarine contained trans fats throughout the entire length of the LA Vet trial.

How Misleading Was the Mazola Advertising?

All of the Mazola marketing focuses on the positive characteristic of having natural polyunsaturated fatty acids rather than the absence of the negative characteristic of having trans fats. It consistently prides itself on being higher in “polyunsaturates” or “linoleates,” and having a high ratio of these to saturated fat.

This can be seen visually in the early television ads, where the cups of corn oil are shown disappearing as the hydrogenation is said to deplete their nutrition.

The 1961 history of the company said the demand for unhydrogenated oil was due to its richness in “the linoleic acid now know to be so important to human nutrition.”

In its late 1960s marketing, it starts comparing itself to protein and vitamins. The “plus” of lobster or steak is “protein,” orange juice’s is “vitamin C,” and that of carrots is “vitamin A,” while Mazola’s “plus” is “polyunsaturates.”

The Mazola brand was so obsessed with increasing intake of polyunsaturated fat that it tried turning “polyunsaturating” into a sexy verb.

1969:

1969:

1970:

They even implied Ringo would be polyunsaturating Beatles tunes in the family den:

It was the embarrassing thing you catch your husband doing with himself in the middle of the night:

What you do with your girlfriends on girls’ nights:

And what you vow to the man you love and honor:

This can explain how Mazola must have internally justified its quiet sleight of hand, where it constantly emphasized the fact that its corn oil or its major ingredient was non-hydrogenated, never even mentioning in the ad copy that it had cottonseed and soybean oil that was hydrogenated, and never showing the ingredients list that bore the incriminating evidence in those ads.

If the whole point was, in the Mazola view, that hydrogenation destroys the “plus” of the natural polyunsaturates, then there’s nothing particularly misleading about this.

Once we become concerned about the “minus” of trans fats, however, this suddenly becomes deeply misleading.

Mazola Patented A Trans Fat-Free Margarine in 1960 But Never Used It

While Mazola as a brand was exclusively interested in getting more polyunsaturated fatty acids into people’s diets, their oil chemist product developers were indeed interested in reducing people’s intakes of trans fats.

Daniel Melnick and Chester Gooding filed a Corn Products Company patent on June 19, 1957 (granted Jan 19, 1960) aimed at using interesterification to produce a margarine with “relatively low trans fatty acid content.”

Interesterification rearranges the three fatty acids on the glycerol backbone of a triglyceride to make them pack together better or give them some other desirable property.

“Hydrogenated fats,” they wrote, “have been subject to frequent criticism in both the scientific and lay press since there is a decrease in essential fatty acid content and a development of isomers of fatty acids, e.g. trans oleic acid, as a result of hydrogenation of oil.”

The specific goal was to get trans fat down below 10%.

The account Melnick submitted to the court in the Mazola/Fleichman’s patent suit was that this patent was never commercialized because it wasn’t “successful,” which presumably means it didn’t produce a margarine that they thought would sell well.

Melnick told the court that his “three prongs” were “reduce saturates, eliminate all isomers of fatty acids, and increase the polyunsaturates.”

“Eliminate all isomers” means to produce a margarine that had zero trans fats.

Chester Gooding filed a second Corn Products Company patent later that year in August of 1960 (granted in mid-1963) with the explicit rationale to replace partially hydrogenated margarines and shortenings with healthier versions that are rich in essential fatty acids and have no trans fats.

This “invention” was a “novel fat” that could be used to produce “margarines and shortenings” that are “characterized by having a high essential fatty acid content” and have “the physical characteristics of fats prepared by partial hydrogenation but which, unlike the latter, contain no partially hydrogenated fats and therefore no glycerides containing isomers of natural unsaturated fatty acids.”

The rationale from the Melnick-Gooding patent was repeated: “Hydrogenated fats . . . have been subject to frequent criticism in both the scientific and lay press since there is a decrease in essential fatty acid content and a development of isomers of fatty acids, e.g., trans oleic acid, as a result of hydrogenation of the vegetable oil.”

The patent covers eight blends. Seven of them use 60, 70, or 80% cottonseed stearine, which is a largely saturated product derived from the refining of cottonseed oil. Blend eight is the exception, and is over 74% corn oil.

The remainder was one of eight different mixtures of fully hydrogenated “interesterified base fat” that could be mixed and matched with the main blend.

The various base fat combinations were made from coconut and cottonseed, palm kernel (very similar to coconut oil) alone, coconut and palm kernel, palm kernel and cottonseed, or coconut and soybean.

Blend 8 using 74% corn oil with the fully hydrogenated and interesterified coconut oil and/or palm kernel oil base fat would have produced a margarine that would be both remarkably similar in its composition to Emdee and consistent with the Mazola marketing that the major ingredient is liquid golden pure unhydrogenated Mazola corn oil, yet would be free of trans fats.

Nevertheless, this patent seems to have languished unused.

No reference to it occurs in the scientific literature until one paper in 1974 and two papers in the 1990s reference it purely to note that it is possible to produce a trans fat-free margarine and that methods to do so have already been patented.

The one time we know they changed their formula, they acted on a five-year-old patent, listed the change in the ingredients list, and introduced it with fanfare on the box label. The absence of any of this for a margarine free of trans fats strongly supports that they never acted on the Gooding patent.

Despite Gooding and Melnick being aware of “frequent criticism in both the scientific and lay press,” public awareness and interest in trans fats was subdued until the mid-1990s. The overwhelming push during the 1960s was to increase intake of PUFA and decrease intake of saturated fat to lower plasma cholesterol levels, and this was where Mazola staked their brand reputation.

Direct Analysis of Mazola Margarine

A 1962 study analyzed seven commercially available margarines. Their seventh was 20.8% linoleic acid and 50.2% trans fat. Since partial hydrogenation destroys linoleic acid and creates trans fat, and since most margarines at the time were low in linoleic acid and high in trans fat, the high linoleic acid content made them speculate that it was based on US Patent 2,983,615. This is the Melnick/Luckmann patent tied to Mazola and Nucoa. This margarine was 50% trans fat.

Daniel Melnick, named on US Patent 2,983,615, co-authored a 1964 paper analyzing a margarine that they directly stated was based on this patent. They strongly echoed the Mazola margarine advertising by saying it was made with “non-hydrogenated corn oil as its major ingredient.” The “other components of the blend,” they wrote, were “lightly hydrogenated cottonseed oil and a partially hydrogenated soybean oil.” This matches the marketing, ingredients, and descriptions of Mazola margarine and it was 25% trans fat.

These two groups used the same methodology to measure trans fats but the first margarine had double the trans fat content of the second, so it is very unlikely these are the same margarine.

The first paper probably refers to Nucoa. It was published in June of 1962, so probably submitted a few months earlier. This was after the 1961 Mazola ads explicitly tying US Patent 2,983,615 to Mazola, but before anything about the 1962 patent suit was made public that would have tied Nucoa to the patent. This would explain why they would use speculative language even though they knew what they bought in the grocery store: Nucoa was made by Corn Products Company and plausibly made according to the Corn Products patent, but wasn’t known to be tied to the patent yet the way Mazola was.

Thus, our best estimate is that Melnick measured Mazola margarine in 1964 as being 25% trans fat.

Which Margarine Was Used In Most of the Trial? Emdee or Mazola?

The LA Vet Trial investigators acknowledged using both Emdee and Mazola margarines in at least four of their papers, none of which explicitly state how much of each was used and whether they transitioned at any point to choosing one or the other.

This is critical, because Emdee Margarine did not have trans fats and Mazola Margarine did.

As such, this question makes or breaks the case of whether trans fats were in the diet of the seed oil group.

We now look at several forms of evidence: textual evidence, historical availability, dietary analysis, and biomarkers.

What the Trial Investigators Actually Said

We start first with the LA Vet trial authors comment that trans fats were “minor components of both diets.”

We then turn to their statements about the use of Mazola rather than Emdee “during the major part of the study” up through 1962.

What Does Trans Fats Being “Minor Components” Mean?

The following quote from the primary 1969 paper shows the authors themselves considered both diets to be low in trans fats:

We did not analyze our diets or tissue lipids for trans fatty acids. Presumably both diets contained some trans fatty acids from partially hydrogenated fats, but such fats were only minor components of both diets. Therefore, the cholesterol-elevating⁵⁹ and triglyceride-elevating⁶⁰ effects of unnatural fatty acids cannot have been an important factor in this trial.

What is a reasonable definition of “minor” here?

Since the next sentence situates this statement with respect to whether trans fats could have had impacts on the trial participants discussed in references 59 and 60, our minimum definition of “minor” has to be “well below what increased cholesterol and triglycerides in those studies.”

Reference 60 is an abstract of an experiment where 1.6-3.9-fold increases in plasma triglycerides were produced by feeding five people diets that were 20-40% of total calories as trans fats. As such, it is not useful for drawing a lower bound beneath which we could consider something “minor.”

On the other hand, reference 59 is very useful for this purpose.

This is a 1961 paper by Ancel Keys and colleagues reporting three controlled experiments, each using 23-27 institutionalized psychotic men, showing the impact of trans fats on serum cholesterol. These experiments used either 30 grams or 100 grams per day of partially hydrogenated oil and compared them to the same amount of the same oil in its natural form, or compared them to blends of other fats and oils that were richer in monounsaturated or saturated fats.

The partially hydrogenated oils were 33-37% trans fat, so the lower bound is 9.9 grams of trans fat per day and the upper bound is 37 grams of trans fat per day.

30 grams of partially hydrogenated safflower oil containing between 9.9 and 11.1 grams of trans fat per day increased serum cholesterol by 10 mg/dL compared to unhydrogenated safflower oil, but this was similar to what would have been predicted by the increased saturated fat in the partially hydrogenated oil. This made it impossible to attribute a specific rise to the trans fats.

By contrast, 100 grams of partially hydrogenated safflower or corn oil providing 33-37 grams per day of trans fats raised serum cholesterol by 25 mg/dL, and this was unambiguously more than what could be attributed to their saturated fat.

These experiments led to the preliminary conclusion that each increase of 1% of total calories in the diet from saturated fats with 12-18 carbons (later revised to exclude the 18-carbon stearic acid) predicts an increase of 2.68 mg/dL serum cholesterol, whereas the same increase in trans monounsaturated fat would predict a rise of 2.1 mg/dL.

The 30 grams of partially hydrogenated safflower oil provided trans fat at 2.7-3% of calories, which the prediction equation derived from the second experiment suggests should have caused a 5.7-6.4-mg/dL rise in serum cholesterol.

This indicates that the trans fats were not failing to increase the serum cholesterol in the first experiment. Rather, the fact that the partially hydrogenated oil had similar amounts of saturated and trans fats made it hard to mathematically tease their impact apart.

“Minor” could mean many things to many people, but here it is meant to dismiss a potential confounder, so it means “we can ignore this because its impact, if any, would be trivial.”

In the context of a randomized controlled trial, it also means “we can assume this was sufficiently distributed across the two groups that it could not have created a meaningful inaccuracy in the estimation of the difference in outcomes between the two groups.”

To satisfy the first definition, the trans fats should have been under 1% of total calories. It is below this point that it would be difficult to detect any impact even under very rigorously controlled conditions. In the LA Vet trial, where the average energy intake was 2496 Calories per day, this would amount to 2.8 grams of trans fat per day.

However, I could see them rationalizing the sloppier definition of less than 30 grams per day of partially hydrogenated oil or less than 10 grams per day of trans fat because this is the lower dose fed in the Keys paper that on the surface was claimed to not show an impact.

In the context of the caloric intake in the LA Vet trial, according to the Keys paper, 10 grams of trans fat per day would translate to a 7.6-mg/dL difference in cholesterol levels. This could only be considered “minor” if it was thought to be spread equally across the two groups.

The average difference in serum cholesterol between the two groups in the LA Vet trial was 29.5 mg/dL. A 7.6-mg/dL theoretical difference is 26% of the actual difference, which is not a “minor” portion of it.

On the other hand, if a 7.6-mg/dL increase was distributed into both groups evenly, to one through shortening and to the other through margarine, I could see them saying that it didn’t matter because the seed oils still created a 29.5-mg/dL difference between groups.

In other words, the confounder would blend into the sameness of the two groups, and the difference that rises above this sameness is what is important.

Therefore, I could see two possible definitions of “minor” here:

• Trans fats were less than 2.8 grams per day in both groups, in which case the difference between the two groups doesn’t matter.

• Trans fats are close to 10 grams per day, but they are evenly distributed between the two groups.

What Does Using Mazola “During The Major Part of the Study” Up Through 1962 Really Mean?

At least four papers from the LA Vet trial reference the use of both Emdee and Mazola margarines.

The only statement indicating the dominance of Mazola is from the 1962 diet paper:

During the major part of this study, a margarine⁷ based on unhydrogenated corn oil has been used.

Footnote 7 then says “Mazola Margarine, Corn Products Company.”

“The major part” implies that Emdee was used in “the minor part.”

However, papers published in the mid-1960s and 1969 simply list both margarines without making any reference to one being more dominant than the other.

The 1962 paper does not tell us whether they started with Emdee and switched to Mazola rather early on or started with Mazola and switched to Emdee rather recently.

By the time the trial was complete, whatever they used from 1963 onward was “the major part” and whatever was “major” in 1962 could have been rendered irrelevant.

The Historical Availability of Mazola and Emdee Margarines

Recruitment for the LA Vet trial began in the summer of 1959, and the first man was assigned to a diet in September of 1959. Everyone was switched onto the control diet eight years later in September of 1967 during a 3-month final observation period.

Thus, the seed oil diet had a margarine demand from September of 1959 to September of 1967.

Emdee was available at least as early as April of 1958.

Emdee almost immediately was used for scientific research. A major early study they boasted about in their advertising was published in the May 16, 1959 edition of JAMA. Emdee was fed to 300 institutionalized psychotics for five months, during which it reduced plasma cholesterol by 23% from an average of 250 mg/dL to 193 mg/dL.

Ads for Emdee continued running in newspapers through late 1968, as shown by this drug store ad from the November 18, 1968 edition of the New Orleans-based The Times-Picayune:

Emdee Margarine completely disappears from newspaper ads after November of 1968. It lingers into the 1970s in catalogs of drugs, manufacturers, and trade names, and it makes its way into a 1990 book about companies and their brands and a 2013 book and its 2021 edition about the politics of nutrition. There is no evidence it was ever sold after November of 1968.

Mazola was released into six markets as “Cornette Margarine” in 1959. The first newspaper ad I can find for this margarine is in the October 13, 1959 issue of The Times Record (Brunswick, Maine).

Cornette had its name changed to Mazola as it was released into the broad market in September of 1960.

Mazola was available throughout the rest of the trial period and is still available today.

In 1970, the year after the LA Vet trial had its main paper published, Daniel Melnick, who was named on the main Mazola Margarine patent, co-wrote an article titled Dynamic utilization of recent nutritional findings: Diet and cardiovascular disease that takes a dig at Emdee:

The first so-called “liquid oil” margarine on the market⁸ was a blend of corn oil with hydrogenated coconut oil. This was a soft, shortening-like product packaged in a tin can for sale in drugstores. Although rich in polyunsaturated fatty acids, it lacked the form, texture, and other sensory characteristics of margarine and, in addition, contained an undesirably high level of the saturated fatty acids from hydrogenated coconut oil.

The next sentence references Mazola:

The first margarine in stick form having a relatively high polyunsaturated and low saturated fatty acid content was introduced in 1959.⁹ Other brands soon became available, giving the consumer considerable choice among products meeting the criteria of the American Medical Association.

The paper has a long accounting of the LA Vet trial itself. It does cover their use of margarine, but doesn’t take the opportunity to tie it to Emdee or Mazola.

While Melnick could be taken to imply that Emdee went out of business because consumers didn’t like it, a more plausible reason is found in the writings of Myra Waldo, a food writer of the era.

Waldo was an early advocate of these margarines. Her cake recipes appeared in the March 12, 1959 New York Post and the March 18, 1959 Dallas Morning News using exclusively Emdee margarine. Her 1961 cookbook, Cooking for Your Heart Health, however, had abandoned Emdee in favor of Mazola because Emdee was “much more expensive and available only in drug stores.”

Since Emdee was available before Mazola, this should lean us toward Emdee being used during the first year until Mazola became available in late 1960.

Analysis of the LA Vet Trial Diets

As a blend of 80% corn oil and 20% fully hydrogenated coconut oil, values for these foods in the USDA database predict that Emdee should be 31% saturated, 24% monounsaturated, and 45% polyunsaturated.

The Mazola/Fleichman’s patent lawsuit materials include a reference to a direct chemical analysis that Melnick had done on Emdee showing it to be 28% saturated fat, a little lower than predicted from the USDA database.

Approximately half of the fully hydrogenated coconut oil should be lauric acid, a 12-carbon saturated fatty acid that is found almost exclusively in coconut and some closely related tropical fats. That should make Emdee 10% lauric acid.

Mazola advertised their margarine as being 19% saturated, 53% monounsaturated, and 28% polyunsaturated.

Based on the production method in the Mazola patent, nearly all of the trans fat should be monounsaturated, and according to Melnick’s analysis it is 25% of the total. Thus, the Mazola breakdown is more precisely 19% saturated, 25% trans fat, 28% natural monounsaturated, and 28% polyunsaturated.

Based on 32 grams of margarine being consumed per day:

• Emdee would provide 3.2 grams of lauric acid per day, while Mazola would provide none.

• Mazola would provide 8 grams of trans fat, while Emdee would provide none.

• Emdee would provide 9-10 grams of saturated fat per day, while Mazola would provide 6.

• Emdee would provide 14 grams of polyunsaturated fat per day, while Mazola would provide 9.

• Emdee would provide 8 grams of monounsaturated fats per day, while Mazola would provide 17 grams, but 47% of Mazola’s monos would be trans fat.

The trial investigators aimed to keep the “iodine value” of the diet constant. The iodine value is a measure of the number of fatty acid double bonds and is based on the ability of these double bonds to absorb iodine. If we calculate a simplified version of this as (monounsaturated fat x 1)+(polyunsaturated fat x 2), Emdee is 36 and Mazola is 35, nearly identical.

This means a change from one to the other would not necessitate a meaningful adjustment to the amount of margarine or to the amounts of other oils to maintain a constant iodine value, so should show up as a change in the fatty acid subtypes of the overall diet:

• If they changed from Emdee to Mazola, saturated fat should have gone down by 3-4 grams, polyunsaturated fat down by 5 grams, and monounsaturated fat up by 9 grams.

• If they changed from Mazola to Emdee, the reverse changes should have happened: saturated fat up 3-4 grams, polyunsaturated fat up 5 grams, and monounsaturated fat down 9 grams.

The LA Vet trial published these distributions of dietary fatty acids.

1962:

1964:

1965:

1966:

1969:

In no case where the 12:0 (lauric acid) is reported does it show up beyond a trace level, and a little blip is consistently shown in the control diet but not in the seed oil diet. There is thus no sign of coconut oil consumption in the seed oil group from 1962 through the end of the trial.

This would seem to rule out the use of Emdee after 1962.

It is uncertain that Emdee used coconut oil 1963-1968, but Emdee was definitely using coconut oil from 1959-1962, so it would seem from the 1962 papers that Emdee had already been cut loose.

Now, suppose also that Emdee did change their formula to use a different fully hydrogenated fat, or that Mazola acted on the Chester Gooding patent to make a margarine free of trans fat. Even if neither were marketing these fats publicly it is at least conceivable that they could have issued a special product for the trial. The resulting margarine would be relatively close to Emdee in its composition, so we would expect to see the same types of changes, just without any lauric acid.

A change toward this type of margarine would mostly show up as an increase of 18:0 and 18:2 at the expense of 18:1.

However, there is no indication that there is any sudden change or any drift in the fatty acid composition at all. This makes it incredibly unlikely that any kind of formula change was made after the publication of the 1962 diet paper.

This suggests that Emdee had been cut loose prior to 1962 and Mazola was used exclusively thereafter.

Biomarker Analysis

In 1962, the “preliminary observations” paper reported that lauric acid was 1% of subcutaneous fat prior to randomization; it dropped to “trace” in the control group after one year and to zero by the second year, but in the seed oil group it stayed at 1% after one year and didn’t drop to zero until the second year.

This suggests that the pre-study “regular” diets of the people contained small amounts of coconut, while the trial diets contained little or no coconut, but that something was sustaining the lauric acid content of the subjects in the seed oil group for the first year of the study that had stopped sustaining it by the second year.

Serum and adipose fatty acids were reported for each year up through year five in 1966, but lauric acid was not included.

In the 1969 main paper, lauric acid is described as “identified inconstantly” both at baseline and at the end of the trial in both groups.

The change from 1% to “identified inconstantly” in the baseline measurements likely reflects the expansion of tissue biopsies from 75 people to 240 people and may reflect the fact that prior to the study some people ate coconut and others didn’t.

The 1969 paper does not report an expanded selection of year-one biopsies and therefore never claims that that first-year reporting of 1% lauric acid in the subcutaneous fat of the seed oil group should be revised.

Nevertheless, the year-one biopsies were only reported for six people in each group in 1962 and therefore could have been strongly influenced by sampling error.

Still, they are consistent with a persistent presence of lauric acid in the seed oil diet but not the control diet for the first year of the study.

Together with the dietary analysis suggesting Emdee had been cut loose by 1962, this suggests that Emdee was the margarine of choice during the first year of the trial but was not used thereafter.

Together with the data on historical availability, this suggests that Emdee was used in the first year because it was the only margarine based on unhydrogenated corn oil on the market in September of 1959, while they switched to Mazola in September of 1960 when it became available.

As to why they made the switch, cost does not make sense because the margarine was provided as a gift.

Either they thought the Mazola Margarine had better sensory characteristics and was better for blinding the subjects, Mazola was simply better at supplying the volume they needed, or they did it because they valued their relationship with Mazola, who was also providing the liquid corn oil that constituted the top oil ingredient of the seed oil diet.

Which Group Had More Trans Fat? Seed Oil or Control?

Since the margarine in the seed oil group appears to have been 25% trans fat after Emdee was abandoned one year into the study, we can quantify the trans fat in the seed oil diet as zero during year one and as eight grams per day years two through eight.

So it was confounded?

Hold your horses.

They do not tell us how much partially hydrogenated shortening was used in the control group, but we need to estimate this before we can even speculate whether there was more trans fat in the seed oil group or the control group.

We now take two approaches:

• textual analysis: what makes sense given how the diets were described, and what comports with their definition of “minor”?

• biomarker analysis: what do the changes in serum cholesterol indicate could have been the “trans fat gap” between the two groups?

Taking a Look at the Diet

Here is a breakdown of the seed oil diet, where the “added vegetable oil” and “added unsaturated margarine” would be replaced by natural animal fats and some portion of partially hydrogenated shortening in the control group.

The main foods I could see them using partially hydrogenated shortening in would be deserts, potatoes, and vegetables.

Deserts accounted for 9% of the fat, potatoes and their substitutes 7%, and vegetables 11%.

If all of these foods were cooked in shortening, shortening would have supplied 27% of the fat. The total fat was 111.2 grams per day, and 27% of this is about 30 grams per day.

The deserts definitely contain some dairy fat, especially in ice cream, so the total shortening should be less than this.

I would say some one-third to two-thirds of deserts were probably foods like pie crusts and pastries where you would expect shortening to be used, so the lower bound is about 3-4 grams of fat from shortening per day (just a third of the desert fat) and the upper bound is about 27 grams per day (these three categories minus a third of the desert fat).

What Fits the Definition of “Minor”?

Because now we are forced to reject the definition of “minor” that no more than 2.8 grams of trans fat were used in each group, we now need to shift to the more liberal definition of “minor” that up to ten grams of trans fats could have been used in each group, but they would have to be distributed equally between the two groups.

This definition is not consistent with 3-4 grams of shortening per day being used in the control group, but it is consistent with around 27 grams per day being used.

We do not have an analysis of the shortening available, but it could easily have been as low as 20% trans fat or as high as 40% trans fat, so there are quite a range of amounts of shortening that could plausibly supply similar amounts of trans fats as used in the control group.

Analyzing the “Trans Fat Gap” Implied by Serum Cholesterol

We now turn to the best biomarker we can use: the difference between groups in serum cholesterol.

This cannot tell us the absolute amount of trans fat consumed by either group, but it can give us an indication of the difference between the two groups.

The preliminary suggestion by Keys that each 1% of calories increase in trans monounsaturated fat would raise serum cholesterol by 2.1 mg/dL has been revised by subsequent research meta-analyzed in 2001 and 2003. The revised prediction is that a 1% increase of calories from trans monounsaturated fat raises cholesterol by 0.31 mmol/L, which is 1.19863 mg/dL.

According to the main 1969 LA Vet trial paper, the Keys equation would have predicted a difference of 36.5 mg/dL between groups, while the observed difference was 29.5 mg/dL.

The authors pointed out that they observed difference is 81% of the predicted difference, and that this is very close to the 78% “adherence” to the diet during the period in which people were living in the home. They therefore suggested that the 22% of the meals that were not eaten at the home (which may or may not have been replaced by a meal outside the home) could have been the reason for the discrepancy.

This makes no sense at all, because in the preceding paragraphs they describe at length how they could not model “adherence” as an explanation of serum cholesterol that achieved a correlation coefficient higher than 0.31. The square of this coefficient represents the explanatory power of one variable for the other. This means that “adherence” to the diet explained 9.6% of the serum cholesterol.

If “adherence” explained less than 10% of serum cholesterol, how could the 78% “adherence” explain why the difference in cholesterol was only 81% of what was predicted by the Keys equation?

The closeness of 78% to 81% is obviously a coincidence.

What we rather have, here, is a 7-mg/dL gap between the difference predicted by the Keys equation and that observed in the study.

Before we consider whether trans fat could account for this, we should also take into account that the LA Vet trial investigators used the equation Keys published in 1965, which attributed the cholesterol-raising effect of saturated fat to be primarily due to lauric (C12:0), myristic (C14:0), and palmitic (C16:0) acids, but did not differentiate between the ability of each of those three fatty acids to raise cholesterol.

A more modern version of the Keys equation by Müller, 2001 attributes a much stronger cholesterol-raising effect to myristic acid than to palmitic acid and attributes a modest lowering effect to oleic acid (C18:1) rather than dismissing it. In this version, the change in total cholesterol is expressed as the following, where each fatty acid is in units of a 1% change in total calories:

My reanalysis of the LA Vet trial data using this equation leads to a prediction of a 25.7 mg/dL difference.

This is actually closer to the observed effect than the 1965 Keys equation, but, critically, the difference is in the opposite direction.

Specifically, “my” 25.7 mg/dL prediction (by simply plugging in numbers into the Müller 2001 equation) is 87% of the observed 29.5 mg/dL difference, whereas the Keys equation is 123% of it. “My” prediction falling 13% below the observed difference is barely more than half the difference of the Keys 1965 prediction rising 23% above it.

This supports the improvements made to the Keys equation over time.

None of these equations are perfect, however, because in the context of a full diet, there could be many other things impacting the cholesterol, so there could be specific food effects in the trial diet that biased the net impact in one direction or another.

For example, cheese consistently lowers total cholesterol relative to butter providing the same amount of milkfat, which might be an effect of the food matrix or mineral content.

Nevertheless, the gap between the predicted and observed difference in serum cholesterol could easily be explained by the difference in trans fat consumption.

Since this gap could be explained by other effects such as food matrix and micronutrients, and since there is certainly some component of random error, this should be viewed as the upper bound of the “trans fat gap.”

The prediction being lower than observed suggests there was more trans fat in the control group.

It means that the seed oils were more effective than expected at reducing serum cholesterol because the control group was more cholesterol-raising than expected due to some of its “oleic acid” actually being the trans isomer, known as elaidic acid.

According to the Müller, 2001 equation, switching from oleic acid to elaidic acid switches the coefficient from -0.170128 to +1.19863 mg/dL, resulting in a 1.368758-mg/dL rise in cholesterol for each 1% of calories making the shift.

Therefore, the 6.2-mg/dL difference in total cholesterol observed versus predicted could be explained by the control group consuming an extra 4.5% of calories as trans fat, which would be an extra 13 grams per day.

There is definitely uncertainty around these calculations and we should not put too much trust in this as an exact estimate. However, I looked through the papers published since Müller, 2001 and I do not see any that challenged the equation as inadequate or provided a better one. The fact that it predicts the observed cholesterol difference with hardly more than half the gap as the Keys, 1965 equation supports its improved utility, which in turn supports the flipped direction toward greater trans fat in the control group.

However, using the Keys papers from 1961 and 1965, we would flip to attributing 3.3% more calories from trans fat to the seed oil group and conclude that they were eating 9.2 grams of extra trans fat per day.

The case to use Müller, 2001 is uncertain but solidly rooted in objective arguments: improvement over time with a larger base of research, unchallenged since, and tighter agreement to the observed results.

The best conclusion is that we should not rule out either difference but we should consider it more likely there was more trans fat in the control group than in the seed oil group.

One way to put this is that our range of possibilities is 9 extra grams in the seed oil group to 13 extra grams in the control group, with probability being toward more in the control group and the exact amount being very uncertain but almost definitely less than these boundary amounts.

Since we should value the Müller, 2001 equation over the Keys 1961/1965 equations, we should be biased toward the right side of the estimate: more trans fat in the control group.

Since the total meals eaten in the home were approximately half of the meals eaten during the study period, we should cut the outer boundaries in half: more likely that there were 6-7 extra grams in the control group than that there were 4-5 extra grams in the seed oil group.

If we give weight to the investigators’ comment that these were “minor” constituents of both diets, this should bias us toward the middle of this estimate rather than the boundary.

This leads me to the tentative conclusion that there was more trans fat in the control group, but not that much more.

Does the Trans Fat Matter?

If I am right that there was more trans fat in the control group, but not that much more, then the trans fats largely fade into the “sameness” across the two groups and the few grams per day pales in comparison to the deliberate randomization of one-third of the fat, to the accidental distribution of twice as many heavy smokers to the control group, and to the poor selection of fats in the control group that rendered them having ten times less vitamin E and almost three times less vitamin E per gram of PUFA.

Another way to look at this would be to ask the question, how much extra trans fat would we expect to need to consume to generate the outcomes observed?

The problem with this is that the outcomes need to be analyzed in the context of the other trials because the benefit to heart disease in this trial is an anomaly.

We therefore turn next to whether trans fats confounded the Minnesota Coronary Survey, the only other double-blind trial, and the one that is distinguished for being largest in size despite not being as long as the LA Vet trial.

Our conclusion here, for now, is that there was probably more trans fat in the control group of the LA Vet trial, but not that much more, and as a result it probably didn’t matter.

• Stocking your cabinets

• Nutrients to focus on for immunity

• At the first sign of a suspicious sneeze

• If you get sick, what to do: week 1, week 2

If you’ve gotten sick and are not recovering like you should, go to my guide on “How to Finally Get Rid of Your Lingering Cough”

For the science on how to leverage nutrients to keep from getting sick, see my course, Nutrition and Immunity.

Gordon Guyatt coined the term “Evidence-Based Medicine” in 1991 and together with David Sackett and others he founded the EBM movement, formally proclaimed in the Journal of the American Medical Association (JAMA) in 1992.

He cofounded the GRADE Working Group in 2000, which has now become embedded in the World Health Organization, Cochrane, and other institutions as the authoritative distribution of EBM’s influence in the world’s institutions.

He started the authoritative textbook on EBM, Users Guide to the Medical Literature, in 2001, which reached its third edition in 2014 and remains the authority of the field.

As such there is no one else alive who speaks more authoritatively on what Evidence-Based Medicine truly means.

You may then be surprised that the familiar “pyramid” of the “evidence hierarchy,” which appears to have been created by an entirely unknown reference librarian at SUNY Downstate in 1997, is, according to Guyatt “totally confused.”

Among other reasons, properly conducted n=1 self-experiments belong at the TOP of the hierarchy, where they have appeared in the EBM textbook since 2001, remaining so in the 2014 edition. Guyatt told me that if he ever produced a fourth edition, n=1 trials would remain right at the top.

We looked together at my own randomized n=1 self-experiment comparing the impact of corn vs quinoa on my sleep and energy.

You may also be quite surprised at how Guyatt described the importance of mechanistic reasoning in interpreting the generalizability of human outcome studies. In some cases, it’s “crucial.”

We talked about how I synthesize mechanistic reasoning, animal studies, and human trials to interpret the way choline status should impact the effect of seed oils on fatty liver and Guyatt weighs in with his verdict on my reasoning.

In this interview, Guyatt and I discuss how to apply the principles of EBM to medicine, health and nutrition, and daily life, paying special attention to the less appreciated aspects of EBM, like the importance of individual values and choice, recognizing what we don’t know, understanding the limits of generalizing from RCTs, weighing mechanistic reasoning against observational studies as means of navigating uncertainty, and how Guyatt responds to critiques from public health and to John Ioannidis’s claim that EBM has been hijacked.

0:29 What was medicine based on before it was based on evidence?
3:29 What is EBM?
5:38 How individual choice is core to EBM but remains underappreciated, and how it took time for it to become a core part of EBM.
10:50 Why the “hierarchy of evidnce” pyramid is “totally confused.”
19:08 Why n=1 trials are at the TOP of the evidence hierarchy
25:09 What might change his mind on statins?
29:15 My own n=1 randomized experiment of corn vs quinoa
33:27 Do randomized controlled trials apply to individual people?
36:06 What is the role of “try it and see if it works” in EBM?
37:38 What is the role of mechanistic reasoning in extrapolating from randomized trials?
46:51 Guyatt evaluates how I synthesize mechanistic reasoning and animal experiments with human trial results to interpret the likely impact of choline status on how seed oils versus saturated fats impact fatty liver disease.
56:08 Does the fact that RCTs aren’t long enough mean that long observational studies can be high-quality evidence?
1:01:08 Whether the upcoming revisions to the GRADE framework are inappropriately reducing skepticism toward observational studies
1:07:31 How he responds to criticisms from public health
1:09:39 Is EBM in “crisis” and has it been “hijacked”?

Mitochondria convert your food to usable energy in the form of ATP, which is used to produce, maintain, repair, distribute, and organize everything in your body.

Abundant health right now, and preserving your health throughout the lifespan toward your longevity, all depends on your mitochondria.

In fact the best explanation for aging is that its a vicious cycle of declining mitochondrial function.

We should always be thinking of mitochondria first. SSRIs, acne treatments, and statins are given as examples. Targeting mitochondria without proper testing has its own set of problems.

This video covers the top things we should all be doing for our mitochondria and how to figure out our own mitochondria’s unique needs.

This is not medical advice and is for educational purposes only.

1:20 Mitochondria govern everything because they convert food into usable energy
7:25 Mitochondrial dysfunction drives aging
10:43 Depression starts with your mitochondria
17:26 The problem with SSRIs
21:59 Acne should start with vitamin A, zinc, B5, and mitochondrial function
28:00 Cardiovascular disease starts with mitochondrial dysfunction
40:26 Statins are mitochondrial toxins
54:00 Targeting mitochondria without testing can be dangerous: three examples.
59:45 CoQ10: no one dose and no one supplement for everyone.
1:04:30 Methylene blue can make your mitochondria worse if you don’t need it.
1:07:48 The power of mitochondrial testing: three examples
1:15:14 Mitochondrial biology
1:17:32 Your mitochondria are pointless if you don’t have creatine
1:18:20 What Mitome is testing
1:20:42 What Mitome reports look like
1:21:57 Energetic bottlenecks are like traffic jams
1:25:01 Organic acid testing of mitochondrial function
1:26:53 Other mitochondrial tests
1:27:44 Five things everyone should do for their mitochondria right now.
1:40:48 We all have unique mitochondrial needs

If you want clarity instead of guesswork, start with the pathway that produces your energy.

Mitome is the first at-home test that measures your cellular energy directly and gives you a personalized roadmap to optimize energy, slow aging, and protect against disease.

Find it here at mito.me
@mitometest

Blood cholesterol levels, along with LDL particle count, ApoB concentration, LDL particle size, LDL pattern, and the ratio of cholesterol between LDL and HDL are all governed centrally by the LDL receptor.

Genetic mutations in LDLR cause familial hypercholesterolemia by lowering LDL receptor expression. One in a million people are born with homozygous mutations, and this can cause heart attacks as young as 18 months old. One in 300 people are born with heterozygous mutations, which shifts heart disease forward from a phenomenon that primarily kills between the age of 60 to 80 to one that primarily kills between the ages of 35 and 60.

The LDL receptor is the main way you bring cholesterol from outside the cell to inside the cell. While the liver only makes about 16% of the body's cholesterol, it is overwhelmingly responsible for controlling the concentration of cholesterol in the blood by taking it up using the LDL receptor. That cholesterol can then be put to productive use, especially for the synthesis of bile acids to support digestion.

However, there are two primary governors of LDL receptor production:

• Thyroid hormone governs it by signaling abundance. This causes the liver to take up more cholesterol than it needs to meet its basic needs, and this is what drives bile acid production to support digestion.

• Cholesterol deficiency within the liver cell governs it by signaling scarcity. The liver takes up cholesterol from the blood not to put it to productive use but simply to replenish its own stores.

Statins work by inducing a cholesterol deficiency in the liver. The liver responds to this by taking up more from the blood to bring its stores back up to normal.

In his 1976 book, Solved: The Riddle of Heart Attacks, Broda Barnes reviewed the history of using thyroid hormone to control both cholesterol levels and heart disease risk. It worked extremely well, but it was discontinued due to what Barnes describes as irresponsible dosing that had killed a few people.

It is important to understand here that you do not need to be “hypothyroid” for thyroid hormone to lower your cholesterol. It will work at any dose. Practitioners using it did not realize they were playing with fire by using a hormone to signal a degree of abundance that was not present in the body. Thus, some of them let the dosing get out of hand because it had nothing necessarily to do with correcting a thyroid deficiency.

Nevertheless, their approach was the same as using statins because both thyroid hormone and statins increase the LDL receptor.

On the other hand, it was the opposite, because thyroid signals abundance while statins signal scarcity.

What everyone involved fails to fundamentally understand is that your LDL receptor does not work in a vacuum. It is powered by mitochondrial ATP production.

If your mitochondria are not working well, your LDL receptor simply will not work.

Doctors who use a pharma-first approach have little respect for nature and doctors who use a statin-first, mitochondria-never approach do so because they do not understand the most basic elements of cellular biology.

As covered in What Everyone Should Be Doing For Their Health, everyone must use a food-first, pharma-last approach, which means everyone needs their doctor to follow the same approach. Do your research until you find a good one.

How Mitochondria Power Your LDL Receptor

As shown in this diagram from Alberts, Molecular Biology of the Cell, the “Bible” of molecular and cellular biology, the LDL receptor clears cholesterol from the blood by binding to LDL particles on the cell surface.

They are present in the membrane bound to the green protein, clathrin, which helps facilitate the membrane invaginating and pinching off to form a vesicle. The protein coat must then be removed, and the vesicle fuses with an organelle known as the endosome, which then uses the digestive organelle known as the lysosome to eat up the contents, digest them, and release the cholesterol.

The figure below shows more detail:

The protein dynamin squeezes the invagination to pinch it off from the cell membrane.

Have you ever tried to squeeze anything without investing energy into it? Here’s something you could try: buy a dynamometer and see what your grip strength is when you feel most awake and energetic. Then try it just before you go to bed, immediately upon waking, or after you finish a grueling workout.

No energy? No squeeze.

Dynamin is powered by the hydrolysis of GTP, ATP’s close cousin. In fact, GTP production is directly fueled by ATP, so without mitochondrial ATP production dynamin won’t function.

Put all the LDL receptors in your membrane you want. If the vesicle can’t squeeze off, the cholesterol is staying in your blood.

Contrary to what some of us learned in high school biology, cells are not sacs of fluid. Things don’t come in and then just float around. They are highly organized microcosms of civilization, with several classes of proteins making up their structural infrastructure and transport superhighways known as the “cytoskeleton.”

The two major cytoskeletal proteins involved in LDL uptake are actin (red) and microtubules (blue).

Actin filaments organize themselves around the invagination to provide structural support to it. They interact with it using myosin VI. This protein is in the same family as the myosin II that performs skeletal muscle contraction. It is directly powered by ATP.

The microtubules make up the transport superhighway network within the cell that, in this case, facilitates the newly formed vesicle moving toward the endosome. The motor proteins that power movement along the microtubules are dynein and various kinesins, which are directly powered by ATP.

Not shown in the figure, the uncoating of the vesicle is performed by heat-shock protein 70 (HSP70), which, like every other heat-shock protein, is directly powered by ATP.

Thus, every LDL receptor that brings an LDL particle in from your blood requires five different classes of ATP-dependent protein to actually move it from the blood into the cell.

No mitochondrial ATP?

No LDL receptor activity.

How Mitochondria Power Your Thyroid Hormone

As I covered in Hormones Are Never In Charge, thyroid hormone is primarily regulated by mitochondrial ATP production.

Caloric intake, especially carbohydrate, drives the pancreas to make insulin and the adipose tissue to make leptin. Adipose production of leptin is at baseline proportional to the total fat stored, so insulin is a gauge of short-term energy status and leptin is primarily a gauge of long-term energy status.

However, these hormones are not directly stimulated by carbs and calories. They are stimulated instead by mitochondrial conversion of food to ATP.

Insulin and leptin both act on the hypothalamus to increase thyroid hormone-releasing hormone (TRH). They do so by increasing ATP production in the hypothalamus.

Insulin and TRH both act together on the thyroid gland to increase the output of thyroid hormone.

Thus, any impairment in mitochondrial ATP production will hurt thyroid hormone production.

Conversely, efficient mitochondrial conversion of food to ATP will drive thyroid hormone secretion because it serves as a highly accurate signal of true energetic abundance.

Optimal Mitochondrial Function Optimizes Blood Lipids

Thus, optimal mitochondrial ATP production exerts a concerted effect on thyroid hormone to increase the LDL receptor and on the five ATP-dependent motor proteins that help bring the LDL particles into the cell.

What happens when you intervene to pump up the LDL receptor without creating the actual abundance necessary for its proper function?

On the one hand, it will not work as well as it would if ATP were abundant.

On the other hand, more LDL receptors will try to use up more ATP, so they will take it from what is available, which will make less available for other things.

Such a solution is like seeing an oil shortage and solving it by importing more cars. Will you incentivize more driving? Yes, you will. Most likely, more people will get from A to B. But now you’ve driven up home heating prices for people who heat their homes with oil, so now you’ll have more people with hypothermia, hypertension, cardiovascular and respiratory problems, and misery.

In other words, you are definitely causing cholesterol to move from your blood to your cells. But at what cost? You are draining your ATP supply and problems are just going to pop up somewhere else.

Statins Are Mitochondrial Toxins

Statins induce cholesterol deficiency in liver cells by impairing the production of mevalonate, a compound far upstream from cholesterol.

Mevalonate is used to produce many different compounds, but two that are central to mitochondrial function are CoQ10 and heme A.

Heme A is the specific iron-based protein that facilitates the use of the oxygen to convert food to ATP in the mitochondria.

By decreasing ATP production, statins hurt methylation. This is because ATP is used to make methylfolate and to activate methionine to S-adenosyl-methionine, the universal methyl donor.

Methylation is needed to produce all mitochondrial proteins. Since methylation is needed to make ATP and ATP is needed for methylation, a decline in methylation that feeds back on mitochondrial ATP production will initiate a vicious cycle.

Methylation is needed to produce creatine, and this is probably why statins decrease creatine synthesis.

Creatine is the mitochondrial power grid. It spreads the impact of mitochondrial ATP production throughout the cell, especially in the general area of the cell known as the cytosol.

It is in the cytosol where S-adenosyl-methionine is made, so creatine supports its production and a decline in creatine status will compromise methylation in a vicious cycle.

Thus, by the time low ATP is hurting methylation beyond a certain threshold, you’ve locked in a vicious supercycle composed of multiple vicious cycles that will cause perpetually declining mitochondrial function.

You now have less ATP and less creatine, which are two hits against LDL receptor activity, even though you have increased the production of that receptor.

To take this back to our oil analogy, this is worse than importing cars to deal with an oil shortage. Rather, this is denying drilling permits so that the supply of oil gets cut in half, then diverting half the oil left over into importing those cars. Now you increase demand after you’ve already decreased supply. You still get more people getting from A to B, but your heating oil shortage has gotten a lot worse and you should probably prepare for a popular revolt among the poor whose children are frostbit and whose grandparents are dying.

In other words, you are hurting the total supply of ATP but diverting what ATP you have toward LDL receptor function. It works to lower cholesterol, but at the cost of other processes that need cellular energy, which is… all of them.

CoQ10 Is Not Sufficient to Prevent or Reverse Statin Myopathy

Statins can cause muscle pain and cramps that can get bad enough in a small number of people to cause exercise intolerance and even rhabdomyolysis. “Rhabdo” is a result of muscle damage that causes muscle proteins to turn the urine a dark color.

As reviewed here, clinical trials only show this happening in 1.5-3% of people, whereas observational studies show that it occurs in 7-38% of people.

The discrepancy can be explained by the fact that people who sign up for clinical trials are healthier than the general population, the trials screen for vulnerability to side effects far more strictly than doctors do for their patients, and you can control the incidence by how strictly you define myopathy. For example, if you count people who have an increase in unexplained cramping, you will get a large number, and if you only count people whose rhabdo put them in the ER you will get a small number.

Meta-analyses conflict as to whether CoQ10 does absolutely nothing or helps substantially with statin-induced myopathy. For example, a 2022 meta-analysis concluded CoQ10 does nothing, and a 2025 meta-analysis concluded that 100-600 milligrams per day has a very helpful effect.

Although the 2025 analysis is more recent, only a single 2022 trial contributing to 16% of the result was published after the 2022 meta-analysis was prepared. The primary difference between the two is that the 2022 meta-analysis required creatine kinase data to verify the myopathy and only included placebo-controlled trials whereas the 2025 meta-analysis required only a survey of muscle pain and included trials that were not placebo-controlled.

Regardless, even if we treat the 2025 analysis charitably, it indicates that CoQ10 is only partially effective. For example, if the subjects reported a muscle pain score of 5-6, CoQ10 would reduce this by 2.5 points, which indicates it could roughly cut the muscle pain in half.

Cutting it in half is good, but if that’s where the result ends, it isn’t a resolution. Moreover, if the vicious supercycle of declining mitochondrial function has not been reversed, it is likely to keep getting worse after the trial is over.

Creatine supplementation has led to the complete resolution of statin myopathy in some cases, but has only been studied in 13 people, published in a case series and a case report.

Metabolomic analysis of statin-associated myopathy cases shows that there are broad shifts in energy metabolism that cannot be attributed to any one single cause and are idiosyncratic between patients. For example, there is a broad shift away from branched-chain amino acid metabolism and toward arginine and proline metabolism, which I suspect reflects impairment in the conversion of pantothenic acid (vitamin B5) to the energy carrier coenzyme A (CoA) as a secondary side effect of the vicious supercycle. There is a broad shift toward altered glycine and serine metabolism, which I suspect reflects impaired methylation. Some people have increased carnitine synthesis and others have decreased carnitine synthesis, showing that the mitochondrial dysfunction is idiosyncratic.

It is possible that some interventions like CoQ10 and creatine can in some people break one or two of the vicious cycles enough to interrupt the vicious supercycle, but neither of these are going to normalize the production of heme A.

There is no supplement that can reverse the inhibition of heme A synthesis, which means nothing can resolve the fact that statins will impair your mitochondrial utilization of oxygen to produce ATP.

Mitome Can Examine What Is Unique to You

As the Founder and Scientific Director of Mitome, I embedded in the reporting the ability to differentiate between impairments in CoQ10-mediated processes, signatures that methylation deficiency is impairing mitochondrial function, and the impact of inadequate complex IV activity that would be caused by a disruption in heme A.

Mitome directly tests the respiratory chain components, where disruptions of CoQ10 and heme A synthesis would first be found, and thus looks at the root cause of the mitochondrial vicious supercycle.

If removal of statins and supplementation with CoQ10 and creatine do not resolve statin myopathy, Mitome could determine what is still missing and what to do about it.

Mitome is not meant to diagnose or treat disease and has not been clinically tested in statin-associated myopathy.

However, it can reveal what the unique limitation in your mitochondrial function is and provide you with an actionable protocol to address it with food and supplements.

Why Mitochondria Must Come First

If you optimize your mitochondrial function, you promote robust thyroid hormone activity to increase your LDL receptor production, and you promote robust supply of mitochondrial ATP production to make the five motor proteins needed for LDL receptor activity do their job.

You also need to be eating enough protein, iodine, and antioxidant nutrients to support your thyroid.

To examine your diet, use the guidelines in What Everyone Should Be Doing For Their Health, track your micronutrients in Cronometer, and run the Comprehensive Nutritional Screening.

If your cholesterol or LDL-C or ApoB or whatever you are measuring remains concerningly high despite adequate micronutrients, and you have also followed the guidelines in What Everyone Should Be Doing For Their Health regarding body composition and stress management, you should experiment with fiber, your fat-to-carbohydrate ratio, and your intake of saturated fat and cholesterol.

In general you need to get your BMI to 18-25 or more precisely optimize your body fat, and, in that context, push your carbohydrates up to a point that does not destabilize your glucose and lactate numbers or make you regain body fat.

More fiber will, in general, help lower cholesterol, but your fiber intake should also be adjusted to what promotes optimal digestion and stool function.

If you have chronic inflammation that is not attributable to excess body fat, you can try a short 12-week course of any of the following, alone or combined: 3 capsules of cod liver oil and 6 capsules of fish oil for high-dose omega-3s; boswellia; 250-2500 milligrams of Specialized Pro-Resolving Mediators; 1000 milligrams once or twice per day of black seed oil; or 1.2 grams (3 capsules) of curcumin as Biocurcumax; or the lactoferrin protocol.

If these do not resolve your blood lipid problem you can try restricting saturated fat, focused on the foods that provide you with the least important micronutrients. You can compare this to restricting cholesterol, again focused on restricting the foods with the least important micronutrients. You can combine these if they both seem to work. While replacing saturated fat with polyunsaturated fat will work, I do not consider it wise.

If after all this you still need to work on your blood lipids you could consider a statin. It may in fact be optimal in certain extreme cases.

However, you should always use a food-first, pharma-last approach to everything, and you should always use a mitochondria-first, statin-last approach to blood lipids.

If you optimize your mitochondria first and you still need more LDL receptor activity, then by the time you intervene, the LDL receptor will be using the most abundant supply of ATP you were able to provide it with.

If instead you use statins — unambiguous mitochondrial toxins — as your first approach, you are killing the ATP supply and then making the LDL receptor take what little there is from the rest of your functions that need it.

It is unlikely that even the most rose-colored clinical trials and case reports are telling the full story. On average, you lose 1% of your mitochondrial function per year. While 100-600 milligrams of CoQ10 or 5 grams of creatine per day may be very helpful in statin-associated myopathy, will that be true ten years after the study was completed when mitochondrial function has another decade worth of decline under its belt? I doubt it.

Age only explains 25% of mitochondrial function, which means that 75% is in your power to optimize.

Focus on getting incredible mitochondria first so you don’t have to try figuring out the cause of mitochondrial failure later.

SSRIs can cause sexual dysfunction and emotional blunting in up to half of people, an unclear incidence of sleep disruption, and a rare risk of suicidality, self-harm, and new-onset psychosis.

Getting off SSRIs can cause discontinuation syndrome in 20-50% of people. It is more common, more severe, and lasts longer when people have been SSRIs for years instead of weeks or months. Many of the discontinuation symptoms represent new-onset dysfunction that cannot be explained by the withdrawal of an effective treatment for the pre-existing depression, such as headache, derealization, depersonalization, cramps, gait abnormalities, and “brain zaps.”

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

My SSRI series led to the conclusion that serotonin’s primary role in the entire body is to coordinate mitochondrial energy production according to shifts in oxygen demand, that SSRIs are primarily mitochondrial drugs, that SSRIs can cause direct and severe mitochondrial toxicity in a small number of people, and that SSRI withdrawal causes new-onset mitochondrial dysfunction in a large number of people.

Many people asked how to transition from SSRIs to using my Depression Protocol, which is a drug-free approach to depression.

What to Do About SSRI Withdrawal Mitochondrial Dysfunction was meant for helping people who experience SSRI discontinuation syndrome or have been experiencing it for a very protracted length. One of its points, however, was that, as long as you’ve been on the drug for more than six to eight weeks, you want the taper to be as slow as possible.

While I am not a medical doctor and cannot manage anyone’s medications, here are some insights derived from the literature about how to make the taper slow, and, most importantly, hyperbolic. Hyperbolic means you don’t go down the same number of milligrams each time; instead, you go down in smaller and smaller increments as you go on.

Mark Horowitz of the Prince of Wales Hospital in Sydney, Australia, and David Taylor of Kings College London Institute of Pharmaceutical Science published a paper on this topic in 2019.

They pointed out that as the dose of an SSRI goes up, the modeled occupancy of the serotonin transporter levels off:

The figure below shows the impact of a linear dose reduction on the top and a hyperbolic dose reduction on the bottom.

When you reduce the dose linearly, your early reductions have a very small impact on transporter occupancy and your later dose reductions have a massive impact.

If you want to achieve a smooth, even reduction in transporter inhibition, you need to make large dose reductions at first, and then progressively smaller and smaller dose reductions as you move on in the taper.

They modeled the range of speeds that could be used, where ten percent of people can taper off in six weeks or fewer, 10 percent of people take longer than eight months, and the median person takes 18 weeks.

In the hypothetical case they gave, the person starts on the median trajectory, but experiences discontinuation symptoms after several dose reductions, so she shifts over to the slower trajectory meant for the 70th percentile.

Henricus Ruhe and colleagues of the Dutch Discontinuation of Antidepressants Task Force replied to their paper with a useful table that models 9- and 10-step taper protocols representing 10% reductions in transporter inhibition per step for the various SSRI and SNRIs:

They suggested that one could start with one week per step instead of the slower approach suggested by Horowitz and Taylor. If the patient reports discontinuation symptoms emerging, they can go back to the previous dose and taper more slowly.

Anecdotally, I have heard of some people taking six months per step, dragging out the taper protocol to five years.

Here is my suspicion:

Take up to as long in the taper as you spent on the drug.

This is based on the literature I reviewed strongly suggesting that incidence, severity, and duration of discontinuation symptoms is driven by the time spent on the drug.

Here is a simple way to approach this. I do not give out medical advice so this is an idea you can bring to your physician to help you get off these drugs if you decide to get off them:

• Follow the dose reduction pattern in the table.

• Make an initial plan for each step to take two weeks.

• If you experience new-onset problems as a result of the taper, go back to the dose prior to the emergence of those problems, and switch to a slower taper.

• You can also make the increment even smaller. For example, you could model the dose reduction after the next, smaller step. Or if you get to “0” in the table, you can replace it with half of the previous dose.

• If you switch to a slower taper you can at first expand it to three weeks, but you should strongly consider taking the entire time you spent on the SSRIs and dividing it by ten. For example, if you have been on SSRIs for eight years, this would be 416 weeks; divided by ten, this makes each step take 41.6 weeks. You could make it simpler by rounding up to 42 weeks, or even just go down one step per year.

If you require a slow taper, you could try the suggestions in What to Do About SSRI Withdrawal Mitochondrial Dysfunction to see if they can improve taper tolerance and allow you to go more quickly.

If you experience a return of depression but not other new problems, you can then use my Depression Protocol for a drug-free approach. In general, it is not contraindicated to mix SSRIs with anything in that protocol, but you should exercise caution with anything else impacting serotonin (e.g. 5-HTP) or monoamine oxidase (dodder, pomegranate, ellagic acid). If you require an extended taper, you should use the the “optimize it” version of the protocol with the assistance of medical supervision as you likely have idiosyncratic mitochondrial dysfunction as either a background vulnerability or as caused by the SSRIs themselves.

Have experience with a successful taper? Post in the comments!

One underappreciated way in which the entire pharmaceutic industry is completely at odds with nature is the concept of the “steady state.” When you dose a drug, you want to find the dose and frequency of administration that leads to a steady plasma level that neither rises nor falls because the drug is replaced at the rate it is leaving the body. This helps maintain it within a therapeutic window that is effective without being overtly toxic.

The problem is that nearly everything in nature is meant to be rhythmic and cyclical, not steady.

This probably underlies why SSRI withdrawal causes mitochondrial dysfunction: SSRIs press the gas pedal of adaptive mechanisms inside cells like sigma-1 receptor activation but they never let go. These are supposed to be stimulated rhythmically to allow for a cycle of stimulus and adaptation, just like exercise and recovery. The steady state pressure put on these systems causes a completely unnatural dependence of mitochondria on the drug that then holds people as captive prisoners.

You don’t need to put the pedal to the metal to cause harm. Just step on the gas pedal a little. If you keep your foot in the same place indefinitely, you have a 100% chance of getting into an accident.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

This Week’s Food Experiment

This week I’ve been testing: are lima beans the mitochondria’s friend or foe?

On the one hand, lima beans are overwhelmingly the richest food source of molybdenum, at least according to the scant literature. In Missing From the Databases: Molybdenum, I include data showing a shockingly low five grams of lima beans per day contain the official requirement for molybdenum. Liver, the only organ meat tested and the only animal food shown to be a good source of this mineral, requires 35 grams to provide the same amount.

Molybdenum gets rid of sulfite, which you make yourself from the protein you eat. Sulfite is a mitochondrial toxin that inhibits numerous enzymes involved in mitochondrial energy production and poisons the mitochondria’s import/export infrastructure, causing them to lose control of what they take in and let out.

On the other hand, lima beans contain linamarin, a cyanogenic glycoside, meaning a sugar-bound molecule that releases cyanide. Cyanide binds to complex IV of the mitochondrial respiratory chain, preventing the mitochondria from using oxygen to produce ATP. This causes a backup in energy utilization that can raise blood glucose. Linamarin, in fact, has been shown to cause diabetes in undernourished dogs.

Worse, inhibiting complex IV can increase sulfite production since this is part of the hypoxia response, thereby nullifying the benefit of the molybdenum.

Further, cyanide is detoxified to thiocyanate, which inhibits iodine uptake into the thyroid gland and thereby hurts mitochondrial energy production by means of downregulating it.

I had been supplementing with 600 micrograms of molybdenum as Mo-Zyme for over a year, and I recently decided to test how much I would need from food alone to prevent my startle reflex from rising, which for me is a sign of molybdenum deficiency. This is driven by sulfite accumulation leading to the formation of S-sulfocysteine, an excitatory neurotransmitter that is relevant to everyone yet totally absent from neuroscience textbooks due to the completely erroneous but universal superstition that it is only relevant to rare diseases.

My experiment on the startle reflex is not over, but my preliminary results suggest that a quarter pound per day of US Wellness 75% Ground Beef With Heart, Kidney, and Liver keeps it normalized. When I reduced this to an eighth pound, below the threshold of startle reflex normalization, I titrated up lima beans in 5-gram-per-day increments, taking just under a month to increase from 5 to 35 grams per day. While each increase seemed to reduce my startle reflex, it felt like I was chasing it upward, having to eat more and more lima beans to keep it down as time went on.

My fasting blood glucose (measured with a KetoMojo) was 87 the day I first added 5 grams of lima beans. My blood glucose rose from 87 to a peak of 118 on 35 grams of lima beans, with the 3-day average reaching 110 the next day, and then it dropped to 89 in my second day after eliminating lima beans.

My lima bean dose during this period correlated to my blood glucose, explaining a whopping 43% of the variation, with a rise of 1 milligram per deciliter for every two grams of lima beans.

If I eliminate the first day of having removed lima beans, where the glucose had fallen to 102, and consider this day still “under the influence of lima beans,” the explanatory power rises from 43% to 50%.

Given that finger prick glucose is intrinsically variable, explanatory power of 43-50% is astronomical.

I conclude that, for me, lima beans are a net mitochondrial toxin.

For more information on lima beans and other sources of cyanogenic glycosides, see Thyroid Toxins: Double-Edged Swords of the Kingdom Plantae.

Coolest Paper I Read This Week

Are common problems driven by rare bottlenecks?

Common Diseases In Clinical Cohorts — Not Always What They Seem is a paper in New England Journal of Medicine where genome sequencing was used in the UK Biobank and five randomized controlled trials to look for rare diseases in people already diagnosed and treated for common diseases like multiple sclerosis, inflammatory bowel disease, and atopic dermatitis.

In 2.86% of multiple sclerosis cases, their multiple sclerosis symptoms could be explained by a monogenic rare disease instead. Further, in these cases there was a 15-fold greater incidence of dementia than expected and a 3.5-fold greater incidence of convulsions.

1.12% of IBD cases could be explained by rare monogenic diseases that hurt intestinal function, and 2.5% of atopic dermatitis could be explained by monogenic rare diseases that cause skin problems.

They uncovered some preliminary evidence that failing to diagnose the rare diseases may account for some failure of the patient to respond to treatments in randomized controlled trials.

This paper is really just uncovering the tip of the ice berg. It doesn’t consider synergistic heterozygosity, complex patterns of inheritance, or the fact that Energy Metabolism Governs Everything and therefore heterozygous bottlenecks in energy metabolism will by the laws of physics necessarily contribute to all diseases.

It suggests that 1-3% of common diagnoses are misdiagnoses but what it’s really hinting at is that all diagnoses are just models to predict treatment response and trying to overcategorize everything into a mirage of binary factual truth is useful until it isn’t, and its usefulness has already been exhausted.

The path foward to understanding unsolved or inadequately treated health problems will be to see them as complex patchworks of biochemical bottlenecks where fixing the weakest link in the biochemical chain will provide the greatest results on the path toward better health. That will always mean maximizing cellular energy production.

Worst Scientific Practice I Found This Week

This is a case of hiding the control group.

This Egyptian paper testing whether Jarrow lactoferrin improves diabetes wins the award for this week. I like Jarrow as a brand, and it isn’t clear that they had any role in the study other than providing the product. At the special request of the investigators, they made a lactoferrin derived from camel milk, presumably to make it culturally relevant in Egypt. This makes it of unclear relevance to the US-sold Jarrow product but isn’t otherwise a bad practice.

The bad practice is this. They claim to have randomly allocated 60 obese children with diabetes to standard therapy with or without the Jarrow product while comparing their values to a set of 50 healthy controls.

Their tables show comparisons of values in 30 diabetics before lactoferrin, the same 30 after lactoferrin, and 30 controls. The result is that we wind up comparing the pre and post values in the 30 diabetics given lactoferrin. This is a very common form of statistical malpractice: the correct comparison in a randomized controlled trial is always the ending value between the treatment and control. This ensures you don’t wind up with conclusions fueled by statistical artifacts like regression to the mean.

But the problem is worse than usual. Are the 30 controls in the table the other 30 diabetics? Probably not, as we don’t have pre and post values for them. But if they are the healthy controls, why are there 30 and not 50 as stated? If they are the 30 diabetics not given lactoferrin, where did the 50 healthy controls go?

Results like these show researchers are just cherry picking whatever will present their data in the best light, presumably because they didn’t like the way it looked when they analyzed it properly.

When reading a paper, always make sure the full populations the authors claim to have studied have their data presented, that the comparisons are the ending values between the treatment and control rather than pre and post, and that any other adjustments or analyses are done after presenting the raw results in the most rigorous manner.

Mitome Story of the Week

I recently talked to a Mitome customer who thought he might have started bonking at marathons because he was getting older. But his Mitome results showed his mitochondria were better fueled with fat than carbohydrate. This allowed him to think through his past results in new light. When he had used chocolate for his intra-marathon fuel, he never bonked. Only after transitioning to carb gels did he stop bonking. It turned out after this analysis his primary problem was a food science problem: how to get chocolate not to melt in his pocket during a marathon.

On the one hand, you might consider this a standard low-carb argument: if you train yourself to burn fat, you don’t have a problem running out of glycogen. But as I covered in Carbs and Sports Performance: The Principles and The Evidence, there are good reasons to think carbs are important to peak performance.

Athletes that succeed on any given diet are self-selected, and the ones who self-select the wrong diet don’t win and get selected out of our attention.

The ability of mitochondria to run better on carbs versus fat is idiosyncratic, and using Mitome to learn which one best describes you can help you choose the fuel on which you can best perform.

As we will see momentarily, however, not all lactoferrin is created equal.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

Lactoferrin is naturally present in milk, saliva, vaginal secretions, semen, mucosal linings, and in polymorphonuclear leukocytes — neutrophils, eosinophils, and basophils — involved in the immune response.

The immune system seems to release it to help bring iron into cells. Iron can feed some pathogens, so an infection drives interleukin-6 (IL-6) to drive lactoferrin release from immune cells and the production of the iron storage protein ferritin inside most cells. The lactoferrin helps bring the iron from the blood into the cells where it gets stored in the ferritin.

Lactoferrin is especially unique in its ability to pick up iron under acidic conditions such as the local sites of infections. It can therefore pick up free iron in infected tissue that the more everyday iron chaperone transferrin would not be able to pick up.

Since lactoferrin accomplishes the mission of keeping iron safely hidden away from pathogens, it sends a “mission accomplished” signal to the immune system by suppressing the continued secretion of IL-6.

This means that for people who are chronically inflamed, lactoferrin — at least, the right type of lactoferrin — can help resolve the inflammation by bringing IL-6 back down to normal.

Each molecule of lactoferrin can bind to two iron atoms, and each 100 milligrams of lactoferrin can bind to 3.57 milligrams of iron. The main dietary source of lactoferrin is milk. (Especially raw milk, since pasteurization destroys 65% of the lactoferrin). In milk, 15-20% of lactoferrin’s iron-binding sites are bound to iron. The portion bound to iron is known as holo-lactoferrin or iron-saturated lactoferrin, while the portion that is not bound to iron is called apo-lactoferrin.

While all natural lactoferrin is a mix of the two, nearly all supplements on the US market are 100% apo-lactoferrin.

In one Italian clinical trial, 100 milligrams twice a day of 20-30% iron-saturated lactoferrin taken just prior to meals improved anemia in women with various forms of chronic inflammation just as effectively as an iron supplement. This was striking because the lactoferrin contained less than one milligram of iron whereas the iron supplement provided 105 milligrams per day of iron as ferrous sulfate. Despite having so little iron, lactoferrin increased serum iron by 48%. It accomplished this by cutting IL-6 in half, which allowed the iron to escape the grasp of ferritin, thereby allowing it to be mobilized for circulation throughout the body and also for hemoglobin production.

That trial was not randomized, but a randomized controlled Egyptian trial in adults with chronic kidney disease and anemia and another in children found similarly that the improvement in hemoglobin was the same or better with lactoferrin than iron and that the lactoferrin reduced IL-6 levels.

I have reviewed the literature more broadly in the appendix.1

While the 20-30% iron-saturated lactoferrin has not been studied head-to-head against apo-lactoferrin, iron-saturated lactoferrin is more resistant to digestion, has a different profile of antimicrobial activity, is more reflective of what is available from natural foods, and is what has been studied for IL-6 lowering and therefore what we should base our conclusions on.

The smaller body of literature using apolactoferrin does not show clear reductions in IL-6.

Thus, the natural, partially iron-saturated form of lactoferrin is what should be used to try to resolve sticky inflammation.

Until recently, the only lactoferrin supplement on the market was an Italian product, Lattoglobina.

However, two brands have recently started selling such a product in the US:

• Double Wood sells a natural partially iron-saturated lactoferrin that appears to be purified from dairy products using a patented process with high purity.

• The Lactoferrin Co sells 95% pure partially iron-saturated lactoferrin isolated from grass-fed milk, as a powder or as enterically coated capsules.

The need for enteric coating is unclear and has not been clinically studied. Taking the lactoferrin on an empty stomach can prevent the stomach from becoming acidic enough to degrade it, and the successful anemia trials approximated this by having the lactoferrin taken just prior to the meals. It is likely the Double Wood product is more pure for those sensitive to dairy, but neither product is dairy-free. It is worthwhile to experiment with the three products from the two brands to compare their utility since their differences lie in gray areas.

100 milligrams twice a day cut IL-6 in half, so it might be that 400 milligrams per day in 2-4 divided doses could completely normalize IL-6. While this has not been studied much, it is worth trying 200-400 milligrams of natural iron-saturated lactoferrin per day in 2-4 divided doses, taken before meals or on an empty stomach, to help resolve stubborn inflammation, especially if your IL-6 is high.

While milk naturally contains lactoferrin, and while the early milk meant for a newborn known as colostrum is especially enriched in it, a capsule of colostrum only contains about 2.6 milligrams of lactoferrin and raw milk (pasteurization destroys over half of the lactoferrin) probably only has around 10 milligrams per cup. Thus, colostrum is a good source if you are drinking it in volume like a newborn would and raw milk is a good source if it makes up most of your calories, but these are not good sources to obtain 200 to 400 milligrams of lactoferrin per day as a supplement to your existing diet.

On the other hand, while whey protein manufacturers do not measure and report the lactoferrin contents of their products and there is about two-fold variation between products according to the literature, we can say that one to two 20-gram scoops of whey protein will provide about 200 milligrams of natural partially iron-saturated lactoferrin, and you can use two to four scoops to obtain 400 milligrams.

The supplements allow you more optionality for your diet since you don’t need to occupy a large portion of your protein requirement and a substantial portion of your caloric load with a protein powder.

I have updated several guides, ebooks, and articles that had only linked to the Italian product to now include the Double Wood and Lactoferrin Co products:

• The Vitamins and Minerals 101 Cliff Notes

• Vitamins and Minerals 101: Iron

• The Iron Deficiency Protocol

• The Crohn’s Protocol

• The Science Behind the Crohn’s Protocol

• The COVID Guide

• Is Lactoferrin Chelating Your Iron?

The Iron Deficiency and Crohn’s protocols and the COVID Guide also include dairy-free options for lowering IL-6.

A meta-analysis of 27 trials including 3,808 patients supports St. John’s Wort as a monotherapy having nearly identical efficacy to SSRIs when used for four to twelve weeks, but with a 41% lower rate of dropping out due to side effects.

A meta-analysis focused on clinical recommendations concluded that “St John’s wort flowers at doses of 600mg to 1800mg (3:1–7:1 extract depending on product) per day standardised to a dose of approximately 0.2–0.3% hypericin and/or 5–6% hyperforin (once to three times per day depending on extract) is recommended for monotherapy in mild to moderate depression.”

However, St. John’s Wort makes little sense as a substitute for SSRIs when it essentially is an example of the closely related drug class of SNRIs. Pharmacological investigations suggest that it is a modest inhibitor of monoamine oxidase and a powerful inhibitor of the reuptake of serotonin, norepinephrine, and dopamine. It may also activate the sigma-1 receptor, a powerful mitochondrial effect exhibited to varying degrees by the different SSRIs. By blocking the cellular uptake of serotonin, we would expect it to have all the same effects on mitochondrial function as SSRIs broadly have.

Indeed, this herb might have a similar withdrawal syndrome as SSRIs do. One case was reported where a 58-year-old postmenopausal woman used 1800 milligrams three times a day for 32 days. She started developing photosensitivity, so stopped using it. Within 24 hours, she developed nausea, anorexia, dry retching, dizziness, dry mouth, thirst, cold chills, and extreme fatigue. These problems peaked on day 3, started getting better on day eight, and led to nine pounds of weight loss.

In a telephone survey of 43 individuals who used St. John’s Wort, thirteen developed withdrawal symptoms when they tried stopping. Nine had to restart it to make them go away, one had them resolve in a week, two in two to three weeks, and one was still suffering from them at the time of the survey.

Twenty of them reported side effects, four reporting photosensitivity, eight gastrointestinal upset, ten sleeping difficulties, two fatigue, seven anxiety and restlessness, and one reported sexual dysfunction. Among five who were using it in conjunction with other serotonin-modulating medications, two developed serotonin syndrome.

39 of them used the herb alongside aged meats and cheeses or over-the-counter cough and cold medications, with two of them acutely developing flushing and pounding headaches from the combination. One blacked out after combining it with alcohol.

St. John’s Wort traditionally had many uses, and in some areas it was used against the melancholy of winter, probably an early name for seasonal affective disorder.

The rationale was that it flowered on the summer solstice at the time of the feast of St. John the Baptist. Since winter caused depression from lack of sunlight, this yellow flower captured sunlight from the summer solstice that could be administered during winter.

Brain serotonin levels are primarily determined by exposure to light. This lines up well with serotonin’s primary role being to support mitochondrial function during hypoxia because when we rise from sleep we are instantaneously experiencing “relative hypoxia” until our mitochondrial function can rise to adapt to the higher metabolic demand of wakefulness.

If St. John’s Wort indeed activates the sigma-1 receptor, it would be expected to increase brain serotonin and thereby act similarly to sunlight.

The fact that one of its principal adverse effects is photosensitivity also lines up with this: by providing too much biochemical similarity to sunlight, one overdoses on the sunlight.

These collectively support the traditional concept that the flower captured sunlight as it bloomed on the summer solstice and then delivered it during winter to those who consumed it.

The photosensitivity of St. John’s Wort can include sun-induced pain, paresthesia, and rashes.

SSRIs can cause various forms of photosensitivity as well, including hyperpigmentation, epilepsy, and skin rashes.

Since photosensitivity is shared across these substances, it probably results from their main mechanisms of action regarding serotonin synthesis and transport.

This might be due to driving tryptophan into the synthesis of serotonin and away from the synthesis of NAD+. Sunlight exposure causes DNA damage, which requires the irreversible loss of NAD+ to repair, thereby synergizing to aggravate NAD+ deficiency in the tissues most exposed to sunlight, such as the skin and the eyes. This principle underlies the photosensitivity of classical pellagra expressed as the “dermatitis” of the “three D’s” (or four, if you include “death”).

Alternatively, elevated catabolism of serotonin to 5-hydroxyindoleacetic acid (5-HIAA) could cause photosensitivity because this compound becomes cytotoxic when exposed to light.

If we put the St. John’s wort data into historical perspective, it seems that traditional flower concoctions probably had lower doses, and use for the melancholy of winter limited the duration of use to twelve weeks.

This is in gross contrast to the modern diagnostic framework that identifies people as depressed instead of identifying “melancholy” as something that occurs due to “temporary deficiency of sunlight,” that demanded a temporary fix. The modern treatment framework then puts people on SSRIs for years, which has no basis whatsoever in the randomized controlled trial literature and is tied to horrific imprisonment of people with the threat of severe and protracted withdrawal.

While randomized controlled trials should indeed be used as the gold standard for inferring causation between a strategy and its outcome, they should not provide the boundaries of safe use in most cases except to potentially restrict them beyond the traditional use of an herb.

While the pharmaceutical-medical industrial complex claims to follow randomized controlled trials to make treatment decisions, this is mostly fake. For example, SSRIs are mostly studied up to twelve weeks but SSRIs are on average prescribed for five years.

Ironically, the medieval European use of St. John’s Wort better conformed to the boundaries of modern randomized controlled trials without even knowing about them, since its use lasted for one season rather than being used indefinitely for years.

Since St. John’s Wort has equivalent efficacy and superior tolerability to SSRIs, it is better to use St. John’s Wort than SSRIs.

However, since it does work in much the same way and it does pose a similar risk of withdrawal-induced mitochondrial dysfunction, it should be used toward the lower end of the effective dose range and for no longer than twelve weeks. It should be used only if still needed after addressing nutritional deficiencies, idiosyncratic mitochondrial dysfunction, and the low-hanging fruit of depression prevention such as boosting protein intake up to the range that powerfully protects against depression.

This is exactly how I have positioned it in the Depression Protocol.

While this relates to much more than tryptophan, getting enough tryptophan is certainly part of it, so we start by looking at the randomized controlled trials of this one powerful little amino acid.

In the first of two tryptophan trials, 115 subjects were randomly allocated to one of four treatments: three grams of tryptophan per day (in three divided doses of one gram each), 150 milligrams of the tricyclic antidepressant amitriptyline (in three divided doses of 50 milligrams each), a combination of both of those treatments, or to placebo.

All treatments outperformed the placebo group, with tryptophan-only performing identically to amitriptyline-only, and their combination modestly outperforming each monotherapy.

The number of subjects withdrawing due to side effects was twice as high in the amitriptyline-only group (7) than in the tryptophan-only group (3), but was zero in the placebo group and was otherwise lowest in the combination therapy group (1). This gives the impression that amitriptyline is considerably more dangerous than tryptophan when used alone but together they each protect against each other’s respective side effects.

Since tryptophan was just as effective as amitriptyline but had less than half the rate of serious side effects, tryptophan outperformed amitriptyline from a cost-benefit standpoint.

I propose the following for why they both protect against each other’s side effects.

Amitriptyline inhibits serotonin transport like SSRIs do, so probably has largely similar effects on serotonin. It could be depleting circulating serotonin and whole-body serotonin, and extra tryptophan could be making up for this by allowing re-synthesis of intracellular serotonin. By diverting tryptophan into serotonin synthesis, moreover, it is inducing a deficiency of tryptophan itself. Since tryptophan is the precursor to niacin, NAD+, and NADPH, amitriptyline could be hurting mitochondrial energy metabolism and antioxidant defense by inducing a deficiency of tryptophan and its derivatives.

Meanwhile, excess tryptophan without the amitriptyline could be metabolized into quinolinic acid, an excitatory neurotransmitter that carries the risk of excessive neurostimulation or excitotoxicity, and to glutaryl CoA, an intermediate in the catabolism of tryptophan that is toxic to mitochondrial energy metabolism and creates an inflammatory stimulus to help with its detoxification.

By increasing the demand for tryptophan, amitriptyline effectively lowers the dose of tryptophan below that which would cause accumulation of quinolinate and glutaryl CoA, and by increasing the supply of tryptophan, tryptophan protects against amitriptyline-induced deficiencies of whole-body and intracellular serotonin and of niacin and its derivatives.

If this explanation is correct, it suggests that 3 grams of tryptophan is an overdose, and that a lower dose such as 1-2 grams per day might be just as effective and safe as the tryptophan/amitriptyline combination.

The remission rate was 81% in the combination group. Thus, over 80% of depression might disappear with simply optimizing tryptophan status!

In the second tryptophan trial, 30 subjects were randomly allocated to fluoxetine (Prozac) with tryptophan or a placebo, but no one was randomized to a placebo-only or tryptophan-only group. The tryptophan was titrated up from one gram to four grams over the first four weeks. Everyone received 20 milligrams of Prozac daily. The trial lasted eight weeks. The two treatments performed identically, decreasing average depression scores by about 40%. Five people dropped out of the Prozac plus tryptophan group and four dropped out of the Prozac-only group, but most of these were unrelated to treatment. The one exception was that one person in the Prozac-only group developed new-onset psychosis.

The second tryptophan trial is not very relevant to our analysis as it didn’t examine the impact of tryptophan without a drug, and if the first trial overdosed the tryptophan the second trial overdosed it even more. The most robust point here is drawn from the first trial, which is that tryptophan alone works just as good as antidepressant drugs and is safer.

Tryptophan is the precursor to 5-HTP, which is the precursor to serotonin, so 5-HTP trials are also relevant.

A small trial of 5-HTP from 1972 with only ten people in it found that three grams of 5-HTP per day for three weeks increased the probability of improvement 13-fold as calculated in a meta-analysis, but what the raw data show is that three people improved in the 5-HTP group and none in the placebo group. Several earlier studies reviewed in this paper showed no effect of much smaller amounts of intravenous 5-HTP.

More recently, 60 patients were randomly allocated to eight weeks of 5-HTP or fluoxetine (Prozac). The 5-HTP was titrated up from 150 milligrams per day to 400 milligrams per day in three divided doses over the first four weeks, while the fluoxetine was titrated up from 20-40 milligrams per day over the same period. Both treatments decreased average depression scores by close to 70% and were statistically indistinguishable from one another.

A similar trial of 69 patients showed 100 milligrams of 5-HTP performed identically to fluvoxamine over six weeks, with average depression scores getting cut more than in half.

While 5-HTP may seem “similar” to SSRIs in that it is a precursor to serotonin, it is grossly misunderstanding SSRIs to consider them similar.

First, SSRIs massively deplete circulating serotonin and most likely massively deplete whole-body serotonin outside the brain and gut. SSRIs that have poor activation of the sigma-1 receptor even deplete brain serotonin.

Second, SSRIs block the entry of serotonin into the cell, which is absolutely critical to its function, whereas 5-HTP gives the cell what it needs to produce serotonin within itself.

Third, a precursor does not force serotonin to be at the synapse, whereas an SSRI does. Thus, the precursor will not have anywhere near the risk of dependence on the unnaturally high concentration or the unnaturally long duration of synaptic serotonin that a drug can cause.

Drugs do not normalize anything. They make things abnormal in ways that nutrients — which are just as effective or more effective — cannot do.

Nevertheless, it is much more natural to obtain tryptophan from whole-food protein than to take 5-HTP.

The tryptophan and 5-HPT trials did not examine the dietary intake of protein, which is a pretty unforgivable oversight for trials examining the impact of supplementing with an amino acid or its derivative.

There is some variation in the tryptophan content of food proteins, but as long as you consume enough total protein and don’t count tryptophan-poor collagen toward it, you will get enough. The average tryptophan intake in the US is 826 milligrams per day, 977 milligrams for men and 679 milligrams for women. However, the bottom 10% only get 502 and the top 10% get 1410 milligrams, which is almost three times higher. Self-rated depression goes down with increasing tryptophan intake, but most of that association is found at lower-than-average tryptophan intakes, indicating that many depressed people will need an extra 300 milligrams of tryptophan per day.

Total protein intake and protein quality are far more sensible to use as a primary analysis because neurotransmitters other than serotonin are important to depression but derived from other amino acids. For example, dopamine and norepinephrine are derived from phenylalanine and tyrosine. The ability to methylate dopamine is supported by methionine.

Observationally, each 10% increase in calories from total protein is associated with a close to 40% lower risk of depression, up through at least 25% of calories.

Using average reported calories of 1900 per day for women and 2500 for men, the 25% figure works out to 118 grams per day for women and 156 grams per day for men.

Percent of calories is a ridiculous and biologically irrelevant way to analyze protein, which likely made the correlation weaker than it would have been if adjusted for lean body mass or weight. Therefore, the average woman should consume at least 118 grams of non-collagen protein and the average man should consume at least 156 grams of non-collagen protein per day to minimize the risk of depression.

These should be adjusted by body mass based on 2.2 grams of protein per kilogram bodyweight on the basis of the classical “standard reference man” representing an adult male who hasn’t been as subject to the obesity epidemic as is the average modern man.

Thus, the 118/156 figures or their rounded counterparts of 120 and 160 should be used as a rule of thumb, but if your bodyweight is substantially different than average due to your height or lean mass or if you are adjusting for children, use 2.2 grams of protein per kilogram of ideal body mass. Use lean mass if you have an unusually high lean mass for your height and you have an estimate of it from a DEXA.

Collagen is important to consume, but is unique in its low quantity of tryptophan, so collagen should be consumed over and on top of this minimum intake of non-collagen protein just described.

Some people report these high protein intakes causing insomnia or some other similar problem. If high protein intake causes problems with insomnia, anxiety, mania, or anything else that looks like overstimulation, the appropriate thing to do is use the Sulfur Protocol, but the Depression Protocol does the heavy lifting by extracting the most useful strategies from the Sulfur Protocol to deal with these issues.

It also covers contingencies for methionine/glycine balance and glutaryl CoA accumulation.

Tryptophan is the precursor to 5-HTP, and comparing the trials above would imply they are equally effective. 5-HTP is converted to serotonin, and serotonin is converted to melatonin, which is not only needed in circulation to promote healthy sleep but is needed inside mitochondria to act as an antioxidant and to support energy metabolism during relative hypoxia.

The ability to support 5-HTP synthesis from tryptophan is complicated by a large number of nutritional cofactors, and while it does not occur in the mitochondria, it does depend on mitochondrial energy metabolism.

The ability to convert 5-HTP into serotonin is far more simple and only requires vitamin B6.

The conversion of serotonin to melatonin is dependent on vitamin B5 (pantothenic acid) in the form of coenzyme A and methylation.

So, it is possible that someone could get enough tryptophan from dietary protein but could still have their ability to make 5-HTP limited by any number of nutrient deficiencies. It is best to ensure adequate total protein intake and resolve any micronutrient deficiencies using the Comprehensive Nutritional Screening and any deficits in mitochondrial ATP production using your personalized Mitome protocol prior to resorting to 5-HTP, but 5-HTP could be tried if there is a suspected genetic problem with its synthesis or if it proves beneficial even after resolving all of these issues.

However, the trials above support equivalency between tryptophan and 5-HTP for the average depressed person, indicating that the average depressed person simply needs more tryptophan instead of 5-HTP, and most of them could probably get the tryptophan they need by eating more protein.

Eating more protein would be the ideal solution because it would support the other amino acid-based neurotransmitters and the methylation system alongside optimizing tryptophan status.

120 grams of protein is 480 Calories, so the average woman who consumes 1900 Calories has plenty of room for this.

160 grams of protein is 640 Calories, so the average man who consumes 2500 Calories has plenty of room for this.

The way to eat more non-collagen protein is to eat more animal flesh, dairy, eggs, and legumes. Even on a vegan diet you can hit these targets with lentils or tofu, and it is all the easier on an omnivorous or carnivorous diet.

This is a powerful baseline from which to launch any strategic resolution of depression.

The problem with this is that there are all kinds of things you could test, and most of the things that actually move the needle beyond the standard of care that the average person gets from mainstream medicine aren’t covered by insurance.

Another problem is that so many labs can be poorly interpreted that it is easy to get misled into making the wrong health decisions.

So how do you juggle decisions of when to test things when you need to balance your budget?

How do you make sure your testing reflects the root cause of your issues and what can be done about them, instead of just telling you whether you have a problem?

And how do you ensure everything gets interpreted properly?

Here we cover the three most important things to test to get game-changing health results.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

When Is the Right Time to Test?

The right time to start testing is when you are already doing what you know to be best, and you are still hitting a point where you need further improvement, yet you are at a loss about which decision to make next.

Thus, focus first on What Everyone Should Be Doing For Their Health.

In my Optimization Framework, the next thing to do is assess your progress. If you have a specific problem, address the known low-hanging fruit. That is, try some things that are well known to work for large numbers of other people to see if they also work well for you. If you still need improvement after that, or if you simply want to optimize, you then enter a rigorous optimization cycle that includes comprehensive measurement of everything relevant.

If your goal has highly specific signs of success, and if you have the patience, then you should follow through with any low-hanging fruit, then should start testing once you reach the limits of that approach.

However, if you want faster answers, you can speed ahead to testing once you have made sure you’re mastering the basics.

Nutritional Testing

Making sure you are getting enough of each vitamin and mineral should be the foundation of any testing you do, because everything in your body is either something you ate or something you made from what you ate.

Macronutrients (protein, fat, and carbs) are just as important as micronutrients (vitamins and minerals), but assessing whether you are getting enough protein is fairly straightforward and your body can very flexibly deal with different ratios of fats and carbs and there are no well-validated tests to tell you if you are currently eating too much of one or the other.

By contrast, each vitamin and mineral has a requirement that can be characterized as an population average and some variation around the average where 95% of people will have a requirement within two standard deviations of it. Since 5% of people will be outliers with very different requirements and most people differ from the average, testing the assortment of very well validated markers of nutrient status is the best way to determine whether you are personally meeting your requirement.

The probability you are missing something is high.

When scientists ask people in the United States what they eat, a whopping 93% fail to consume enough of at least one vitamin or mineral [1].

It looks even worse when we break it down by specific nutrients:[2]

• Over 90% of people don’t get enough choline, a nutrient that protects against fatty liver disease and supports learning, memory, and relaxation.

• Over 60% don’t get enough of vitamins K, E, or D, needed for healthy bones and to protect our tissues from wear and tear as we age.

• Over 40% don’t get enough magnesium, needed to fuel every process in the body.

• Over 30% don’t get enough calcium, needed for healthy bones; enough vitamin A, needed for healthy eyes, skin, and sleep; or enough vitamin C, needed to protect against heart disease and cancer.

The best approach to nutritional testing is to use my Comprehensive Nutritional Screening. You then combine this with Testing Nutritional Status: The Ultimate Cheat Sheet for interpretation. If the labs in the comprehensive screening are not available to you, you use the Appendix on page 81 of the Cheat Sheet, which lists the markers you would need to assemble a comprehensive screening from scratch based on the labs that are available to you.

The core of this testing is as follows:

• Serum or plasma concentrations tell you whether a nutrient is present in the diet and being effectively absorbed.

• Intracellular concentrations tell you whether a nutrient is getting into your cells effectively.

• Functional markers tell you whether the nutrient is doing its job.

If you measure functional markers without concentrations, you don’t know if the job isn’t done because there’s no vitamin, because it didn’t make it into the cell, or because there’s another problem. Or, if you have the concentrations but not the functional marker you may verify that there is enough nutrient present but completely miss that there is some other problem such as a genetic impairment that is raising your need for the nutrient. By contrast, if you have all three, you can integrate the information to make a robust conclusion about whether a nutrient’s job is getting done, and, if not, why not and what to do about it.

Glucose, Ketones, and Lactate

I use a NovaBiomedical Lactate Plus for finger-prick lactate, KetoMojo for finger-prick glucose and beta-hydroxybutyrate (ketones) measured at home.

Elevated glucose is considered the hallmark of diabetes, which afflicts over a half billion people worldwide.

This is actually somewhat misleading, because diabetics don’t just have high glucose. They have high ketones, high lactate, high free fatty acids, and high triglycerides too.

Diabetes itself is very unlikely to be a single disease, and is instead a generic hodge podge of hundreds of unique problems in energy metabolism that all converge on a fundamental impairment in effectively turning food energy into cellular energy.

If you find out at an annual checkup that your glucose is high, this tells you that you may be on the path toward this condition, but it doesn’t tell you anything about why or how to change course.

While there are many types of low-hanging fruit that can help, such as optimizing your body composition and exercise program, being able to test your energy metabolites at home allows you to test what they respond to, which then tells you what you can do.

Once you move the measurement from your doctor’s office to your home, you now empower yourself to take charge.

You can run experiments on a daily, weekly, or monthly basis to figure out exactly what is working, discarding the things that don’t work and leaning hard into the things that do.

While continuous glucose monitors (CGMs) are currently trending, until there is a continuous glucose-ketone-lactate meter on the market I advocate using finger prick measurement of glucose, ketones, and lactate simultaneously.

The reason is that ketones and lactate add context to the glucose measurement to help its interpretation. I would rather have fewer measurements with more context and better interpretation than more measurements with less context that are harder to interpret.

CGMs also run the risk of putting way too much focus on postprandial excursions outside the timeframes that are well validated as meaningful and can also facilitate people running away from glucose spikes without giving their bodies time to adapt to varying carb loads, which are very well established to require weeks of adaptation.

For interpreting glucose, ketones, and lactate, use my protocol dedicated to that purpose here:

Mitochondrial Respiratory Chain Testing

As I wrote in Energy Metabolism Governs Everything, every single health issue is profoundly influenced by the ability to produce cellular energy in the form of ATP.

The respiratory chain of the mitochondria is where you use oxygen from the air you breathe to produce 90% of this ATP.

You can think of the mitochondrial respiratory chain as the cell’s power plant. In turn, creatine forms the mitochondria’s energy grid, carrying the energy produced in the mitochondria throughout the cell, and glycolysis forms a local power generator.

Testing the respiratory chain can tell you why your glucose, ketones, or lactate are abnormal.

Those may even be normal, but you might not be making enough ATP to invest in things like utilizing your cholesterol instead of leaving it to run high in your blood; making your gut motile enough and your nutrient absorption efficient enough to have well formed stool instead of constipation or diarrhea; or anabolic tissue synthesis to support healthy skin, bones, teeth, and gums.

It can also explain why you might not be oxidizing NADH, an important energy carrier, which could have many knock-on effects like hurting your activation of vitamin A and thereby making you more vulnerable to getting sick frequently, developing autoimmune disease, and getting cancer.

I developed Mitome to take respiratory chain testing out of the niche field of rare disease diagnosis and reframe it to look for unique idiosyncratic bottlenecks in each person’s respiratory chain. Instead of classifying you as “normal” or “out of range” Mitome is looking for patterns that provide unique insights of how you can best optimize your cellular energy metabolism.

Most importantly, I designed all of the protocols to help you know which nutritional testing you could benefit most from and how you can use glucose, ketones, and lactate to fine-tune the protocol to your own needs.

Many people have asked what the difference is between Mitome and tests such as the Genova ION + 40 or the Mosaic (formerly Great Plains) Organic Acid Test. The ION and OAT contain many markers of the catabolic breakdown of macronutrients that happens inside the mitochondria, but they do not contain any markers of the mitochondrial respiratory chain. When you break down fat, carbohydrate, or protein in these systems, you then deliver the breakdown products to the respiratory chain. If you have a problem in the respiratory chain and in the breakdown of macronutrients, the last thing you want to do is fix the breakdown of macronutrients first, because then you will greatly increase the work you are sending to the impaired respiratory chain.

It’s as if you had traffic jam at point A and point B when everyone is traveling from A to B. The last thing you want to do is resolve the traffic jam at point A while doing nothing to fix it at point B, as you will double or triple the problem at point B and probably 10x or more the accidents and road rage.

You always want to fix a downstream problem before an upstream problem, so that when the upstream point starts flowing freely to the downstream point, the downstream point can handle it.

Mitome tests the most downstream point there is in your mitochondrial energy production, so it represents the point you want to optimize first.

Why These Beat Many Other Tests

Many other tests that conventional or functional medicine focuses on are worth testing, but should be seen as effects rather than causes.

For example, you might want to optimize your cholesterol to optimize your heart disease risk. Before we get into deciding whether we should look at total cholesterol, the total-to-HDL-C ratio, the HDL-C-to-triglyceride ratio, or ApoB, we should first acknowledge the elephant in the room: cholesterol levels are powerfully controlled by thyroid hormone and are mechanistically controlled by the activity of the LDL receptor. In fact, statins work by making your liver cells deficient in cholesterol, which makes them increase the LDL receptor without the need for thyroid hormone to do so.

As I covered in Hormones Are Never In Charge, thyroid hormone is produced in response to hypothalamic ATP content. The LDL receptor itself requires ATP to fuel the motor protein dynamin, which constitutes the molecular motor that moves the LDL receptor into the cell.

Thyroid hormone is influenced by a wide variety of vitamins and minerals. In fact, I have never seen a woman on thyroid hormone who didn’t have low serum iodine, and I find it disturbing that practitioners are prescribing thyroid hormone without measuring iodine status first.

ATP production, too, requires most of the nutrients.

Both thyroid and the LDL receptor together are fundamentally powered by ATP production in the mitochondria. For thyroid hormone, ATP production in the hypothalamus is the driver, and for the LDL receptor, ATP production in the liver is the driver.

As such it makes no sense to go on statins — which themselves are mitochondrial toxins — before optimizing the nutrients you need for thyroid hormone and ATP production and the identification of the idiosyncratic needs of your respiratory chain.

Bonus: Putting Them Into Proper Context

In order to put these into their proper context, you want to run two “bonus tests” at home:

Bonus 1: Self-Rated Health Scores. “What gets measured gets managed.” Don’t get lost in proxy markers. Rate yourself on the health metrics that matter most to you on a scale of 1-5. Make sure your health strategies are improving these and not just the lab numbers. This is the ultimate determinant that you are getting the health results you want.

Bonus 2: Track Your Diet in Cronometer. If you are doing nutritional testing, you will make much better inferences if you characterize the nutrients you are eating. This rounds out the functional markers and concentrations in serum/plasma and cells by showing you whether the nutrients are even in your diet in the first place. Use my Cronometer Guide to get this set up correctly.

Integrating both are covered in my Optimization Framework.

Negotiating Time and Money

If you have unlimited time and money, you should run all of this as a baseline and after each major turning point in the evolution of your health strategy. More data is always better so long as you have trained yourself not to get lost in the weeds or have anxious reactions to data overload.

Most of us need to negotiate the tradeoff, however, and most of us are more limited in one area than the other.

One important point, though, is “I have no time” or “I have no money” often means “this is not important enough to me to spend my time or money on it.” My own experience is that the greatest waste of time and money I’ve ever made has been in putting off testing of this nature. Time spent in less than optimal health means problems compound themselves over time. Better to invest and accumulate compound dividends than to let problems drag on and accumulate compounding debt payments.

With that said, tracking your diet is the most time-intensive step and is the part you can forego if you are trying to save time. I would highly recommend you at least do a rough outline of your diet. For example, go into Cronometer, and in a single day, estimate what you eat in a month. Generate the nutrition report for the month, and you’ll now have a daily average of your nutrient intake for the month. Whenever you change your diet, copy that day to a day in the new month, and edit it a little bit to make it more representative of the new diet. That saves a lot of time over meticulously tracking each morsel of food every day, but it still provides you a general sketch of where your nutrients stand.

If you genuinely have no money to put toward this, you are best to focus on home experiments with affordable foods. For example, how does your self-rated energy respond to increasing the amount of liver or nutritional yeast in your diet or to consuming diets run more on carbohydrate or more on fat?

If you are trying to hit a middle ground on what you spend, I would run my Mitome test and combine it with testing glucose, ketones, and lactate (GKL) for three days in a row at baseline and at the turn of each of your health strategies.

There are two reasons to test Mitome first:

• Since any problem you fix in your energy metabolism eventually dumps on the respiratory chain, the respiratory chain is what you should fix first to enable anything else to be fixed safely.

• I made the protocols with cost-saving people in mind. Rather than telling you that you have to optimize your nutrients first or telling you that you have to do testing afterward, the protocol guides you through how to implement each step while judging its effects with your home testing and rating alone and also tells you when you would be best off running additional nutritional testing alongside that approach. When the protocol recommends testing, it gives you the interpretive framework you need for the recommended tests.

Running the comprehensive testing costs about $2400, so the ability of Mitome to select out the nutritional tests that are most necessary can save money as part of the return on the initial investment into the Mitome test.

The reason to include the GKL is that it adds the highly repeatable pre- and post- home testing to evaluate how your strategies are impacting your energy metabolism. This is critical to tailoring your individualized program.

You get faster insights when you test everything together at every step of your strategy, but if you have more time than money you can invest more time in self-experiments on self-rated endpoints, and if you have more money than time you can focus more on the comprehensive screening + Mitome, with the middle ground being Mitome + GKL.

The light way to use these is to do intermittent testing to evaluate the things you do using a “try it and see if it works” approach.

If your problem or goal is worth the time, or if you are simply obsessed with optimization, you take one of these approaches and then run with it using my Optimization Framework.

If you are in an emergency, you might need to throw the kitchen sink at the problem until you have some relief, and you may be necessarily in the care of someone else.

If you are suffering from a debilitating problem, you may feel the need to throw the kitchen sink at it and may not have the mental bandwidth to optimize.

On the other end of the spectrum, when you’re getting the best results you could ever dream of, the rule that applies is “if it ain’t broke don’t fix it.”

When you aren’t in the midst of an emergency and still want to improve something about your health, the first thing you should do is see if you can truly check off all the boxes in What Everyone Should Be Doing For Their Health. If fixing some low-hanging fruit you are missing gets you what you want, you can now revert back to “if it ain’t broke don’t fix it.”

Perhaps you are skeptical that one or two of those boxes is necessary. You can actually test whether checking them off helps you with the strategy described in this optimization guide.

One reason to optimize would be that you are a data junky and love optimizing everything. Or, you’re a mad scientist and love running experiments on yourself. If that’s you, dive into this strategy with with your whole heart. You’re going to have lots of fun.

For most people, the reason to optimize is that you are doing what you know to be best practice, but there are still areas where you want to improve.

Maybe your sleep could be better. Or your energy levels. Or you want to gain ten pounds of muscle. Or you’re grappling with a biomarker you are trying to optimize for longevity.

Whatever it is, this is the guide to optimizing anything. We focus on health, but the basic principles apply to everything that could be optimized.

The key to optimizing optimally is to apply the scientific method to yourself, while artfully balancing rigor and practicality.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

Download the Ultimate Guide to Optimizing Anything

The Ultimate Guide to Optimizing Anything is a streamlined 8-page quick guide. Click here to download it.

Print This Article out and Read It In Natural Lighting

This full article is 33 pages long. You can print it out here to read it in natural lighting:

How To Optimize Anything

1.4MB ∙ PDF file

Download

Download

This Guide Isn’t About WHAT To Optimize

This guide provides an overarching framework to apply to anything you want to optimize.

If you want to start reading about what you should optimize, start with my guide to the top three tests to run for your health:

Otherwise, read on about optimization!

Decide Whether Optimization is the Right Strategy

You can use this decision-tree chart to help you figure out if optimization is the right strategy for you at this time.

“If it ain’t broke, don’t fix it,” “throw the kitchen sink at it,” and “try it and see if it works” are all important strategies to use outside of optimization.

If you are happy with your results on your current regimen, there is no reason to alter it.

If not, you might need to throw the kitchen sink at the problem or simply try something and see if it works. Throwing the kitchen sink at something makes a mess that is impossible to decipher on the back end, so it only makes sense in an emergency or a completely debilitating situation, and both of those things call for expert guidance or expert care. Low-hanging fruit could either be of the general kind, as outlined in What Everyone Should Be Doing For Their Health, or of the problem-specific kind.

There are many problems and there is no end to the number of problem-specific things you could simply try and see if they work, but an example would be that if you are unhappy with the amount of muscle in your chest and you don’t do many pushing exercises, adding some pushups or bench press sets might help you out a lot. Whether something is truly “low-hanging fruit” can often be judged in retrospect on the basis of whether “try something and see if it works” actually worked. If it is well known, though, that almost everyone else who tries a particular strategy gets success with it even if they don’t perform rigorous data tracking, that is a very good sign that “try it and see if it works” could work for you.

If you’ve gotten to this point and still want better results, now is the time to enter the optimization cycle.

Decide What to Optimize

Define clearly what you are going to optimize.

This is the dependent variable in the scientific experiment you are performing on yourself.

Measure What Matters

As once said by the late, great founder of modern management theory Peter Drucker, “what gets measured gets managed.”

Don’t get lost in proxy markers. If you are trying to optimize feeling rested when you wake up, you must rate yourself on how rested you feel when you wake up. You cannot substitute hours slept because it is easier to measure. If you are trying to increase your libido, you cannot measure your testosterone as a substitute. Be very clear about what you are optimizing, and measure it directly.

This doesn’t mean you don’t track total hours slept or testosterone. Those are quite possibly causes of how rested you feel when you wake up or how strong your libido is. The point is simply to always measure the thing you care most about.

If what you are optimizing is already in numeric form — for example, number of hours slept or fasting glucose — you don’t need to convert it in any way.

If what you are optimizing is not yet in numeric form, you need to convert it into a number. For example, each morning rate yourself on a scale of 1 to 5 on how rested you feel.

You should use a scale of 1 to 5 instead of 1 to 10 because you want to have clear definitions of each score, and it is a lot harder to come up with 10 clear definitions than 5.

You can self-rate up to five things. More becomes too cumbersome.

You can focus on optimizing one or two of them at a time.

The other three or four can be leading candidates to be optimized in the future, and if you can see wanting to optimize them in the next six months, you should start rating them now. Or, if you’ve already optimized them, you should track them to make sure they stay optimized.

If you have other things being tracked automatically and effortlessly, you can have as many as you want. For example, if you have a sleep tracking device recording your REM, deep sleep, and resting heart rate, and HRV, and it all gets tracked in an app, it doesn’t take any mental bandwidth for you to track it and there’s no need to translate it into your spreadsheet unless it is a current, past, or leading candidate for a future optimization target.

Pre-specify the definitions of your numerical self-ratings now. Spend twenty minutes thinking it out clearly. Define what a 1, 2, 3, 4, and 5 mean. Glance at this every time you rate yourself until you have it thoroughly memorized and check back with it once a week after that to make sure your “memorized” version doesn’t drift. This holds you to a consistent rating pattern and prevents your perception from drifting along with your expectations over time. It also makes it so that you never have to think too hard about how to rate yourself ever again. Just look at the definitions. If you’re not sure whether you’re a 3 or 4 today, you’re a 3.5. Done. Not having to think too hard or too long about it makes it a sustainable practice.

You can borrow someone else’s definitions from a questionnaire. But should you? Generally, no. Do this if you feel at a total loss about how to rate yourself, but your definitions are best tailored to what is relevant to your own health goals. Some institutions’s definition of fatigue will never be as relevant to you as your own definition of fatigue, because you are the only person who is intimately familiar with how your own fatigue manifests and feels.

Make your definitions so that you expect to have a monthly average of 3 when you first start tracking. Make them so you expect to see at least a few 2s and at least a few 4s over the course any given month, but you would have to get substantially worse to see a 1 or substantially better to see a 5. You can also set it so that 5 is horrible and 1 is great, whatever is more intuitive to you.

Rate yourself every day. Either at the beginning or the end of the day. Make a ritual out of it. It should take one or two minutes.

Why should you bother with these numerical self-ratings? Five reasons:

• It massively increases precision. Sometimes things change very slowly over time. If your average daily energy level each week is going from 3 to 3.1 to 3.2, it’s probably noisy enough that you can’t even tell the difference. But when you look at your average weekly data you will indeed be able to see it.

• It prevents expectation drift. If you get healthier over time, you will expect to be healthier and you will fail to be impressed by your health. If you get less healthy, the reverse will happen. When your criteria are pre-specified and easy to look at, you will keep your ratings consistent and allow you to make objective comparisons.

• It allows you to run averages. If you can’t calculate a weekly or monthly average, you are left with a vague sense that “I seem to get startled too often but it’s always going up and down.” But if you’re able to calculate that the weekly average went down from 4.0 to 3.7, you know you’re on to something that seems to be reducing your startle reflex.

• It allows you to visualize your data. Simply creating a line chart can allow you to see an upward or downward drift in what you’re measuring very easily where you’d probably miss it going on your general sense alone. Or if you test something in each of several conditions, you can make a bar chart that allows you to quickly get a sense of whether your variable is having an effect.

• It allows you to make statistical comparisons. There are easily available, easy-to-use online statistical calculators where you can copy and paste your data to see if it’s statistically significant. You can do this with numbers, but you can’t do this with anything else.

Make Your Measurements Every Day

Whatever you are rating should be rated every single day.

Anything you can measure at home must be measured every single day.

I have seen some people advise their clients to weigh themselves weekly so they don’t get confused and discouraged by random daily fluctuations in their weight. This is terrible advice. All it does is make your data seven times less useful.

If you take this approach, you will not know if your weight is going up or down at all for three to four weeks, and it will take you three months to know how fast or slow you are using weight.

If you have come to the point where you have decided that optimization is right for you, you need to train yourself to take the data daily, but focus on the averages.

If you are an expert optimizer, you might play around with single-day experiments, but in general you should ignore single-day data and focus on averages over time.

Never Second-Guess Your Data

I have seen many people take a glucose measurement. It’s too high! They take another. It went down. Oh good, it wasn’t as high as the first number said it was.

You cannot do this. Accept your data as-is.

You must always pre-specify your approach and then stick to it.

Why? Because violating what you have pre-specified introduces bias. If you take a finger prick glucose measurement and it is 99, and then because it is higher than you like you take another one and it comes in at 94, you have turned an unbiased measurement into a biased one.

The bias is this: you never took a second measurement because you liked it too much or it was too normal; you only took a second measurement when you didn’t like the first. That bias will lead you to overestimate how normal your measurements are and underestimate how abnormal they are because you are only second-guessing the abnormal ones.

If you suspect you have a problem with your data collection instrument, the solution to this is to perform a separate validation experiment with calibration fluids or with multiple pricks of different fingers or comparisons to ear lobe data or whatever you want to test. It is not to second-guess and repeat just the measurements you didn’t like.

I have also seen some people decide to do three measurements and take the average. While there is nothing wrong with doing this, you have to realize that if you are optimizing something you are going to have to take a lot more than three measurements. You may well increase your precision by taking the average of three measurements, but it may not be worth your time when you are going to have a larger number of measurements across time to take the average of. Regardless, if you do take the average of three measurements, you have to do it every day, or in a pattern that you have pre-specified, not just on days where you don’t like the first measurement.

Measure What Matters: Example

As an example, I track my energy at the end of each workday.

I organize my tracking to be part of my workday. On rest days, I rest from tracking. As such, my definitions are centered around the interaction between my physical fitness and my work productivity.

My ratings assume that I was able to keep up with my workout volume, but this part is circular, because I design my workout volume around what I can tolerate without experiencing a reduction in my work productivity.

I rate myself at the end of each day.

Here are my pre-specified definitions:

1 — I struggled to find the motivation to work.

2 — I got a solid six to ten hours of work in, but much of it was spent lying down.

3 — I got a solid six to ten hours of work in, and all of it was upright, but not necessarily active.

4 — I got a solid six to ten hours of work in, and at least 2-3 hours of it was spent in activity. For example, I worked while walking at 1.4 miles per hour on my treadmill, or in an active lunge with a straight-side glute contraction and a flexed-side hamstring contraction, or in a half lotus (which is a serious hip stretch for me), or in some other position.

5 — I got a solid six to ten hours of work in, and nearly all of it was spent in activity, and two to three hours of it were spent in double-activity. An example of double activity would be walking at 1.4 miles per hour on my treadmill while also doing a core contraction and practicing breathing into my upper back or doing a single side stretch of my shoulder, arm and fingers while the other hand controls my scrolling during reading.

My experience is that if my energy level is abundant, I can do many things with my body with plenty left over for my mind. If my energy level is scarce, doing any type of activity is extremely distracting to my mind. So, I prioritize putting myself in the physical position that allows my cognitive work to get done, and I rate my energy by how conservative I had to be or liberal I was able to be with my physical activity while getting that cognitive work done.

You may have a very different work style or a very different level of energy, so you might pick a totally different set of definitions.

The point is to pre-specify the definitions that are relevant to you, write them out clearly, and hold yourself to them.

How to Organize Your Data

Make a spreadsheet in Google Sheets, Excel, or a suitable alternative.

Each day is a row. Make a column for each measurement, and label the column appropriately. Freeze the top row so that the labels are always visible even if you have many rows of data.

Put the definitions for your self-rated tracking items in the top row so that they are easily accessed every time you rate yourself. You can either hide them in the column header of that item so that you view them by double clicking, or you can put them off to the right side so that you can scroll horizontally to see them.

In this example, the width of “Energy” is adjusted so that you can only see the column title without double clicking. Once you double click, it opens up the definition. This was made simply by putting the definition right after the column title but not selecting any text wrapping and adjusting the column width so that only the label could be seen.

In this example, by contrast, I put the definition of energy on the right-hand side, wrapped the text, and widened the column enough for it to appear in paragraph form.

Either of these are fine and you could in fact combine them if you wanted to. The point is that you need the definitions very precise, to the point, and easy to check so that you can hold yourself to them over time.

Major protocol changes, such as changing the dose of something or adding something, should be added intermittently in their own column.

Your protocol changes are the independent variables of the scientific experiment you are performing on yourself.

You will likely want a “notes” column at the end that serves as a journal for anything notable that doesn’t fit in another column. However, you have to be conscious that if you write a lot in the notes section, you aren’t going to read it very often.

—> Anything important as an outcome should be a number, not a note.

—> Anything important as a protocol change should be six to seven words max, and should be infrequently noted because you shouldn’t be changing your protocol every day.

For example, “increased creatine from 3g to 4g” might be in the Protocol Changes column, and two weeks later you might write “increased creatine from 4g to 5g.” Or you could simply call them “Creatine 3g,” “Creatine 4g” and “Creatine 5g.” You can then compare, say, your average sleep on 3 grams versus 4 grams versus 5 grams in about 30 seconds, but you could never interpret all your daily notes in 30 seconds.

The main purpose of the notes is to look through or keyword search in the future if you ever get stuck on an interpretation, or if you fundamentally change how you view something you had been recording in your notes and need to collect all the instances you mentioned it. Or you can add what you were thinking that made you change your protocol in case you forget why you did something. You might also list notable confounders in this section so that if your primary interpretation of something doesn’t pan out you can go back and look at the confounders to see if they help reveal a secondary interpretation.

The main point is this: take the notes, but consider them extra, not something constituting your primary data to analyze.

The Ultimate Optimization Spreadsheet

Click below to download the Ultimate Optimization Spreadsheet:

The Ultimate Optimization Spreadsheet 2025

96.5KB ∙ XLSX file

Download

Download

Do not open it in Excel or any other desktop program. Immediately drag and drop it into Google Drive and double click to open it as a Google Sheet.

Otherwise, features will be lost.

Make the following edits:

• Change self-rated and extracted variable names to reflect the variables you are measuring.

• For self-rated variables, replace each one’s name both in its data column as well as in its definition cell in the upper right, then fill in your succinct and precise definitions for 1, 2, 3, 4, and 5 for each of them.

• Click on the top chart. Click … in the upper right corner, edit chart. Then click “include hidden/filtered data.” Keep the chart editor open and do this for the rest of the charts. It will now only require you to click once on the chart and then click once on “include hidden/filtered data” and keep doing this until you do it for the eighth and final chart.

Now all you have to do is enter your data.

Hidden columns will automatically calculate and graph your daily data and the 3-day, 7-day, and 30-day rolling averages, and your protocol averages will be assembled into a table on the right-hand side based on when you enter protocol changes into column A.

Here is a video to help orient you to it:

At 18:43, it explains how to add extra variables and keep the features if you need to track more than eight.

Characterizing Your Background Data

Somewhere you should have your entire regimen characterized. For example, your diet and supplements can be tracked in Cronometer and your workout could be tracked in an app like Jefit or, if you work with a trainer, Everfit.

If you run lab data that is impossible to run daily, such as the data that comes with an annual checkup or your own self-organized quarterly lab testing, this can live in the lab portal. For example, Quest now does a great job keeping track of your results and formatting them as change-over-time line graphs.

If you are testing dietary components, you really should consider Cronometer tracking mandatory. If you don’t want to put in the time weighing out your food and tracking it meticulously each day, you should just sit down for ten minutes and log what you consider to be your likely intake of foods over the course of a month into one single day, and generate the nutrient report for the entire month, giving you at least a general idea of what your micronutrient intakes average to over the course of a month.

Your optimization spreadsheet is for analyzing your primary outcome data (your dependent variables and your primary protocol changes (your independent variables) and does not need to have the totality of your regimen, as long as the totality of your regimen is easily accessible in one or several other places, you do not need to track it all in your spreadsheet.

However, it should be available in such a way as to be able to easily characterize it for any time period that is relevant, and to be able to extract data if you decide they are relevant enough to include some portion of them in your spreadsheet for ease of comparison or visualization.

If you are a data junky, feel free to extract all of that background info into one giant spreadsheet, but this should not be in your optimization sheet. It should be a different file altogether or it should be its own tab within your health spreadsheet, with the optimization sheet being the main tab you use.

You should extract a variable from your background data into your primary spreadsheet if you believe it may be either a) an intermediary between your protocol and your outcome, or b) a confounder that complicates the interpretation of your protocol’s effects providing it also is a) variable enough that it is worth tracking daily and b) irregular enough that it doesn’t simply average out.

For example, say you are optimizing your sleep and are testing creatine. Your carb intake might not be part of your experiment, but it might vary quite a bit from day to day and you might become convinced its variability is influencing your sleep, acting as a confounder of the creatine. You may then want to be able to extract it from Cronometer into a column of your spreadsheet.

However, if you keep your carb intake constant enough that its influence is muted, there is no point in extracting it to your spreadsheet, and if its ups and downs are so regularly cyclical that they average out over the relevant timespan — for example, your protocol steps are at least one week long and your average daily carb intake is roughly the same from week to week — then there is also no point in extracting it.

Similarly, you could be testing whether black seed oil increases your energy, and you expect that it could do so by lowering your inflammation, shifting iron from storage into the synthesis of energy utilization pathways. In that case, you should extract your annual or quarterly iron measurements into your optimization sheet, and should strongly consider measuring them more frequently to help optimize your dose of black seed oil.

If you can keep your background variables more controlled, it will help you get quicker results in your experiments.

For example, I eat the same thing every day weighed out to the gram on my workdays, and I relax my diet on my rest days. This allows me to change one thing in my diet and clearly distinguish its effect since nothing else in my diet is varying.

You do not have to do this.

The solution to greater background variability is to collect more data. Since you are only a single person, the only way to collect more data is to run the experiment for a longer period of time, giving you more days to measure.

You should assess the extent of this background variability. For example, if you eat a rotation of foods and the rotation is fairly consistent, then your intake will average out over the period of your rotations. If you rotate seven different dinners across a week, your weekly average intakes will even out. If you are less deliberate but do not have infinite appetite for variation, you are probably averaging out at least on a monthly basis.

If your dietary variability is clearly adding noise to the variable you are tracking, simply extend your experimental trials so that they span the spectrum of variability. If your diet tends to repeat week after week, make your dietary interventions so that each trial lasts at least a week. If your diet tends to repeat month after month, make them last a month.

If you are optimizing something that is clearly influenced by your menstrual cycle, you are going to need each experimental trial to last through at least an entire menstrual cycle.

Decide What to Test

Having chosen what to optimize as an outcome and having set up your data tracking correctly, you now must choose what to test as an intervention. This could be one of the steps in any of my Protocols Series, or it could be something from ancestral traditions, something someone recommended, something that you found a scientific study supporting, or something you came up with yourself based on biochemical reasoning or intuition.

Make a very brief description of the intervention you are testing and include it in the “Protocol Changes” column of your sheet.

Make sure to also log it wherever you are keeping the full regimen of its category logged. For example, if it is a dietary or supplement strategy and you track your diet and supplements in Cronometer, make sure to also log the change in your Cronometer.

Test One Thing At a Time

If you test more than one thing at a time, you will not know which thing is responsible for any effect you observe.

Isolating your changes to one intentional change at a time does not automatically give you a robust ability to determine causation, but it does remove the major barriers to such a determination that are under your control.

As covered in the Background Data section, you will still have many confounders operating in the background. Your job is to carry out each trial long enough to let these confounders average out so that your signal rises above the background noise.

This does not mean you have to test one isolated nutrient at a time. Testing a multivitamin, a B complex, or an infinitely more complex whole food, is a completely valid test. However, you must make your inferences about the things you are testing. For example, adding pizza is testing “pizza,” not “carbs.” Adding a Thorne Basic 2/Day Multivitamin is adding that specific multivitamin, not just any multivitamin and certainly not one of its nutrients. The point is, you are not adding a multivitamin and adding pizza to increase your “carbs” and then trying to determine which one is causing the subsequent changes in your health that you observe.

If you add pizza and you hypothesize that its effects are due to the extra calcium in the cheese, and you want to be able to know if that is true, you can run subsequent trials for other forms of cheese providing the same amount of calcium, or for other food sources of calcium such as bone meal, or for isolated calcium supplements. You do not need to test these things, but if you want to be able to know it is “calcium” that produced the result, you will have test them to properly form that conclusion.

There are many synergies involved in nutrition. They are complex and require expertise to understand. This can be self-made expertise. It does not require a degree to understand, but it requires considerable study.

In my Protocols, I combine items into a single step when I believe their synergy outweighs the value of testing them separately.

If you do not have adequate expertise to judge what is synergistic and what is not, it is best to defer to whole foods, supplement formulations on the market, and expertise you are able to tap into through your network or online.

As a general rule, trying to increase a vitamin or mineral through food offers a major margin of safety against violating a principle of synergy that you could very easily violate with a supplement.

Moving very slowly through dose ranges as advocated below offers another margin of safety.

You can obtain another margin of safety simply by grouping nutrients together by class. You can expect that all nutrients have some potential synergy with each other, which makes a high-quality multivitamin one of the safest supplements to try. B vitamins all have intimate interactions, so a B complex is safer than an isolated B vitamin. The groupings in Testing Nutritional Status: The Ultimate Cheat Sheet can further help you see synergies to expect based on the overarching categories into which nutrients are placed.

Whatever your intervention is, no matter how simple or complex, define exactly what it is, make your inferences exactly about it, not some fanciful assumption about the broader thing it is meant to represent, and isolate your intentional changes to that one thing. Allow your trial to go on long enough to ensure all the things outside of your control in the background average out over time.

Slowly Titrate Your Dose to the Point of Maximum Benefit

You can expect everything in life to have a dose-response curve that looks something like this:

At a low enough dose, it is doing nothing.

Somewhere after that you have some sort of dose-response curve where a bigger dose gives you more benefit.

You then experience 1) diminishing returns, where an increase in dose still gives you more benefit but the slope becomes less steep, 2) a plateau, where increasing the dose does nothing, 3) a loss of the original benefit, and then 4) an adverse effect.

The curve can thus be broken into six major sections:

There may be things in life where we do not see the full curve play out, but this is probably because we lack experience of certain doses. For example, there might be some degree of lead exposure that is beneficial through the principle of hormesis — the concept that a little bit of something bad is good for you by stimulating your adaptive defenses — but this might be underneath the lowest possible exposure in the modern environment and therefore be irrelevant. Or, we might lack the willingness to study this due to the downside consequences of being wrong about the dose.

Nevertheless, we can assume that the general principles of this curve play out across all things, and the practical point is that they are going to play out for anything that we are actually going to test on ourselves.

This could apply to resistance training and hypertrophy. Not enough stimulus, no adaptation. Within some range, more sets, more muscle growth. Then, a marginal set does give you more muscle growth, but a lot less additional benefit than the previous number of sets gave you: diminishing returns. At some point you are overreaching and then overtraining and you are stressing your body out so much that you compromise its ability to grow or even wind up in a straight-up catabolic state.

This could apply to time management. You can spend so little effort analyzing your time management that you learn nothing. At some dose, you start learning how to manage your time more efficiently and have remaining bandwidth to execute on your new insights. At some greater dose you spent so much time analyzing how you spend your time that you had no time left to actually do anything.

It applies to everything.

With a vitamin that has a toxicity syndrome, such as vitamin B6, it is more obvious that at some dose you get into “harm.” What is less obvious is that any nutrient can ultimately cause imbalances if its dose gets too high. For example, if you pound fat-burning nutrients like B2, B5, and CoQ10, you could eventually wind up in a state where you are burning more fat than your body would otherwise deem appropriate and this might hurt your skin by draining its lipid content or hurt your glucose metabolism by constantly shoving fatty acids in the way of your carbohydrate oxidation pathways.

Many people say you want to use “the minimal effective dose.”

It depends how you interpret this phrase. It sounds on the surface like it would put you here:

If you interpret this as “the minimal effective dose necessary to achieve the desired outcome,” then that would presumably put you at the beginning of the plateau point, since your desired outcome should be maximal benefit:

However, my position is that you want to arrive just before the point of diminishing returns:

The way you get there is you titrate up until you hit diminishing returns, then pull back to the most recent dose that provide a large, obvious benefit.

If you interpret “minimal effective dose” to mean “the minimal effective dose to achieve the closest thing you can to maximal benefit while also maintaining the largest window of safety from loss of benefit and adverse effect which puts you at the point just shy of diminishing returns” then by all means call this “the minimal effective dose.”

The reason you want to be at this point is that this gives you nearly all of the benefit and gives you an extra margin of safety against an adverse effect.

Think of it this way: you are climbing up a hill. When you get to the summit, you can walk around a little on the flat top. However, walk a little further and you are on the edge of a cliff. One misstep and you can fall.

The primary misstep you are likely to have made here is that your view of how long you had to spend hiking up the hill to actually know where you landed had been unrealistically short. You think by holding the dose steady you are staying in the same place, but actually you are slowly drifting toward a higher and higher cumulative body store of the nutrient over time, and slowly upregulating pathways that are either designed to get rid of the extra or use what is available toward some purpose. If you stabilize before the plateau, you have a large margin where you can slip forward and nothing bad happens. If you stabilize on the edge of the plateau, however, your margin of safety is much smaller and you are more likely to stumble your way into an adverse effect over time.

More guidance will be given on how long to spend in each trial in the next section.

First, we clarify some more points in how to determine where on this graph you are.

Your initial dose should always be the lowest expected to have any effect.

If what you are testing is a well characterized nutrient found in food, your lowest dose should be based on the dietary reference intakes characterizing the basic requirements for the nutrient and, if available, studies suggesting variation in the amount derived from eating different foods impacts health outcomes.

On the other hand, if basic requirements are not well characterized or not characterized at all, the lowest dose should be based on randomized controlled trials showing human health outcomes of particular doses when available, and otherwise from case reports or anecdotal experience.

In my Protocols, I synthesize all these types of evidence to use the best available evidence for the purpose and characterize the most probable dose range to work within to gain the proposed benefit.

However, your requirement will not always and not even usually be the average requirement derived from existing studies.

Rather, you use the available data as a starting point and then you determine your requirement by slowly titrating the dose to just shy of the point of diminishing returns that you personally experience in the context of performing this scientific experiment on yourself.

The increment of your dose increases should match your starting dose.

For example, if you start with 6 milligrams of riboflavin per day, your next increment should be 12, and then 18, and so on.

You can then modify this according to your results. Specifically, if your initial dose does nothing and you determine that you are beneath the primary benefit curve, you can speed up your dose increase until you start seeing a benefit. Once you see the benefit, slow down.

One of the problems with interpreting dose increases is that a higher dose allows you to fix a deficiency faster, but once it is fixed, the higher dose is just a bunch of extra over your maintenance requirement. In these cases, it can help to use a loading dose or what I sometimes call a “flush” to flush maximal levels of the nutrient through your body over a short timeframe.

This is well characterized with creatine, where four to five days of 20 grams per day leads to muscle saturation that otherwise takes a month of 2-3 grams per day. When you know this to be the case, you want to reach muscle saturation in either one of these ways and then use a bottom-up slow dose titration to determine your optimal maintenance dose.

The purpose of this is just to clear out any state of “deficiency” as fast as you can so that the time you spend titrating your dose is used efficiently to find your maintenance requirement.

If you add a supplement or increase the dose and nothing happens, there are essentially two possibilities.

One is you are beneath the point of primary benefit:

Another is that you are moving along the plateau point:

The only real way to know the difference is to increase the dose and see if you get a benefit or an adverse effect.

If you had started at the plateau point as your baseline, you have already accrued the benefit, so as you move on to the loss of benefit part of the curve, it will be experienced as an adverse effect.

Ultimately, you should derive the probability that one or other is true based on what is known from existing literature and human experience about the dose you are using. I continually update my Protocols to give you my best understanding of what to expect, so if you are using one, the bottom of the dose range should put you near the bottom of the primary benefit, not the top, so if nothing a happens your best assumption is that you need to increase the dose more to obtain the benefit.

However, if you are working on your own, it is easy to be misled. For example, there are many studies of riboflavin using 100 to 400 milligrams per day, but there are almost none testing 400 milligrams per day head to head against 3-5 milligrams per day and there are never any studies that give lower doses of vitamins extra time to work than the higher doses, which is what you need to do to truly show superiority of the higher dose. Go figure, what is primarily tested is never what actually needs to be known.

My current research is leading me to believe that almost no one has long-term maintenance needs for riboflavin above 5-6 milligrams per day, and most genetic factors that increase the riboflavin requirement above the population average move it from 1.6 milligrams per day to 3.2 milligrams per day.

If you decide to test the effect of riboflavin on your cramps and you have been consuming five milligrams of riboflavin per day from food for the last five years, there’s a pretty good chance that additional riboflavin will do nothing and it’s because you have already accrued the benefit of higher-than-average riboflavin intakes.

By contrast, if you are consuming the population average of 1.6 milligrams per day of riboflavin, or less than that, or you have only been consuming five milligrams per day for a month or two, there is a high chance that boosting your riboflavin gives you additional upside.

On the other hand, I cannot rule out that there are rare genetic mutations that really do require 400 milligrams per day. So I include this as an upper dose in my protocols, but also make it clear that the average person probably tops out the benefit at 5-6 milligrams per day.

Ultimately, the only way to know where you are is to continue the test to see if dose increases make things better or worse.

Another interpretive problem is that we don’t know whether an adverse effect is an intrinsic problem with the dose or is due to a contextual factor being out of balance.

For example, you may experience an “adverse effect” by fixing a problem too fast and your body needs to undergo various adjustments to allow the fix to operate in a way that provides a clean experience of a benefit.

For example, you may be deficient in creatine and this is leading to whole-body intracellular potassium depletion because you cannot efficiently use cellular energy in the form of ATP to bring potassium inside of cells. Too much creatine too fast could start fixing this, making potassium go into your cells and causing the serum levels to drop, causing heart palpitations. Your potassium supply becomes the relevant contextual factor, and you can either a) drop the creatine dose to move potassium into your cells more slowly, b) wait it out and allow your body to adjust, which might take weeks or even months, or c) supply the extra potassium so that you have enough to go around.

If there are predictable contextual factors that could be driven out of balance while fixing a problem, I include them as if-then contingency plans in my protocols.

If you are unsure, you can simply treat it as an adverse effect. If you hit an adverse effect, you pull back on the dose.

If you immediately hit an adverse effect at the lowest dose, cut it in half or a third. Keep going down to the lowest practical dose to see if you can find the “primary benefit” part of the curve. If you cannot find a dose that has benefits without adverse effects, discard the strategy.

In order to properly find the point of diminishing returns, your dose increments have to be small enough to not pop you suddenly from “primary benefit” to “adverse effect,” and you have to move slowly enough so that you determine the effect of any given dose based on its final resting point, not a snapshot of your health in motion as the dose moves you from one point to another. That means that any health changes you observe must fully stabilize before moving from one dose to the next, or from one protocol step to the next.

In the next section, we look at how to determine how long you give any dose to show its final effects.

The most important point to take away from this current section is this: start with the lowest dose that has plausible efficacy, titrate up as slowly as is practical until you hit diminishing returns, then pull back to the last dose that had a large benefit.

How Long to Spend At Each Dose

Whenever you make a change, there are biochemical, molecular, cellular, physiological, and epigenetic changes that are distributed over different time horizons. Some of these happen over fractions of a second, some over minutes, some over hours, and some over days. This is complicated by the fact that any effect has knock-on effects that are secondary, tertiary, and so on, cascading into changes with their own time horizons. At some point we hit knock-on effect layers that we don’t know enough to predict and simply cannot guess their time horizons.

Some of these could involve tissue remodeling, and these might take a longer time and be somewhat more stubborn.

The epigenetic fraction of these changes has the most variability in its time horizons. Many of these changes occur over a span of hours and ebb and flow with, for example, the fasting-feeding cycle. Others occur during development, and the window of time in which they occurred raises their perceived importance to render them permanent or at least quasi-permanent and difficult to reverse. It is likely that highly sustained signaling becomes progressively more difficult to reverse the longer it is sustained. For example, if you have a situation that persists for years, it could conceivably take years to reverse.

Effects are often multi-phasic.

For example, during water fasting, you will deplete your liver glycogen over the course of 24 hours, and during the first day you will progressively shift from using stored carbohydrate (glycogen) to using newly synthesized carbohydrate (gluconeogenesis) over time. Your ketone synthesis will ramp up, but over 3-5 days ketones will shift from feeding muscles to feeding the brain, and muscles will shift from using ketones to using fatty acids. At five days, you are solidly “keto-adapted,” a position you have reached much more quickly than people using moderately ketogenic diets because water fasting is a much stronger stimulus for adaptation. Yet, as you get weeks into a water fast you will have a secondary phase of muscle catabolism, and as you get months into a water fast you will get into a fatal third phase.

If you take something less well-characterized, like 400 milligrams of riboflavin, you might have an early phase characterized by rapidly fixing a deficiency. You then may have a second phase that lasts weeks or months where you gradually increase your ability to burn fat. This might be net beneficial until it hits a tipping point 3-6 months in where it leads to lost fat in the face, creating a more hollowed out look; eczema from inadequate skin lipids; or progressively worsening glucose tolerance as fat begins to displace glucose as a preferred fuel but your carb intake has remained as high as it always was.

The longest randomized controlled trial of a dietary strategy in existence is the LA Veterans Administration Hospital Study, which looked at the impact of substituting “seed oils” for traditional fats. The initial benefit to cardiovascular disease wore off over the course of the eight years. The rise in cancer did not even start emerging until year two and mostly took off after year 5. The total mortality was unchanged throughout the study but looked at the end like it was starting to favor traditional fats around year eight. The authors concluded the trial wasn’t anywhere near long enough to see the true effect. This trial emphasizes that dietary impacts can take years to reveal their full effects.

The point here is that there are too many relevant time horizons for us to say definitively when an effect has truly stabilized, so when deciding how long to stay on a given dose you need to artfully mix the science of dose-response with the practicality you require.

There may be cases where we already know a lot about how long something should take to stabilize. In this case, we should initially assume as our base case that we are similar to other people. For example, we know how long creatine takes to saturate muscle stores and we know how long keto-adaptation takes. If we know we should expect these lead-in phases, we should let them run their course and then start the clock looking for stabilization.

Once we are beyond any expected lead-in requirement, I recommend giving any given dose at least one week.

However, it then becomes your job to look for changes in your health and ensure they have the chance to stabilize.

A week should allow most of the first-phase epigenetic changes to settle in. While there may be other phases down the road, they tend to be months out rather than linearly accumulating. If they are linearly accumulating, you will see this clearly in your health metrics failing to stabilize. If they are months out, you will waste too much time if you wait for them. You can always revisit your protocol if something seems to be out of balance by reducing the dose of something or eliminating it altogether if it seems to have run its course.

You can protect against benefit-to-adverse-effect transitions by avoiding megadoses unless they prove absolutely necessary, by focusing on foods instead of supplements, by being slow and methodical in your initial dose titration, by never ignoring adverse effects even if they are subtle, and by aiming for the point just shy of diminishing returns instead of the absolute maximal benefit, and by not blindly copying the dose that someone else on the internet is raving about.

There are various ways to assess whether something has stabilized.

First, you want to take stock of what you already know about the variability in what you are measuring. For example, if it is your past experience that your deep sleep ranges from 1-2 hours per night and that when it hits 2 hours it often stays there for two weeks and then reverts back to the mean of 1.5, you must take that into account. If you decide that 5 milligrams per day of apigenin has increased your deep sleep from 1.75 hours a night to 2 hours a night, you cannot decide it has stabilized after one week of staying right at 2 hours a night when you know from past experience that your deep sleep often does that without apigenin. You will have to take this knowledge into account and impose a rule that your deep sleep hitting 2 hours a night does not even register as a signal above the noise until it surpasses your previous all-time best of two consistent weeks of 2 hours per night. Then you start the clock for it stabilizing.

Once you take into account your expected background variability, you can look for stabilization visually or mathematically. If you plot your data as a line graph, you can simply look at whether the line is straight, and whether it has clearly spent more time being straight than it usually does prior to having introduced your intervention or your current dose of that intervention. Or, whether it has some ups and downs but you could easily draw a straight line through the average of those ups and downs that is straighter than any previous line you could have drawn or simply is in a new, better place than any straight line you could have drawn through the data.

The way I prefer to do this is as follows:

• Determine what margin of error you will tolerate. For example, say I want my morning glucose to be 78-85. If I’ve gotten it down to 82, I can probably tolerate a span of 3 points up or down for my own satisfaction.

• Calculate the 7-day rolling average, the 3-day rolling average, and the average for all measurements taken on that particular dose. If you are carrying the experiment on for more than a month, add to this the monthly average.

• If all the averages are within the accepted margin of error of each other, see if there is a trend from longer-term averages to shorter-term averages to your most recent measurement that all point in one direction. For example, If so, give yourself a little more time. If not, you have stabilized.

• If your averages are within the margin of error but all pointing in one direction, simply keep taking measurements until they are no longer pointing in a direction.

If you see an initial adverse event, you can either immediately pull back on the dose, or you can give it three or four days for it to turn around. Never let an adverse effect keep getting worse for more than three or four days, and never let it get dangerously bad.

If you are seeing a benefit, stay on the dose until the duration and magnitude of benefit provide a clear signal that rises above the background noise and until the benefit stabilizes as assessed visually or mathematically.

Move your dose up to the next increment, and stabilize again. Keep repeating this until you reach the point of diminishing returns, and pull back to the most recent dose before that. Then you can move on to your next strategy.

The Ultimate Optimization Spreadsheet automatically visualizes your daily data alongside the 3-day, 7-day, and 30-day averages and puts your protocol step average into a table. In this case, you want to see that the averages are clustering together and flat, and that there is no recent divergence from that pattern by the daily data or 3-day average.

When to Randomize

If you arrive at some new level of health that has proven to be a stable benefit, some might question how sure you can be that your protocol is exactly what did that, but you have the benefit and that is what matters.

Suppose, however, that you initially seem to have a solid benefit and it starts to lose its stability over time.

You may start questioning whether the effect was real.

This is a good time to consider performing a randomized controlled trial on yourself.

For many strategies, especially foods and exercise, it will be impossible to blind yourself and control with a placebo, even with the help of a friend or professional. However, randomizing repeated periods will help evenly distribute the confounders across your trials and will give you a much better ability to make a conclusion about causation.

You should randomize when all of the following are true:

• You are unsure about the effect of something despite having tried it using the approach outlined above.

• It seems plausible that there is an effect that is large enough to be worth doing this randomized trial to figure out.

• There is some downside to “just doing it anyway” such as time or money you don’t want to waste, or some tradeoff effect that you are also unsure about.

• The timeline on which you expect to observe the effect, including any lead-in time required for the effect to onset and any washout time needed for the effect to stop persisting after you stop the intervention, is one you are willing to repeat some six to twelve times.

One example could be that you believe that 200 grams of net carbs per day leads to a half hour better sleep than 100 grams of net carbs on the same night you consume them, but you have about two hours of variability in your sleep that are determined by whether your kids wake you up during any given night. Every time you think you are going to see this stabilize, your kids start waking you up more, so you really aren’t sure you’re right.

In this case you expect the carb influence to act on a one-day time horizon, and 30 minutes of extra sleep is actually very meaningful, but you are having trouble distinguishing that 30 minutes from the much larger variability caused by the kids.

If you randomize your carbohydrate intake, you can effectively separate it from the influenced of the ups and downs of your kids, and do a statistical test to give greater confidence.

If your kids are acting in a multi-day cycle where any two days in a row they are acting relatively similarly, you can do a paired statistical analysis where each pair of opposite-carb days are matched with one another to increase the statistical power.

Or, let’s say you are investigating the same question but your menstrual cycle impacts your sleep even more than the carbs. You can do a paired statistical analysis where each pair of opposite-carb days occurs within a tight two-day block of your menstrual cycle, allowing the stats to automatically control for your cycle.

If you know you have a regular, cyclical variable as a confounder, you can also do post-hoc analyses where you visually inspect the data to see if, for example, carbs only add to your sleep during ovulation, or if they benefit your sleep in the follicular phase and hurt it in the luteal phase.

You will need to randomize blocks of time that allow your effect to take place. If it takes you four weeks to keto-adapt, and a keto diet benefits your sleep after the four-week time point, and it takes one week to unwind the keto-adaptation, you are going to have to randomize 7-week periods just to get one solid week of keto-sleep. You might consider this deeply impractical.

Similarly, if you are trying to build muscle, it might take you twelve weeks of a program to get any measurable sign of progress, and if you randomize pairs of training regimens it could take you years to get clear differences. This is effectively impossible, or at least a ridiculously unwise strategy compared to tweaking your training regimen with expected improvements every twelve weeks and keeping what seems to work while discarding what doesn’t.

Or you might have to make an executive decision around something that seems a bit inconvenient but could be practical if the information is valuable enough. For example, if your observational period indicated that you can consume a pound of beef per day for four days before it starts increasing your anxiety, and the effect washes out after two days of no beef, you will need to randomize pairs of one-week periods just to get one day of beef-induced anxiety. This might take a while, but if you are genuinely unsure whether otherwise seemingly random panic attacks are driven by beef consumption this might be very important to you and well worth randomizing.

Here is how to randomize:

1. First, determine whether you fit the bullet point criteria outlined above and only randomize if each bullet point is true.
   

2. Determine what your intervention and control is. If you are adding something new, your control is without it and your intervention is with it. If you are increasing the dose of something, your control is the lower dose and your intervention is the higher dose. If you are simply changing from an old protocol to a new one, your old protocol is the control and your new one is the intervention.
   

3. Determine what your primary endpoint is. This is the target you are trying to optimize.
   

4. Determine from your observational period (that is, your attempt to try the regimen and stabilize on it as outlined in the previous three sections) whether there is a necessary lead-in time (the time on the intervention until the effect onsets or stabilizes) or washout time (the time off the intervention before the effect wears off). Make the periods you are randomizing long enough to get your observed effect after any necessary lead-in time, and schedule any necessary washout periods between your randomized periods.
   

5. Determine how many pairs of periods you will randomize. As a default, I recommend doing six randomized pairs, so 12 total intervention periods.
   
   If you want to go the extra mile in setting this up for statistical success, you can use a sample size calculator. If you arrived at the randomized controlled trial because you did an observational (try it, track it, and see if it works) period first that produced confusing results, this is your preliminary data. Your preliminary data will be sequential over time rather than paired, but you are going to be randomizing pairs of trial periods. Your preliminary data from the observational period will be more fit to use for a parallel design sample size calculator, while a paired t test sample size calculator might actually be better for your trial design.
   
   My recommendation would be to use the parallel design calculator and be aware that this might recommend a higher number of pairs than you actually need. This is a good thing, however, because it ensures better stats. Use your spreadsheet to calculate the average and standard deviation of your primary outcome for your control condition and your intervention condition. If the standard deviations are considerably different, use the condition with the larger standard deviation for “group 1” and use the one with the smaller standard deviation for group 2. Fill in the average and standard deviation for group 1, and the average for group 2. If you want to err on the side of a more burdensome study to gain more statistical power, shift the power from 80% to 90% and change the alpha from 0.05 to 0.01. Otherwise leave them at their default settings. Then click “calculate.” The total number per group is the number of pairs you have to randomize.
   
   If the number of pairs recommended seems impractical, try this: pre-specify that you will analyze your data at the half-way point. If you have convincing results you will stop the trial there. If you have results that seem promising but uncertain, you will carry it on until the full number of trial periods have been finished.
   

6. Make two orders: control then intervention; intervention then control. You can randomize by flipping a coin or by using a random number generator. Assign one order to heads and the other to tails if you are flipping a coin; assign one order to 1 and the other to 2 if you are using a random number generator. Write these down before you flip the coin or use the generator. Since we have a small number of trials, I like to pre-specify that every order is used an equal number of times. With this rule, the coin or the generator is used to determine the sequence up until the point where one order has already occupied half of the slots; after this point, the other order is used for the remainder of the slots. With the rules in place, flip the coin repeatedly until you’ve determined your full sequence. Or, set a random number generator to a minimum of 1 and a maximum of 2 and hit “generate” until you’ve determined your full sequence. Write down the details of how you randomized, what rules you pre-specified, what the actual coin flips or random numbers were in what sequence, and how you translated that into your trial order.
   
   NEVER ignore coin flips or random numbers generated or keep hitting them extra times. Either your rules were pre-specified or you need to start all over if you messed something up. Once you modify the design post-hoc, you have introduced bias. Pre-specify your rules and then obey the coin flip or number generator.
   

7. Record all your data in your optimization sheet normally. Use the “Protocol Change” to note your randomized experiment and which trial period you are on in any given day. Keep track of your experimental design and results in a second tab of your sheet dedicated to that experiment. Copy over the relevant data from the first sheet to the second so that the second sheet has your experiment organized according to your experimental design. At the end, you want to be able to easily calculate averages from your control and your treatment.
   

8. At the end of your experiment, calculate the averages for the control and intervention so that you can assess whether the effect is meaningful to you. For example, if something added 20 minutes of sleep that is probably a good gain, especially if you feel more well rested in the morning or more energetic during the day, and that is more important than the statistical significance. Ultimately, however, that is something you have to judge based on your own values and desires and what costs might be incurred by carrying out the intervention over the long-term.
   

9. Use the Graphpad t-test calculator to test for statistical significance. Choose “enter or paste up to 2000 rows.” You can then copy and paste from your spreadsheet the control values in the group 1 column and the intervention values in the group 2 column.
   
   You should now reflect on how “paired” your pairs were. On the one hand, you randomized the pairs so you are completely justified in using a paired t-test. On the other hand, you probably only have genuine biologically significant pairing if you have real trends over time that were impacting your primary endpoint. For example, if your menstrual cycle impacts the trials and they were all done in one month, each pair will be tightly tied to a specific part of your menstrual cycle and pairing makes a lot of sense. If you did your trial in March and it was the first month of the year you spent outside, you might have cumulatively increasing vitamin D status and other benefits of sunlight exposure and each pair of trials will be tightly tied to a specific point of that accumulation. Pairing again makes a lot of sense. On the other hand, if no such cycles existed and you have a lot of random variation that hit you randomly across your experiment and never lasted more than one day at a time, pairing might not be best. You can try the calculator with both “unpaired t test” and “paired t test” and see what makes the most sense to you.
   
   If you use a paired test, you have to enter the data at the level you randomized. For example, if you randomize pairs of ten-day periods, you need to enter the average for each ten-day period instead of the daily data, and the 10-day periods that were paired with one another need need to occupy the same position in their respective columns of data.
   
   In general, getting below 0.05 is considered statistically significant, but this cutoff is arbitrary. Your goal is to convince yourself, not publish this in peer review, so you can be a bit liberal in your interpretation of the stats. A lower p value gives you higher confidence your results are real.
   
   The calculator gives “two-tailed” results. This assumes that your results could be different in either direction. If you have a strong hypothesis about the direction of your experiment and would be willing to give up the ability to test whether the opposite result were true, you can cut the p value in half to get “one-tailed” results, which will be much more likely to be statistically significant.
   

10. If you are convinced by the results of your trial, reflect on whether the benefit is worth whatever cost might be involved. If it is, incorporate it into your regimen, and move on to your next optimization target or strategy. Don’t randomize again until “try it, track it, and let it stabilize” gives you another set of confusing results that are worth randomizing.

Randomized Self-Experiment: An Example

In my own optimization studies, I found that 24-hour whole wheat sourdough seemed to better promote sleep than white rice, nixtamalized corn, Ezekiel bread, or any other carb source I had tried when equating for net carbs. However, I had found this intermittently without intentionally testing it.

Unfortunately, I have also determined that anything with fructans in it hurts my gastrointestinal function to at least a mild degree, and this made 24-hour whole wheat sourdough a tradeoff due to its fructan content.

Nixtamalized corn has proven to be the whole-food carb source I can best tolerate as a generic source of carbs if I am to, for example, replace fat calories with carb calories without a big impact on other aspects of my diet.

Thinking the impact of sourdough on my sleep was probably real, I considered that its pro-methylation potential due to its choline and trimethylglycine (betaine) could be responsible. The only other carb source with that type of pro-methylation profile is quinoa, due to its high trimethylglycine content. The main downside of quinoa is its substantial oxalate content.

I tried replacing corn with quinoa, and I tried comparing this to replacing fat sources with enough extra egg yolks to equate the sum of choline and betaine.

My initial results made it look like quinoa was unique in optimizing my sleep, but may have also been hurting my energy:

Sleep is in total hours per night and energy is self-rated as described earlier.

However, the longer I stayed on the quinoa the more this differential seemed to collapse:

When I looked over the data, it appeared to me that these ebbs and flows could more easily have been explained by the fact that I was increasing my calories after transitioning from a leaning-out to a weight-maintenance phase, that my workouts were being adjusted according to my current goals, thereby impacting how much recovery I was needing, and that my sleep was largely following how much I needed rather than how much I was able to get.

The results seemed sufficiently confounded and confusing that I decided to randomize.

I did not conduct a power calculation. Instead, I decided to run six pairs of trials with equal numbers of corn-quinoa trials as quinoa-corn trials.

Since I do food prep in batches that last four days, each pair of trials lasted eight days, either with four days of corn and then four days of quinoa or with four days of quinoa and then four days of corn.

I pre-specified that I would have an equal number of each order, and used a random number generator to generate the sequence. 1 meant corn first, 2 meant quinoa first. The random number generator gave me the sequence 2,1,1,2,2. Since there was only one slot left and 2 had already been used three times, the last spot automatically went to 1 by virtue of my rule. This led to 2,1,1,2,2,1, which translated to quinoa, corn, corn, quinoa, corn, quinoa, quinoa, corn, quinoa, corn, corn, quinoa.

I did not observe any trends indicating a need for a lead-in period or a washout period, so the trials straightforwardly reflected my four-day food prep cycle.

Since I am randomizing blocks of four corn day then four quinoa day versus blocks of four quinoa days then four corn days, and because the main purpose of the pairing is to avoid heavy clustering of the trials over time, the data can either be analyzed at the daily level in a unpaired t-test or in the four-day block level in a paired t-test. The idea behind pairing is that any given eight-day period had something similar within itself that was generally different in the other 8-day periods, and that might line up to something like the ebb and flow of my workout and recovery periods.

My trial is currently 83% complete, and these are the preliminary results:

The sleep is not even close to statistically significant at p=0.57 for an unpaired test and nearly identical results for a paired test.

My energy difference is not statistically significant, but it is a lot closer, reaching 0.24 or 0.25 depending on whether I use pairing. These are two-tailed t-tests, which aim to express equal skepticism toward quinoa helping or hurting my energy. Given that my initial results suggested quinoa hurt my energy and that I could attribute this to its oxalate load, I think it is fair to say I had a specific one-sided hypothesis and could use a one-sided t-test, which would be half the p value of a two-sided test and be 0.12.

This means that there is a 12% chance I would have observed results with an energy difference equal to or greater than this if quinoa has no genuine impact on my energy.

For my purposes, I conclude that quinoa does mildly hurt my energy. This wouldn’t fly in a peer reviewed paper, but my odds are 88% versus 12% so I’ll take the 88%.

I do intend to finish the last two trials to arrive at the final stats, but I am leaning toward excluding quinoa from my default food selection and keeping corn.

A note about the sample size calculations. In my preliminary data, my sleep after eating quinoa was higher on average, but it was also more volatile. Had I used my preliminary standard deviation for a sample size calculation, it would have told me to randomize 23 pairs. One downside to using the sample size calculation is that if you are purposefully running a randomized experiment, your standard deviation will probably go down. For example, if I use my current standard deviation within my experiment alongside the preliminary group averages from my observational data, it suggests I need 12 pairs to make a 7.0 versus 7.57-hour sleep difference statistically significant. This is twice what I have but half as many as the preliminary results would have suggested. Thus, if you get a wildly high number for the sample size calculation you can probably cut it in half if you practice rigorous controlling of variables during your randomized experiment.

Again, a good compromise here is to take the number recommended, and pre-specify that you will analyze your data at the half way point and stop if the results are convicning.

You don’t want to stop at a random point you didn’t pre-specify, because this will create bias by making you likely to stop whenever the results look the way you want them to.

At this point, the difference in my sleep has collapsed to 0.1 hour or six minutes. That won’t become significant by doubling the number of trials and I wouldn’t care if it did, because six minutes of sleep isn’t worth changing my diet over.

By contrast, the p value for the difference in energy might stay as low or decline in value further after completing the last leg of the experiment, and while it might not reach statistical significance, I expect the odds will continue to favor corn over quinoa for energy, which is convincing enough for me to stick to corn as a default carb source.

Nevertheless the primary conclusion from this experiment is that the initial perception that quinoa was helping my sleep was wrong, and I effectively refuted my hypothesis quinoa would increase my sleep due to its choline and betaine content.

When You Are Done Optimizing

You are done optimizing once you are getting the results you want or you have run out of strategies.

If you are now getting the results you want, revert to “if it ain’t broke don’t fix it.” You should take a rest from changing things, and should focus on maintaining your results.

If you are doubtful that everything you have accumulated into your regimen is necessary for those results, you can use this optimization protocol to slim down your regimen. Focus on stability of your previous optimization targets as you use the exact same protocol framework to remove things instead of adding them.

Once you are confident in your ability to maintain your benefits, you are now free to exit the optimization sequence, or to pick another optimization target. If you pick a new optimization target, you should still track your previously optimized targets. Make sure optimizing your new target either doesn’t interfere with your previous gains, or only does so in a way and at a magnitude that you are willing to accept as a tradeoff.

If you are not yet happy with your results but you have run out of strategies, you may want to take a rest and “deload” from optimizing. Once you feel rested and ready to go again, focus on studying and learning until you find something new to “try and see if it works” or to use with this optimization protocol.

Staying Optimized Over the Long-Term

Over the long-term, you may find that a particular strategy wears off or that your needs have changed, or that something that was helping you for a period of time is now hurting you.

If a strategy wears off you may simply need a new one. If your needs changed, you may need to optimize with a different target in mind. If something that helped you short-term is hurting you long-term, you should use this optimization protocol to test the impact of removing it the way you tested the original impact of including it.

In my Protocols, I will build into specific protocols things to look out for when you’ve been using the strategies for too long.

Some things to be suspicious of:

• Doses of supplements that are much higher than my custom nutrient targets.

• Superfoods that provide certain nutrients at levels much higher than those targets.

• Foods that provide necessary nutrients but that you do not digest well.

• Any herbal medicines or pharmaceuticals.

• Exercise regimes that have been too much for too long.

Even if you exit the optimization cycle to simply enjoy life, you should check back on a quarterly basis with your previous optimization targets, as long as you still value them, to see if you are staying optimized.

For example, you should not need to optimize your glucose every day, but if you are optimizing it you definitely should measure it every day until it is optimized. If you spent two months optimizing it and then exit the optimization cycle, you should, on a quarterly basis, take whatever number of measurements you previously found was needed to capture representative variability. If you had optimized it to 82 mg/dL, catching it at 95 mg/dL is better than letting a problem under the hood go on until you’re at 110 mg/dL.

Or, perhaps you had optimized your strength on certain lifts and moved on to other physical feats. Check back with those lifts from time to time so you have a leading indicator of having lost your previous progress.

You may in the future reflect back on your optimizations and find that some of them were well worth it, and others were fleeting obsessions that you no longer see the value in. The more you reflect, the more you can whittle down the number of things you need to check in on.

The more you do this, the more you refine your conception of what is and isn’t worth optimizing, and the greater the return you get whenever you decide that it is worth taking another trip into the cycle of optimization.

The Next Step: Identify What You Want to Optimize

Start with my top three tests to run for game-changing health results!

You may be struggling to improve a problem and making little progress.

Or, you may be giving up and taking a pharmaceutical shortcut that will later come back to bite you.

Or, you may be missing the forrest for the trees in search of optimizing niche things that hardly matter when you haven’t yet mastered the basics.

Or maybe you’re here because you simply want to cover all your bases and suspect you may be missing one or more.

Regardless of which category you fit in, we should all from time to time review our health routine and make sure we’re covering all the basics.

These are the things we should all be doing for our health as a baseline, not things we should be measuring or things we should be optimizing. If your measurements are coming in looking bad or you’re trying to optimize something that just won’t optimize, come back to this and see if you’re hitting all the basic action items.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

The Health and Wellness Checklist

Here’s your simple checklist. Make sure you read the detailed description below before deciding whether you can check off each item.

☐ Eat a nutritious, largely whole-food diet that is digested well and adequate in all the macronutrients and micronutrients. Limit alcohol to no more than 15 drinks per month.

☐ Get plenty of direct outdoor sunshine in the morning, a little bit in the afternoon, spend most of your day in natural lighting, and avoid artificial light at night.

☐ Maintain a healthy body composition.

☐ Breath through your nose and breathe fully.

☐ Live an active lifestyle with healthy movement patterns.

☐ Sleep well.

☐ Take a food-first, pharma-last approach.

☐ Get yearly checkups from your primary care doctor, dentist, and a movement specialist. Women should stay up to date with gynecologists and most people should add hearing and vision screening as they get older, or at any point if they have reason to believe their hearing and vision is suffering.

☐ Treat your eyes and ears well.

☐ Work in healthy positions.

☐ Be a digital intentionalist, and spend time in the real world.

☐ Be clean, from both industrial and microbial toxins.

☐ Nourish positive relationships with our family and friends and a positive outlook within ourselves.

☐ Take bites of life only as big as you can chew, but always be taking bites.

Eat a nutritious, largely whole-food diet that is adequate in all the macronutrients and micronutrients.

What follows is not a list of foods that everyone must eat, but rather how everyone should eat if they want to meet their nutrient needs without thinking any harder or further about it.

Your diet should contain protein diversified across meat, fish, shellfish, and other invertebrates, such as insects if you eat them, eggs, and dairy products. Most of these foods should take up a third of your plate at each meal when used as your main source of protein, but eggs should take up about half your plate when used as your main source of protein. When eating eggs, the white should always be fully cooked solid. When using liquid dairy products as your main source of protein, you need at least two glasses worth at a given meal.

If you eat animals other than shellfish, try to eat nose-to-tail. At a minimum, try to include liver and edible bones or bone broth.

If you don’t eat animals, put special emphasis on legumes (lentils, peas, and beans) that you digest well. These need to make up the bulk of your calories to provide enough protein.

After you meet your protein requirement, you can flexibly decide whether you feel better with the remainder of your diet being fueled by carbohydrate, fat, or a mix of the two.

Limit your total number of alcoholic drinks to more than 15 per month, and, if you do drink alcohol, group the drinks in a way that interferes least with your sleep and quality of life.

To ensure you are getting adequate vitamins and minerals from food, follow these principles:

• Eat an average of one to two ounces of liver per day (you can batch this once per week).

• Eat an average of one or two oysters per day (you need this every day unless you also get your protein for the day primarily as red meat or cheese).

• Use a tablespoon or two per day of unfortified nutritional yeast in anything that could use a cheesy flavor.

• Diversify your carbohydrates among whole grains, legumes (lentils, peas, and beans), starchy tubers such as potatoes, and whole fruits. If you’re going to cut one of those out, you can cut out whole grains. Grains and legumes (alongside nuts and seeds) should be soaked, sprouted, soured, or fermented.

• Eat several cups per day of vegetables, diversified across the color spectrum. Make sure at least one of those cups of vegetables is dark green.

• At least a cup of fruits and vegetables should be raw to maximally preserve vitamin C. The most popular sources of vitamin C-rich foods that could fulfill this purpose are bell peppers, strawberries, oranges, pineapple, papaya, and lemon.

• You need three servings of calcium-rich-foods per day. These can include dairy products like milk, yogurt, kefir, or cheese, but not dairy products that isolate the fat (butter, cream) or the protein (whey protein, casein). They can include edible bones, such as those in canned fish or the soft parts on the ends of chicken bones, or whole bone meal powder. They can include some greens, mainly napa cabbage, Chinese mustard greens, and bok choy.

• Include foods or supplements that aid in digestion at every meal. Examples include ginger, lacto-fermented vegetables, kombucha, raw apple cider vinegar, and Swedish bitters.

• Salt your food to taste. Assume you are not getting enough salt unless you do, especially if you are not eating bread and cheese.

If you are not hitting one or more of these, you should take a multivitamin. I have given guidance on multivitamins in Will Multivitamins Help You Live Longer?

Eat at least eighty percent of your food cooked at home. If you cook with oil, use traditional fats and oils like red palm oil, coconut oil, olive oil, or fats from grass-fed animals (butter, ghee, tallow, lard), not “seed oils.” When you eat out use restaurants that are confirmed to not use seed oils. Across the board, minimize deep fried foods and desserts as special treats.

Select your foods based on what produces stool quality consistent with 3-5 on the Bristol chart most of the time. Your bowel movements should not require excessive strain nor come with excessive urgency. You should at least be in the range of having two to twenty bowel movements per week, but should ideally have one to two per day.

Eat your food sitting down in a relaxed environment, chew it thoroughly without rushing, go for a light walk after eating or at least stay upright for 40 minutes, and avoid laying down within two to three hours of eating.

Get plenty of direct outdoor sunshine in the morning, a little bit in the afternoon, spend most of your day in natural lighting, and avoid artificial blue light at night.

Every morning, get 30 minutes of outdoor unprotected sunshine. Get it as early as possible after you wake up. Try to ditch your glasses if you wear them. Even if it is overcast, go outside. Even if it is raining, try to go outside under shelter.

In the afternoon, usually around 1 PM, the UV light is the strongest. Get at least two minutes of unprotected sunshine during this time to maximize the synthesis of vitamin D. This dose should be less than what is required to even temporarily turn your skin a little pink. However, the darker your skin is the more time you need. For example, very dark skin can require up to two hours to maximize vitamin D synthesis.

Outside of this small afternoon dose, you should protect your skin from the sun using clothing made from natural fibers for most of the day, especially from 10 AM to 5 PM. For example, wear a sun hat when in the sun.

Nevertheless, you should spend as much of your day in natural indirect sunlight as possible. Ideally this is outdoors, where overhead shade protects your skin from excess UV rays but outdoor sunlight is still what is enabling you to see everything around you.

If you must work indoors, to the extent possible, work in an environment that has abundant natural light coming in the windows so that you can see everything you need to see without using any artificial lights. Avoid tinted windows and “environmentally friendly” low-emissivity windows. “Environmentally friendly” windows hate your mitochondria and block the infrared light needed to nourish them.

Preferably after the sun sets, but at least for two hours before bed, avoid any artificial blue light. You can switch your ambient lighting over to red or amber lights and you can wear amber-colored blue-blocking glasses if you have to expose yourself to standard lights (like the refrigerator light or if you have rooms like the bathroom where you don’t switch over the ambient lighting).

Maintain a healthy body composition.

If you spend a lot of time among general populations in the west, a good rule of thumb is that you want to be in the upper half of muscle mass and in the bottom 20-40% of fat mass.

If you plug your height and weight into a BMI calculator, it should be at least 18.5. It should be below 25 if you are white, 24 if you are black, and 23 if you are east or south Asian, with some studies arguing for a very tight 18.5-19.2 optimal for south Asians.

If you look like you carry a “beer gut” when you look in the mirror, you have too much visceral abdominal fat and need to lose fat. It is healthier and less risky to carry “rolls,” and for women it is healthy to carry fat in the hips and thighs.

If you measure your body fat mass and percentage with DEXA, underwater weighing, or bod pod, the percentage for men should be 12-17% and for women should be at least 17%, ideally above 22%, and up to 30%. If you convert your height to square meters (from inches, square your height in inches and multiply by 703) and divide your total fat mass by this number, you can derive your fat mass index. It should be at least 4 but under 10. You can then plug your height, weight, and body fat percentage into a fat-free mass index calculator. Your fat-free mass should be at least 20. If you care more about longevity than aesthetics you should not push it higher than 24.

Breath through your nose and breathe fully.

By default, your breathing should occur through your nose.

There are exceptions, with the primary two being when you are unable to breathe through your nose and when your oxygen demand from exercise exceeds what you are able to get through nose breathing.

When you inhale, your default should be to fill your entire torso with breath, allowing your ribs and diaphragm to expand in all three dimensions, without letting your belly hang out and while keeping your shoulders relaxed.

By default, you should take long, full inhales and exhales that last around 6.5 seconds each.

You don’t want to be thinking about your breath all the time, but you can occasionally check in with how it’s behaving and set apart small amounts of time each day to practice proper breathing if needed.

Live an active lifestyle with healthy movement patterns.

You should have an active lifestyle that incorporates all of the essential movement patterns and challenges to your system.

You do not need to track your workouts or optimize for specific skills, but your movement regime needs to fulfill all the challenges necessary to keep your body working properly, because everything in your body follows a “use it or lose it” principle.

Thus, if you simply want to maintain a series of sports games for fun, this is fine, but review them to make sure that all of the following get challenged. Fill in any gaps you discover.

It is definitely possible to find creative ways to hit more than one of each item in a single exercise, but if you aren’t a trainer or a biomechanics expert you should seek one out if trying to design the most efficient program you can.

Every ten days, your exercise should include the following types of cardio sessions:

• At least one intense 30-second sprint using any type of exercise that leaves you huffing and puffing afterward.

• One cardio session where you sustain one pace for an hour that you probably couldn’t have held for another 10-15 minutes without slowing down or giving up.

• One cardio session where you hit a mix of different intensities (a sports game or group workout class could fit).

Every week, you should hit at least one heavy working set of resistance exercise for each of the following domains:

• Upper body vertical pull (e.g. pullup).

• Upper body vertical push (e.g. overhead press).

• Upper body horizontal pull (e.g. row).

• Upper body horizontal push (e.g. pushup).

• Lower body pull (e.g. deadlift).

• Lower body push (e.g. squat).

• Horizontal rotation (e.g. banded or cable trunk rotation).

• Diagonal rotation (e.g. regular and reverse chop and lifts).

• Side bending at the torso (e.g. side plank, dumbell side bends, or a kettlebell windmill).

• Abduction and adduction of the hip (e.g. standing banded or cable abduction and adduction).

• Internal and external rotation of the shoulder (e.g. banded or cable internal and external rotation).

• External and internal hip rotation (e.g. clamshells and reverse clamshells).

• The six directions of the neck (bending up and down, rotating to each side, and side-bending to each side).

If this seems too complicated, you might want to look for a group workout class that naturally hits many of the bases you tend to miss when you design your own workouts. For example, a pilates class will often hit almost all of the strength points listed above. A Lagree class will mix this with sustained or mixed cardio intensity for an hour.

The resistance exercise sets only “count” if you are within 80% of what you could have done before being unable to do any further work. For example, if you hit eight reps and could have gotten 10, it counts. But if you hit eight reps and you could have gotten 12, it was a warmup and doesn’t count toward the minimum of one set per week. To estimate this properly, you should have experienced at least several times in your life a rep where you hit a wall where you spent five to eight seconds pushing with all your might through the failure point in order to complete it. If you haven’t, you are probably underestimating your capacity.

Make sure that at least every 7-10 days elements of your exercise challenge your balance, your ability to quickly change directions, your ability to be quick on your feet, and your ability to respond to unexpected stimuli outside your control (such as having to catch a ball that someone else throws).

Get as much light movement in as possible (for example, comfortable walking), aiming for at least one if not several hours per day.

If you track your workouts, you should not be losing strength, endurance, speed, balance, or agility over time. If you are, it could be a sign that something else is wrong, but the most obvious place to look is that you are training too much. Try reducing your exercise volume, duration, frequency, or intensity, or choose easier exercises, and see if it helps.

If you don’t track your workouts, make sure that they are not leaving you tired all the time. If they are, it could be a sign that something else is wrong, but the most obvious place to look is that you are training too much. Try reducing your exercise volume, duration, frequency, or intensity, or choose easier exercises, and see if it helps.

You should not be experiencing joint pain and it should certainly not be getting worse over time. If you are, you may simply be moving wrong, or you may simply have put together a poor selection of exercises to create an imbalanced program. Have a trainer or your movement specialist (see below) review your movement quality and program design to look for flaws that could be causing joint pain. You also need to look at how you are distributing the load of your workouts. Make sure that the hurting joint is getting a long enough multi-day rest period before it gets worked hard again. Don’t use anabolic steroids, which make your muscles too strong for your joints. Don’t do the same exercise all the time while constantly pushing it to the highest load possible, which is joint suicide.

You shouldn’t be getting injured, whether acutely or chronically. If you are, have a trainer or your movement specialist (see below) thoroughly review and revise your practice to focus on injury prevention.

On the other hand, if your joints hurt and you aren’t exercising, you need to start exercising for joint health. If you are falling and injuring yourself and don’t exercise, you need to start exercising for balance, strength, and to train safe ways of falling.

Every once in a while, try a new sport or group activity. If you notice you are bad at something, or are losing the ability to do it, and it is important to you, find a way to add it to your training regimen.

If hitting all the bases above exceeds your total capacity for exercise, reduce the frequency of the exercises or the number of components you are hitting until you find something sustainable. First try doing them less often, such as every two weeks instead of every one week. If you find that this is not frequent enough for you to get better at anything, then cut down the number of exercises and focus on doing at least one thing frequently enough to keep getting better at it. Try to train your way into handling a well-rounded regime that does hit all the bases.

Your movement patterns should be checked at least yearly by a specialist (see below). Whatever you learn about healthy movement patterns from your specialist should be incorporated into your training program to improve the quality of your movement.

Sleep well.

Allow yourself at least a nine-hour window to sleep. Prioritize it above anything else you might do at night and be consistent about it every day. Start this sleeping window at the same time every day.

Frame this in a period of mental relaxation and zero exposure to artificial blue light. This winding down period should be two to four hours long.

If you notice patterns in how eating impacts your sleep — for example, if going to bed hungry or going to bed after eating a big meal hurts your sleep — eat in the way that maximizes the quantity and quality of your sleep, with the primary endpoint being that you wake up feeling rested and ready to tackle the new day.

Do not set an alarm clock. To the extent possible, schedule your day so that you have room to wake up a little later than usual and you will not be late for anything important. If you have to use an alarm clock as a safety measure, schedule your day so that you can set the alarm two hours after you expect to wake up and you will still have time to get ready for anything important even in the event you wake up to the alarm.

Take a food-first, pharma-last approach.

You should always try to meet your nutrient requirements from food before you try to meet them with supplements.

When you use supplements, you should use the least number of supplements at the lowest doses needed to achieve your purpose.

If you can solve a problem with foods and supplements you should always do so before you try to solve it with pharmaceuticals.

None of this is meant to say that there is no place for supplements — even at megadose levels — or drugs. However, you should maintain a bias to solve problems with foods and natural remedies first, with supplements second, and with pharmaceuticals last, and should be very conservative about the need to progress from one type of solution to the next.

Get yearly checkups from your primary care doctor, dentist, and a movement specialist.

The doctor and dentist are largely self-explanatory: you want to know if you are developing a problem and address it in a timely manner.

Inflammation in your oral cavity can be a systemic source of inflammation, so the dentist is not merely an aesthetic appointment.

Ideally your doctor follows your food-first, pharma-last approach.

Ideally your dentist is familiar with Weston Price and Sir Edward Mellanby’s research showing that good nutrition can lead to refilling of cavities without surgery, and errs on the side of putting the least number of foreign objects in your mouth possible using the least toxic materials available, and does not believe root canals are safe. Often searching for a “biological dentist” can help turn that up.

You need someone to assess your joint alignments and movement patterns to make sure your activity regime reinforces good movement patterns that will lead to a pain-free life from now into your old age. Try seeking someone associated with the Postural Restoration Institute, the Institute of Physical Art, or the Ida Rolf Institute.

Treat your eyes and ears well.

Use maximal ear protection when you are exposed to unnaturally loud, repeatable noises. For example, at the gun range or concerts.

Avoid putting sound inside your ears whenever you can. For example, listen to a podcast with external speakers and put phone calls on speaker when you can get away with it socially.

If you read regularly, take frequent breaks to look at things that are far away. You could set a timer every 20 minutes as a reminder to spend one minute looking at things in the distance, or you could use a task-based reminder to avoid interruptions, such as finishing a section or chapter of what you are reading.

If you do regular reading on your phone or computer, you need to address not only distance, but also screen exposure. Aim to do as much of your reading on physical paper in natural lighting as possible. One way to negotiate this would be to read online when you know you have to frequently flip between many sources, but to print something out and read it when you know you are going to read a document from beginning to end.

Work in healthy positions.

If you work on a computer, you need to minimize the crowding together of your head, your hands, and everything in between in space. Have your primary setup keep your monitor at eye level and your hands resting on an external keyboard at slightly below elbow level when your shoulders are in a relaxed and retracted position. Have the distances set while you maintain a neutral spine with your ribs stacked over your hips and your head centered over your torso, not sliding forward or looking downward. Your keyboard should be designed to allow you to separate your hands and rotate them to a less flat position, like the Kinesis 2.

If you are able to maintain this setup, it will help minimize the stress of a disorganized posture on your upper body, but your upper body still won’t get enough movement in this position, so you will need to take frequent breaks to get your upper body moving.

If you are staring at a screen for hours, you will also need to give breaks to your eyes to look at things that are far away. You may therefore want to take breaks every twenty minutes or after certain types of tasks are complete to get in “movement snacks” that include head and torso rotation as well as external rotation of your shoulders and elbows and supination of your wrists, while also getting some “eye snacks” in by looking at objects in the distance.

Even while doing computer work, however, you can find many healthier positions with your lower body to alternate. For example, if you have an adjustable standing desk with a treadmill, you can alternate between standing, walking, sitting, lunging, squatting, kneeling, sitting on your shins, sitting cross-legged or in a half lotus or lotus, and sitting in a 90-90 while you work.

If you need to use a laptop without an external keyboard because, for example, you frequently work at coffee shops or while traveling, you should take more frequent breaks for full-body exercises to break up the monotony of a poor working position.

If you do a different type of physical work where these ideas don’t apply, get input from a movement specialist on how you can best position yourself for your work.

If you do stationary work, you will need to find ways to work in movement snacks. If you do physical labor, you will need to find ways to mitigate overuse injuries from doing one type of movement over and over, or from using one side of your body so much more than the other.

Be a digital intentionalist, and spend time in the real world.

A digital minimalist tries to minimize their connection to the digital world.

A digital intentionalist thinks, “is this connection to the digital world useful to me?” and adopts the digital custom only when the answer is clearly yes.

Our bodies and minds are designed for the physical world, not the digital world.

If we spend lots of time on the computer or phone, it puts us into a stressed state. If we “forrest bathe,” it reduces stress.

Our social media feeds are designed to addict us with dopamine hits that facilitate doom scrolling.

No one goes out to a 3-hour group dinner and laments having “doom dinnerred.”

In general, then, we want to opt for the “real world” over the digital world at least as a preference, opting for the digital world only when we intentionally choose it because we see a specific benefit to it.

While we have a long way to go in understanding the impacts of electromagnetic frequency (EMF) on health, the principle that is emerging from the research is that we want a high exposure to the native EMF generated by people, animals, the earth, and by the sun as experienced from earth when following the light exposure guidelines described earlier. On the other hand, we want a low exposure to non-native EMF from electrical devices.

It may be that grounding is an important aspect of being connected to the real world, alongside attending social events with real people and spending time in nature. This means being barefoot or wearing shoes made from leather you sweat in or made from modern “grounding” materials while being connected either to the earth itself or to a conductive material like the salt water of the ocean or unsealed concrete.

At a minimum, for EMF:

• Put at least 8-12 inches of distance between yourself and your phone or your laptop, unless you have an EMF-blocking pad in the way. Even then, maintain whatever distance you can, such as having your laptop on an EMF-blocking pad on top of a pillow instead of on your lap if you are working while reclining on a couch.

• Turn your wifi off when you sleep.

• Turn off and unplug devices you are close to when you aren’t using them.

• Whenever practical, charge your devices in between using them rather than while using them.

At a minimum, for the real world:

• Spend some time in nature at least once a week, even if it’s just a walk on the beach or through the woods.

• Eat dinner with family or friends at least once a week, or do something else with genuine social value.

Other good ideas in this area are to try grounding, have your home evaluated for dirty electricity, choose areas to live that are less dense with power lines and 5g antennas, practice “phone fasts,” and anything else you think of to anchor you more in the real world than the digital world.

An important part of connecting to the real world is being barefoot. Being barefoot on the earth can be a way to practice grounding, but another important part of going barefoot is that you get far more neurological stimulation to your feet when they have to adjust to hard surfaces without shoes in the way. This trains your proprioception, which is your sense of how you are positioned within three-dimensional space. Robust proprioception is extremely important to injury prevention. Cleanliness concerns about this are covered in the section below.

Another important part of connecting to the real world is your exteroception. Whereas proprioception is your sense of yourself and your position with respect to your environment, exteroception is your perception of the environment itself through your five senses. The recommendation in the exercise section to make sure your exercise regimen trains “your ability to respond to unexpected stimuli outside your control (such as having to catch a ball that someone else throws)” will help train your exteroception. Using a sport that requires traversing an area while navigating around obstacles or making decisions of where to go based on things you observe — for example, hiking, skiing, snowboarding, an obstacle course, and many others — will help train your exteroception. Ditching your GPS and going for a ride can train it. If you are always “in your head” you should consider training it with something like my Walking Sensory Meditation.

Be clean, from both industrial and microbial toxins.

Being clean by washing your hands, taking showers, keeping your house in order, throwing old food out of your fridge, cleaning the remains of food waste from your eating environment, and cleaning the kitchen and taking out the trash is important to prevent overgrowth of pathogenic microbes on your skin, mucous membranes, and in your home. The reason pathogenic microbes are pathogenic is because they produce toxins.

We therefore want to consider water—>soap—>antimicrobial cleaners similarly to foods—>supplements/herbs—>drugs.

You may be surprised what can be cleaned with water and a cloth alone.

A soap with no fragrances or other additional ingredients is not toxic at all, though it does have the downside of dissolving the lipids of your skin. It is obvious you need soap to clean dishes because the soap dissolves the greasy elements in water just like it dissolves your skin lipids into the water. Soap is also more effective at washing pathogenic microbes off your skin, so you should use it after your hands get dirty or after you go to the bathroom and before you eat. In fact, you almost never need to kill pathogenic microbes on your hands (with obvious exceptions like surgery prep or after coming into contact with something egregiously pathogenic) because they haven’t colonized your skin and soap is incredibly effective at washing them right off into the drain.

However, it still makes sense to minimize unnecessary contact with soap. For example, if you are washing dishes by hand for a whole house all day, wearing dish gloves will be a great way to protect your hands from too much soap (as well as too much water).

Rarely do you actually need antimicrobial killers. When something kills a microbe it will inevitably have at least a little toxicity to us, so you should be judicious about using them, because the entire point of being clean from pathogenic microbes is to be clean from their toxins.

Stagnant environments breed microbial overgrowth. You want your digestive motility to be fast enough to prevent overgrowth of bacteria and fungi in your small intestine. You want light to enter your home and any water to be dried rather than to sit stagnant so you don’t get overgrowth of toxic mold. If your room is a mess, there is probably something funky growing underneath something you haven’t picked up over the last year that comes bearing gifts of toxins.

By the same token we don’t want industrial toxins bombarding us from all the entry points into our living space, so filter your drinking water, your shower water, and your air.

Synthetic clothing does contain toxins, but how often they leach into your skin probably depends on how much you sweat in them. On the other hand, working up a sweat in a material that has no sweat-wicking properties can get fungal toxins being produced in your skin quite quickly. So make choices to minimize unnecessary synthetic fibers, but be aware that industrial toxins and fungal toxins are equally problematic. Make the choice that minimizes total toxin exposure. If you wind up with an itchy rash after your workout, you haven’t done that.

You have to consider that a microbial infection leads to self-reproduction of the toxin factory within your skin, whereas synthetic fibers can only leach toxins from the clothing to the surface of your skin acutely and you may be able to wash much of them off the surface before absorption. Therefore, it is more important to wear whatever material will prevent you from getting an ugly rash after you workout than it is to avoid synthetic fibers, but once you have this part under control, you should still err on the side of natural fibers.

Try to use glass, stainless steel, and ceramic over plastic for your cooking and the storage of your food and drink. Cast iron is good for cooking if you don’t have an iron overload problem.

In the previous section, we discussed the benefit of going barefoot for grounding and proprioception. If you do this for grounding, you do want to consider the quality of the ground you are on. You can avoid the risk of soil parasites by avoiding land with fecal matter from wild animals, pets, and application of unheated manure. You should definitely wash your feet before entering your home. A convenient way to do this is to carry water-based disposable wipes to clean your feet as you switch back into your shoes. However, if you suspect you have been exposed to fecal matter you should wash thoroughly with soap and water as fast as possible.

While you cannot ground on the floors of your home, you can certainly develop proprioception by walking barefoot inside your home. If you do this, you certainly want your floors to be clean, which means you should follow the Asian practice of no shoes in the house, and dedicated slippers for special dirty areas. For example, special slippers for the bathroom, and, if you have one, the balcony, while having dedicated areas for bare feet. In the Asian practice this might be limited to tatami flooring, but if your goal is to develop better proprioception, this could be all the flat floors of your house. Just wash them off with water before you get into bed or otherwise put your feet places that will rub against your face, since no matter how clean your floor is you would never eat off it.

The best way to negotiate these concerns is to use grounding shoes for grounding, use Asian traditions to keep your home floors pristine, walk barefoot on your pristine home floors, and wipe your feet with water wipes before bed.

Nourish positive relationships with your family and friends and maintain a positive outlook.

Humans are social animals. Every introvert needs to socialize and every extrovert needs to introspect. Make sure you make time for both. Seek friends who bring out the best in you.

Health requires a disposition toward hope, optimism, and gratitude. Experiment with whatever helps you maintain this disposition and keep what works. Prayer, meditation, and gratitude journaling are some things you could try.

Have a strong sense of purpose, but do something trivial.

We can see the importance of purpose and calling to longevity in the famous story of John Adams and Thomas Jefferson both dying on the 50th anniversary of the Declaration of Independence on July 4, 1826. The last words of Adams were “Thomas Jefferson survives,” though Jefferson had in fact died several hours earlier.

I also saw this when my grandmother died. About a week earlier, she told me she was depressed because she didn’t understand why she was still on earth when my grandfather had left her several years earlier, but that she felt the family needed her. Her sense of purpose was conflicted so she was left in a state of limbo. While she got sick and fell in the days that followed, she passed from this life after internally reconciling herself that her purpose on this earth had been completed and it was time to meet her husband and Lord in heaven.

If a strong sense of purpose can determine whether we defeat or submit to death, it certainly must be a central driver of health.

At the same time, it is important to be able to do something trivial. Something strictly for enjoyment.

Many of us are so obsessed with the purpose of everything that we cannot do something that we have no answer for when asked “what is the purpose of this?”

For example, having a hobby is great, but if you are constantly trying to optimize it and improve your performance, you aren’t genuinely resting in the pure enjoyment of it. Find something that you do for fun and only for fun. You don’t even need to call it “fun.” Just something whose purpose is nothing other than the experience. Even if that is simply resting.

Take bites of life only as big as you can chew, but always be taking bites.

Our comfort zone is where we go to shrivel and die.

Outside the outer atmosphere of our comfort zone is where we go to give up.

Just outside our comfort zone is where we go to grow.

What we want to do is exit our comfort zone just far enough to obtain a growth stimulus, then come back to our comfort zone to rest and recover. Then rinse and repeat.

We can apply this to exercise, cold plunging, sauna, intermittent hypobaric hypoxia training, learning a language, mastering a new discipline, learning a new hobby, or any other type of growth.

To some extent, these will occupy specific stress buckets, and to some extent we have a total stress bucket.

If you aren’t growing, you probably aren’t taking off a big enough bite.

If you keep getting worse at something the more you bite, your bites are too big.

Always make sure you are challenged and always make sure you are properly resting and recovering from the challenge. This balance is what sets us on the path to always learn and grow.

Checked All the Boxes?

If you checked all the boxes, great! You’re doing all the basics right.

If you feel like you’re on top of the world with your health, rest on your laurels a bit.

But if there are still problems to solve or goals to reach, consider entering an Optimization Cycle.

Learn more about that in the next article:

• That Prozac is a performance-enhancing drug but inferior to exercise as an antidepressant

• That serotonin is a whole-body signaling compound acting primarily outside the brain on cells that are not neurons to control the hypoxia response, which is fundamental to the regulation of energy metabolism

• That SSRIs are primarily mitochondrial drugs rather than “psychiatric” drugs

• That protracted SSRI withdrawal is a form of mitochondrial dysfunction

• That SSRIs can act as mitochondrial poisons, despite their paradoxical ability to also stimulate positive mitochondrial functions

• And that hypobaric hypoxia training is 25 times more effective at promoting psychological resilience than pharmacological antidepressants.

The latter finding is strongly suggestive that depression is a variation of “hypoxia intolerance.”

Many people may not find this easy to grasp because “hypoxia” sounds like you must be getting strangled or drowning under water.

But the hypoxia response is always active to some degree in various places in the body as a central means of regulating energy metabolism in a completely healthy human. This is because you always have variation in how much oxygen is reaching various places in your body and what their demands for mitochondrial energy production are. Any relative mismatch between the supply and demand for oxygen will elicit some stimulation of the hypoxia response to restore equilibrium.

While some of the effects of SSRI withdrawal are a return of the depression or anxiety that the SSRI was used to treat in the first place, and while some are a result of readjustments in neurotransmitters taking place, the protracted withdrawal that lasts months or years rather than a few weeks is clearly a form of mitochondrial dysfunction that is lasting so long due to the rising of vicious cycles that are locking in the dysfunctional state.

In this tenth and final article in the SSRI series, we brainstorm some things that could be tried to remedy this mitochondrial dysfunction.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

What the Limited Literature Says

The first and most robust defense against SSRI-associated mitochondrial dysfunction is to avoid going on SSRIs in the first place.

The second and next most powerful defense is to stay on SSRIs for as short a time as possible, and to limit the course to less than a year and, if possible, to a few months or even a few weeks. The likelihood of severe and protracted withdrawal increases dramatically with the time spent on the SSRI, and the difference between one year and eight years is likely enormous.

The third defense is to initiate a withdrawal that is as long and drawn-out as possible under the supervision of a very experienced physician.

There is an exception to this: if you’ve only been on an SSRI for less than eight weeks, you may be best off getting rid of it as fast as possible. This is because somewhere between one and eight weeks is the minimum length of time necessary to make withdrawal a problem. You don’t want a slow taper to turn you from someone who didn’t need to worry about tapering into someone who has now been on SSRIs for months and needs to worry about each dose-lowering event.

There is some support for using buproprion (Wellbutrin) and PDE5 inhibitors (e.g. Cialis or Viagra) for SSRI-related sexual dysfunction, which may generalize to post-SSRI sexual dysfunction (PSSD). Titrating up Terra Soul Gelatinized Maca from ¼ teaspoon to 3 tablespoons per day could boost dopamine and norepinephrine and act as a more natural form of buproprion. Titrating up citrulline from 2 grams to 10 grams per day could boost nitric oxide function and act as a more natural substitute for PDE5 inhibitors. There is anecdotal support for androgen therapy, and my 5 Ways to Boost Testosterone Naturally could come in handy for more natural alternatives.

A systematic review published earlier this year found that going back on the original SSRI helped in 47% of cases of protracted SSRI withdrawal, while going on other drugs, such as benzodiazepines, pregabalin or propranolol only helped in 18% of cases. Three to six months of cognitive behavioral therapy helped in three out of three cases reported. The GABA-stimulating effects of benzodiazepines and pregabalin might be replicated in natural and more gentle ways with 100-800 milligrams per day of oral GABA, or chamomile.

New Ideas Based on Optimizing Mitochondrial Function

This series has suggested throughout that mainstream progress on this topic is hamstrung by the commitment to the “psychiatric” theory of SSRIs and that real progress can only be made by looking at these drugs through a “mitochondrial” lens.

Here are some ideas to try based on peering at the topic through this lens.

Near Infrared Light

One case report showed that low-powered laser in the near infrared range on the scrotum and spine of a PSSD patient with no feeling in his genitals led to a 40% increase in penile sensitivity.

Infrared light acts in part by feeding photons to complex IV (cytochrome oxidase), which improves its ability to facilitate ATP production. Further, near infrared also structures mitochondrial water to improve its viscosity in a way that facilitates greater production of ATP by ATP synthase.

You can apply therapeutic near infrared light from five to thirty minutes per day (700-1000 nanometers) on each part of the body. For example, Saunaspace (expensive, whole body), RedRush Pulse (less expensive, sit by it a lot, turn to face different sides), or LUMEBOX Red Light (even less expensive but small and you have to apply it to specific areas).

The best source of near infrared light is the sun, but you have to be very careful to manage not getting burned, and these devices allow a greater dose of near infrared without exposure to burning rays.

Recent research suggests that infrared applied to one area in your body can benefit another. For example, when the light is applied to the chest, it improves vision. However, it makes the most sense focus on applying the therapeutic light to the most relevant part of your body, as long as you don’t shine it directly in your eyes. For example, try applying it to your head to see if it helps with brain zaps.

Creatine

Creatine is your mitochondria’s power grid.

Creatine travels through the cell orders of magnitude faster than ATP, so it is the primary way you spread the energy produced by the mitochondria throughout the rest of the cell.

As I covered in Your Cells Are Starving For Creatine, everyone needs to optimize their creatine status. If you aren’t consuming one to two pounds of animal flesh per day, or if you aren’t consuming 500 grams per day each of quinoa and tofu, you almost certainly need to supplement with creatine.

The success story I shared earlier in the series, who reported at 8.5 months that “I've still been pretty amazed how effective your protocol has been. On the whole I feel like a whole different person from a year ago. It's wild” was using creatine as a central part of his protocol.

He had to start at 100 milligrams of creatine for a long time due to the twitches it caused him. He managed these with my strategies in Handling Creatine Side Effects.

You can start with a five-day loading dose of 20 grams per day followed by a 3-gram-per day maintenance dose, but if you have adverse effects you should start at a dose that doesn’t cause them, even if that means reverting to something as low as 100 milligrams. The loading dose helps you achieve muscle saturation quickly. You can also achieve muscle saturation by using 3 grams a day for a month. Creatine should be spread out into three or four doses per day taken before meals (not after meals), after exercise, or on an empty stomach to facilitate maximal absorption. After achieving muscle saturation, you can titrate up at one gram per day per week as high as 20 grams per day, but should stop at the point of maximal benefit, which might be around 3-5 grams, could be 7-8 grams, and for some people might be 20 grams.

The Mitome test had showed that this client had impaired transport of methyl groups into the mitochondria, and since creatine supports methylation, this may have made him respond especially well to the creatine. However, everyone needs to optimize their creatine status so this is a safe and fruitful path to embark on without testing first.

Melatonin

The animal evidence covered in Melatonin Is Your Mitochondria’s Guardian Angel suggests that secretion of melatonin at night by the pineal gland is a major source of mitochondrial melatonin throughout the body as cells clear it from the blood by taking it up and transporting much of it into their mitochondria. Less predictable stimulation of the hypoxia response during periodic hypoxic stress supplies the remainder of the melatonin by driving circulating serotonin into mitochondria where it is converted to melatonin.

SSRIs likely deplete mitochondrial melatonin by depleting serotonin at a whole-body level and in most specific tissues outside the brain and gut, and by blocking the entry of serotonin into the mitochondria for melatonin synthesis.

The depletion of mitochondrial melatonin could lead to a vicious cycle where mitochondrial function is damaged from constant hypoxic signaling that respiration is unnecessary and a constant state of melatonin deficiency, thus hurting the ability of mitochondria to synthesize melatonin from serotonin, leaving them unable to lift themselves up by their bootstraps once the SSRI is taken away. Supplemental melatonin may therefore help break the vicious cycle.

Human trials of melatonin supplementation have used 5-10 milligrams per day for up to three months, 10-50 milligrams per day for up to one month, and 100-400 milligrams per day for up to two weeks, and 1000 milligrams on one single day. Case reports have used as many as 1600 milligrams per day for 24 days in individual patients. High-dose melatonin increases the risk of drowsiness, headache, and dizziness by 40%, but it does not increase the risk of severe adverse effects.

Overall, high-dose melatonin appears safe for a limited period of time, with the safety of using 10-50 milligrams for up to a month demonstrated rather robustly.

Most studies use melatonin at night, though some have used it in the day. I think it’s clear that melatonin should be used before bed so as not to disrupt the cyclical rise of plasma melatonin at night and its fall during the day. If you supplement before bed, the plasma level will rise more than usual, but this will also be rapidly cleared by cells that take it up and can deliver it to their mitochondria.

The use of melatonin in this manner should be thought of as a short-term repletion of the mitochondrial melatonin pool. This can be done with 10-50 milligrams of melatonin taken before bed for up to one month. You should be able to stop this cold turkey with no problems, but if you notice any decline in sleep quality during withdrawal you could try tapering it slowly.

Strategies to Improve Mitochondrial Turnover

SSRIs can dysregulate mitochondrial turnover by promoting mitochondrial biogenesis (the creation of new mitochondria) through specific serotonin receptors and by promoting mitophagy (breakdown of damaged mitochondria) through direct activation of the sigma-1 receptor.

SSRIs with low sigma-1 activation are likely to overstimulate mitochondrial biogenesis relative to mitophagy, while those with high sigma-1 activation are likely to overstimulate mitophagy relative to mitochondrial biogenesis. Withdrawal from these SSRIs could do the opposite.

So, withdrawal from fluvoxamine and sertraline, which are strong sigma-1 activators, would be expected to cause a loss of sufficient mitophagy, while withdrawal from paroxetine, citalopram, and escitalopram would be expected to cause a loss of mitochondrial biogenesis. Fluoxetine would be in the middle.

However, this also will depend on an individual’s genetic variation in all of the relevant receptors.

Mitophagy and mitochondrial biogenesis are regulated by the supply and demand for mitochondrial energy metabolism. In general you need both demand for new mitochondria and supply of the substances needed to produce mitochondria.

Exercise is an excellent way to stimulate the demand side of mitochondrial biogenesis, where induction of an energetic deficit by the exercise can increase mitophagy and subsequent resting and refeeding can increase mitochondrial biogenesis.

Cognitive behavioral therapy could be seen as a form of exercise since it will create demand for mitochondrial energy production in all of the neural networks that support the cognitive and behavioral changes being stimulated in the same way bodily exercise does this for neural networks that promote strength and for energy production within muscle itself.

Your exercise program should be designed relative to your capacity. If you are chronically fatigued and feel incapable of serious exercise, take what you can do, cut it down to 60-70%, and begin an exercise cycle where you aim to slowly do more over time so that by the end of twelve weeks you can do more than you had been able to at the start. If you see no change to your capacity, you may need to work harder. If you see yourself getting weaker and more fatigued as you do more work, you need to reduce your volume, frequency, duration, or intensity, or select an easier exercise, and rest more. Find the right pace of rest and recovery and take off bites you can chew.

My fasting-feeding reset is designed around a cyclical promotion of mitophagy in the fasting state and mitochondrial biogenesis in the fed state. It uses exercise and supplements to amplify each state on a cyclical basis. This is a well balanced approach. However, given the possibility that SSRI withdrawal has left you with a pronounced imbalance, it is important to pay close attention to your body’s responses and avoid excessive stress. If you find your resilience declining rather than improving, deemphasize the portion of the cycle that seems to be causing the most bodily stress. For example, you may want to greatly reduce the stimulation of the fasting part of the cycle while indulging more in the refeed or vice versa.

While various supplements can be leveraged to globally increase mitophagy or mitochondrial biogenesis, this risks making a major imbalance worse.

I recommend testing Mitome before doing this, which will tell you whether you are more imbalanced in one direction or the other and will put the relevant supplements directly into your protocol.

Another approach is intermittent hypobaric hypoxia training. As I covered in This Is 25x More Effective Than Antidepressants—And It's Not a Drug, animal studies indicate this is a powerful promoter of psychological resilience and it may be addressing the root cause of depression far more directly than SSRIs do.

While I strongly promote everyone engaging in some degree of hypoxic stress as part of their health regimen, this is another case where Mitome testing would provide value. Too much hypoxia can hurt the respiratory chain, and idiosyncratic differences in the hypoxia response can underly the different responses to different SSRIs. Mitome can see whether your hypoxia regimen is excessive or just right and the protocol will help you adjust it.

5-HTP for Certain SSRIs

SSRIs with poor sigma-1 activation like paroxetine, citalopram, and escitalopram deplete serotonin even in the brain, and recovering from these might warrant supplementation with 200 to 900 milligrams per day of 5-HTP, with or without Seeking Health B6 as P5P 25 milligrams once a week to once a day.

Fluoxetine is in a gray area, whereas sertraline and fluvoxamine not likely to do this.

Serotonin synthesis is likely to be restored with adequate protein (at least 1.2 grams per day per kilogram of ideal bodyweight), adequate micronutrients, and the proper assortment of stressors covered above like exercise, cognitive behavioral therapy, and intermittent hypoxia.

However, the ability to support 5-HTP synthesis is complicated by a large number of nutritional cofactors, and while it does not occur in the mitochondria, it does depend on mitochondrial energy metabolism.

The ability to convert 5-HTP into serotonin is far more simple and only requires vitamin B6.

Serotonin depletion is less likely to lock in a vicious cycle than mitochondrial melatonin depletion, which is why this is considered a possible adjunct for certain SSRIs rather than a general principle applied to withdrawal from any of them.

Runner Up: Natural Sigma-1 Activation

A few more things deserve some mention but their roles aren’t quite clear.

Chronic activation of the sigma-1 receptor by fluvoxamine or sertraline could leave it chronically underactivated during withdrawal. This receptor is never supposed to be chronically activated, so what SSRIs are doing to it is completely unnatural.

We know that it should be activated periodically during stress, so this is another case where healthy stressors like exercise and intermittent hypoxia become important. We don’t know what naturally activates it, but DHEA-sulfate, choline, myristic acid, and sphingolipids have all been suggested. Choline can be obtained from liver and egg yolks; myristic acid from coconut oil and butter; and sphingolipids from the fatty portions of a wide variety of foods, with unique compositions found in organ meats, fish, shellfish, mushrooms, whole grains, and fermented foods.

Many Relevant Nutrients

As with any other health problem, you need to make sure you are hitting all of your nutrient targets. Track your diet in Cronometer to make sure you are hitting them on paper, and run the Comprehensive Nutritional Screening using the Cheat Sheet for interpretation to make sure your regimen is working to support your body’s unique needs.

Everyone Is Different

Everyone has variation in the specific strengths and weaknesses of their mitochondria, and assaults on the mitochondria that precipitate the onset of chronic mitochondrial dysfunction will reveal these deficits. Testing the mitochondria in the dysfunctional state can provide insight into what the limiting bottlenecks are, and can therefore add a highly individualized layer to your protocol that makes it specific to correcting your own dysfunction. Mitome tests your mitochondrial function with a cheek swab and delivers a personalized protocol that you could integrate with the SSRI-specific strategies outlined above.

Share Your Experience in the Comments!

If you have SSRI withdrawal mitochondrial dysfunction and have tried some of these strategies, let us know what worked and didn’t work in the comments.

Tapering Off and Resolving Depression Without Drugs

If you are on SSRIs now and need a taper protocol, go here:

If you want to beat depression without the use of drugs, go here:

Moreover, we will see that the impact of all three can be experimentally blocked in animals by blocking their impact on energy metabolism, which shows that the “psychiatric” model of how these drugs work to alleviate depression is wrong, and the “mitochondrial” or “energetic” model is correct.

This shows us that even when Pharma succeeds, it is flying blind and getting lucky.

The most important point, however, is that this nearly 25-fold (24-fold to be precise) better method is not a drug at all. It operates entirely outside the pharmaceutical model and is instead somewhat more akin to exercise, though the only fully natural way that our great-great-grandparents could have achieved it would have been to periodically travel into the mountains.

The core comparative study we will look at was done in mice, but we will put it in the context of the human research.

While randomized controlled trials are the best way to determine whether an intervention causes a benefit in humans, animal studies are extremely valuable in their greater ability to infer mechanisms. “Mechanism” is a fancy way to say “why and how something works.” When you understand why and how something works, you can make much better decisions about how and when to use it as a tool. RCTs should be used to assemble what we know can work into our broader toolkit, while understanding how and why these tools work should be used to make wise decisions about what tool to use for the job.

The intervention we will look at is called hypobaric hypoxia training.

This is the opposite of hyperbaric oxygen therapy.

It is used to induce tolerance to low blood oxygen levels caused by low atmospheric pressure in, for example, military pilots.

We saw earlier in this series that serotonin’s fundamental primary action is to act throughout the entire body to facilitate hypoxia tolerance, so it should not be surprising if we find that hypoxia is fundamental to depression.

Hypobaric hypoxia training can have adverse effects, including drowsiness that can onset hours after the training and could interfere with driving or operating machinery. If done without proper safety precautions, it carries all the same risks as altitude sickness. This often include transient headache, nausea, shortness of breath, an elevated stress response, getting sick, decreased cognitive performance, and lethargy while they more rarely can include delirium or even psychosis. Generally these respond to restoration of the oxygen level, so it would take some catastrophic misuse to generate the worst of these problems using a simulated altitude machine, whereas getting stuck in the mountains at a level of altitude you cannot tolerate can be harder to remedy quickly. Used properly, properly dosed hypoxia/altitude stress may be a safe and very powerful weapon to combat depression.

This article is not meant to suggest this as a first resort for depression — the basics of nutrient adequacy, a healthy lifestyle, and healthy mental practices should be the first resort — but rather to examine the fascinating prospects for this therapy and use it to properly understand the impact of SSRIs.

Nevertheless, I do believe that optimal health can best be produced with defined doses of hypoxic stress from periodic excursions to mountains or use of simulated altitude at sea level.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

Altitude, Hypoxia, and Depression

Ira Katz of the Albert Einstein College of Medicine in the Bronx, NY proposed in 1982 that hypoxia could cause depression by interfering with neurotransmitter metabolism. His evidence was largely limited to experimental studies showing the dependence of neurotransmitter regulation on oxygen and indications that depression might result in some medical conditions that cause hypoxia.

In 2018, authors from the University of Utah and the Salt Lake City Veterans Affairs Medical Center collected the evidence tying altitude to depression in the Harvard Review of Psychiatry. In the United States, rates of depression and suicidality are highest in the mountain states, and correlations adjusted for age, race, and sex suggest that a state’s peak altitude explains 38% of its suicide rate while the altitude of its capital city explains 55% of its suicide rate. On a county basis, further controlling for income and population density, altitude explains 25% of the suicide rate. The effect size is large: high-altitude areas have a 3-fold higher rate of suicide than low-altitude areas. Similar results have been published for Spain, Saudi Arabia, South Korea, and in Peruvian electrical workers, though two studies failed to find associations in Turkey and India.

Six cases of new-onset panic disorder have been documented in travelers to the mountains of Nepal, anxiety has been found increased in mountaineers doing 55 days of simulated altitude training at the equivalent of 9000 meters, and depression and anger increased in US Marines doing simulated altitude training for thirty days at the equivalent of 2053 and 3600 meters.

Finally, chronic illnesses that cause hypoxia such as asthma and chronic obstructive pulmonary disease (COPD) cause more depression and suicide risk than other chronic illnesses. Both asthma and cigarette smoking are specifically associated with depression and suicide when current. Past asthma has no such association, and the association with cigarette smoking wanes during smoking cessation and returns if smoking is resumed.

With the benefit of 36 more years of research, these authors were able to tie hypoxia to brain energetics and relate it to the positive studies on creatine for depression, whereas Katz was operating in the emerging “neurotransmitter” model of the brain, which had been taking the psychiatry world by storm in the wake of the popular obsession with psychedelics in the 1960s and 70s.

Studies conflict about whether airline pilots have an increased or decreased risk of depression. Self-reported depression is higher among pilots, whereas official databases say they have less depression. My suspicion is that the databases are less accurate because pilots don’t want to get taken off the job by seeking a diagnosis. Suicidal airline pilots have killed 562 people in crashes since 1972, so pilot depression is considered a security risk and the incentive is to hide it.

The blood cells of airline pilots have massively increased expression of the serotonin receptor. Stuck in the “psychiatric” model of serotonin as expected, these authors commented that “a dysfunctional serotonergic system leads to abnormal neurotransmission and the development of psychiatric disorders. Although our study was performed on peripheral blood cells, some authors have proposed that alterations in PBMC could be a reflection of alterations in the central nervous system (CNS).”

Thus they failed to appreciate that the massive increase in the expression of the serotonin transporter of non-neuronal cells outside the brain in pilots is showing the extreme significance of serotonin to the whole-body hypoxia response, not its role in mediating “positive vibes” within the brain.

Military pilots are often required to do hypobaric hypoxia training every few years, but its impact on depression is unstudied apart from the study of US Marines mentioned above. There are 25 papers that discuss the use of hypobaric hypoxia in airline pilots but they do not cover depression.

24 Times More Powerful Than Antidepressants

The central mouse study we will consider compared the antidepressant effect of hypobaric hypoxia training to the tricyclic antidepressant imipramine, with some side experiments that looked at ketamine.

The mice were subject to 70 days of chronic mild stress, including restraint, forced cold plunging, deprivation of food and water, tricking them with an empty water bottle, tilting their cage at a 45 degree angle, hours of strobe light exposure, and keeping them in a dirty, feces-full cage.

Multiple tests were used for depression and anxiety, including their preference for sugar, the degree to which they groom their fur, their ability to stay swimming when forced into a bucket of water instead of giving up, their willingness to explore an open area, and the ease with which they became afraid of loud noises.

They were split into four groups with treatments thrown into the middle of the chronic stress period: injection with imipramine or a saline control for four weeks, or exposure to four hours a day of hypobaric hypoxia equivalent to 5000 meters (16,400 feet) or a normoxic control for two weeks.

The difference in how long the treatments lasted is stunning:

• The antidepressant effect of imipramine faded away one week after treatment.

• The antidepressant effect of hypobaric hypoxia lasted at least 12 weeks after treatment.

The actual treatment shows an at least 12-fold longer-lasting effect of hypobaric hypoxia.

The imipramine was used for twice as long, so we could say from one perspective that it was 24-fold more powerful. However, the experiment was stopped there, so it lasts at least this long, and perhaps much longer.

They performed a number of more limited side experiments that showed the following:

• Blocking the hypoxia response at a molecular level abolished the antidepressant effects of hypobaric hypoxia training, imipramine, and ketamine.

• Chemically activating the hypoxia response by injecting cobalt chloride into the brainstem produced an antidepressant effect that lasted at least seven days.

More Powerful Than Ketamine?

This study showed that activation of the hypoxia response mediates ketamine’s effects, but did not compare the robustness of the response to ketamine over time.

Nevertheless, there is a significant body of human literature on ketamine that we can compare this to.

Ketamine is a dissociative anesthetic. While it had been used for human surgery in the past it is now mostly used for veterinary surgery. It is used illegally and recreationally for its dissociative and hallucinogenic effects. Dissociation refers to detachment from emotional and physical reality.

A single infusion produces a powerful antidepressant effect, but its utility is limited by the fact that the doses required generally produce some degree of dissociation and often cause a rise in blood pressure. The impact of a single infusion lasts for one to two weeks.

Ketamine is primarily known as an NMDA receptor antagonist. Since schizophrenia involves NMDA receptor hypofunction, there is analogy between the ketamine-induced state and the schizophrenic state, which earns it a place in the category of “psychotomimetics.”

Controversy exists over whether the dissociative effects mediate the antidepressant effects. If they do, this would fit into a “psychiatric” model of its impact rather than an “energetic” model because modulating neurotransmitter systems to achieve perceptual changes and psychological experiences would be the central mechanism of the drug.

On the one hand, most people experience some degree of mild dissociation during ketamine-based antidepressant treatment. On the other hand, while this does predict the antidepressant response, correlation coefficients indicate that it only predicts 4-17% of the response. This correlation is hardly evidence of causation. The dissociation could reflect a combination of the dose and distribution of the drug as well as the the sensitivity of the individual’s receptors to it. That is, it could essentially be an index of the “individually-adjusted dose.” The dissociative effect also represents a powerful capacity for a placebo effect. Regardless, 4-17% of the explanation of the variation is simply a weak proportion of the antidepressant effect. The perceptual changes are either entirely incidental to the antidepressant effect or play a minor role in it.

Given that its impact is blocked by molecularly blocking the hypoxia response in mice, the “mitochondrial” or “energetic” model is a much better explanation of the ketamine effect, just as it is for SSRIs and tricyclic antidepressants.

That the antidepressant effect of ketamine tends to last one to two weeks suggests hypobaric hypoxia training could be 6-12 times more effective than it, using the mouse study.

But this is based on single infusions.

There are isolated case reports of multiple ketamine infusions leading to remaining depression-free for six months to a year. However, larger studies show that this is not the normal result:

• In a study of 10 people receiving six infusions over two weeks, the antidepressant effect lasted three weeks after the treatment was stopped.
  

• In 11 people receiving 10-51 infusions over six to 49 weeks, four people continued on the maintenance regimen, four stopped because it stopped working, two stopped for unknown reasons, and one stopped because their 51st infusion caused a level of dissociation and restlessness that made them not want to use it again. This indicates that you cannot maintain the effect perpetually because it will stop working before you stop using it in many cases, or you may run into a wall of adverse psychological effects.
  

• In 15 people receiving six infusions over 12 days, one dropped out after the first infusion because it caused fatigue an irritability, one dropped out after the second because it didn’t work, and one received no benefit even after the sixth infusion. Of the remaining twelve, nine achieved remission and three partially improved. In half of the twelve who got some benefit, half were still benefitted four weeks out while the other half relapsed in an average of 16 days. Nevertheless, many of them had declining neurocognitive and mental health scores between one and two weeks after treatment ended, and everyone was in decline after the 2-week mark.
  

• In 12 people receiving six infusions over two weeks and then four weekly followup infusions, five people dropped out before the initial six infusions were over. Two withdrew because of adverse effects, one stopped because it wasn’t working, and two withdrew because the ketamine was making their depression get worse. Two subjects partially improved and five achieved remission. In the four weeks that followed ketamine treatment, only one remained in remission. The remitters all remained better four weeks out than they had been pre-ketamine, but they were also losing their peak benefit within one week after the last ketamine treatment.
  

• In 100 people receiving six infusions over 12 days, 13 people dropped out during the infusion period and 10 more dropped out during the two-week followup. One dropout reported becoming suicidal and another dropout reported becoming manic. 26 achieved remission and 20 more had a positive response. The antidepressant effect lasted robustly through the two weeks of followup.
  

The most robust finding is that a ketamine infusion produces an antidepressant effect lasting one to two weeks. There is some indication that multiple infusions produce a longer-lasting effect that appears to max out in three to four weeks but be quite variable between individuals.

So we can be generous to ketamine and say that hypobaric hypoxia training lasts three to four times as long as a round of multiple ketamine infusions.

Many people are poor candidates for ketamine from the getgo. The ability to indefinitely sustain infusions to avoid this wearing off seems doubtful, and dragging them out over a year seems to often cause the ketamine to lose efficacy and sometimes become harmful.

Other Hypoxia Studies in Rats

The authors of the mouse study discussed above had conducted another earlier study in rats showing that intermittent hypoxia is similar in the magnitude of its antidepressant effect to Prozac and imipramine and that it lasts at least seven days after stopping the treatment, but did not compare its duration to that of the drugs.

A different group showed that in female rats but not male rats, altitude at 4500 feet or simulated altitude at 10,000 feet and 20,000 feet for one week caused a dose-dependent worsening on tests of depression and anxiety that were conducted at 4500 feet where the lab was situated. In another study this same group showed that female rats become anxious and anhedonic after altitude exposure.

The difference between these models is that the group showing benefits ran the tests a week after the hypoxia, whereas the group showing negative results seems to have run the tests immediately after removing the rats from their chambers.

If you view hypoxia stress similar to exercise, these findings are easy to reconcile. Right after exercising, your ability to complete the same task has gone down. Once you recover, your ability has gone up.

Intermittent Hypoxia Is an Essential Stimulus to Produce Positive Adaptations

The best way to synthesize all of this is that impaired oxidative energy production in the brain drives depression and anxiety, while SSRIs, tricyclic antidepressants, and ketamine are in variable ways helping to modify the response to this problem.

While the literature even on the antidepressant effect of ketamine is far more robust than the literature on the same effect of hypobaric hypoxia (where we only have rodent studies), and while the literature on SSRIs is yet more vast, the rodent studies on hypobaric hypoxia indicate it could be a vastly superior way to produce a long-lasting antidepressant effect.

There are hundreds of reasons that anyone at sea level could have impaired brain energy metabolism, but in people who live at altitude, chronic hypoxia is excessive and plays a major role in depression and suicide.

The increased risk of depression and suicide at moderate altitude coexists with a lower all-cause mortality risk, which shows that this degree of altitude stress is causing a tradeoff, showing the dose has some benefits but is not optimal.

While chronic altitude stress raises the risk of depression and suicide, small doses of altitude stress can improve the brain’s ability to thrive under such stress, and thereby support psychological resilience.

In my review of the literature on simulated or real altitude raising testosterone in men, I concluded that 10-12 hours of 6500 feet using a simulated altitude tent three times a week is the best-performing protocol used in humans, which makes me hope that this can be consolidated to 5-6 days per month spent hiking, skiing, snowboarding, or doing whatever you want at the top of a mountain about 6500 feet above sea level.

In male boxers and cyclists, stacking the hypoxia stress for five or six days starts to cause the acute loss of the initial testosterone benefit, while spacing the protocol out three times a week allows the benefits to be maintained indefinitely, at least as far as has been studied. However, I believe that 5-6 days in the mountains would slightly push you into “overreach” from which you would bounce back more hypoxia-fit, testosterone-loaded, and psychologically resilient, while the rest of the month at sea level would allow you to reset your capacity to handle another round.

In non-athletes and recovering drug addicts, testosterone increases linearly during a month spent hiking the mountains of Tibet. While you could propose that natural mountain exposure is healthier than simulated altitude exposure, this cannot explain why depression and suicide is higher in people naturally living at altitude. Therefore, I believe the answer is that high-level athletic activity is already adding enough into the physical stress bucket that they become very sensitive to exceeding the optimal dose, whereas people who are sedentary are chronically physically understressed and can take longer to accrue the initial benefits.

Nevertheless, athletic activity is the vocation of everyone who has a body, as Nike cofounder Bill Bowerman once pointed out, so I think we should take the studies in boxers and cyclists as a provisional target for the optimal dose of hypoxia stress. This indicates that you want the equivalent of 15-20% of your time being spent at 6500 feet. We need more research to understand how changing the exact altitude changes the proportion of time that should be spent in it. For example, maybe we could spend two or three days per month at 12,000 feet.

To replicate the protocol of the human studies, you can use an altitude tent during sleep and part of the evening for 10-12 hours three times a week and can probably substitute a simulated altitude mask.

If you live at altitude, you may already have an excessive hypoxia response. If you cannot change where you live, you can take the opposite approach. Start using an oxygen machine at a low, practical dose such as an hour a day, and titrate it up to where you start seeing improved mood and psychological resilience. Then, continue titrating it slowly by increasing the dose a little each week. When you reach the point of diminishing returns, stay there or pull back a little. Hypothetically you would want to achieve something equivalent to sea level for 80-85% of your total hours, but you may find that the minimal dose producing the maximal effect you are seeking is lower than that.

In the first article in this series, we saw that exercise is a more effective antidepressant than Prozac in mice, while Prozac is an exercise-enhancing drug that increases muscle mass and grip strength.

We then saw that a central role of serotonin is to mediate the hypoxia response.

We built on this to see that SSRIs modulate this role of serotonin in promoting mitochondrial tolerance of oxygen deprivation and even enter the cell to have their own impacts on mitochondrial function.

We now see that intermittent hypoxia has a much longer-lasting antidepressant impact than any pharmaceutical drugs.

All of this converges on seeing depression as fundamentally about mitochondrial energy production and the impact of these drugs being far more about intersecting with mitochondrial energy production than modulating psychological effects of specific neurotransmitters.

The final post in this series addresses how to deal with SSRI withdrawal-induced mitochondrial dysfunction. Read it here:

Adult-onset multiple acyl CoA dehydrogenase deficiency (MADD), also known as glutaric aciduria type II, is considered a rare genetic metabolic disorder with an incidence of one in 250,000, characterized by fat accumulation in skeletal muscle, exercise intolerance, and pain or weakness in the muscles.

Many cases may include other features, such as episodic vomiting, fatigue, difficult or rapid breathing, seizures, hypoglycemia, inflammation of the prostate, sensory neuropathy, and damage to the liver, heart, and skeletal muscle.

It turns out that sertraline (Zoloft) is 45 times more effective at inducing this otherwise genetic disorder than genetics themselves are.

In southeastern Sweden, there had been zero cases identified between 2004 and 2013.

Sertraline (Zoloft) was approved in Sweden in 2006 and in 2012 it became recommended as the first-line treatment for anxiety and depression. This increased the prescription rate by 48%. As of last year, 3.5% of people in Sweden were on sertraline.

Whereas zero cases of adult-onset MADD were diagnosed in southeastern Sweden between 2004 and 2013, nine cases were diagnosed between 2014 and 2024, but only two of them had a clear genetic basis.

In 2023, a North American group published a conference poster showing a woman who had a MADD-like biochemical profile on sertraline. The profile completely normalized with two measurements taken off sertraline. Going back on sertraline made the biochemical profile diagnostic for MADD again.

This prompted the neurology unit of the Linköping University Department of Biomedical and Clinical Sciences in southeastern Sweden to retrospectively analyze these cases for an environmental cause. All seven of the patients with no genetic explanation were on sertraline, and no other drug was taken by more than two people.

Whereas MADD is considered an extremely rare genetic disorder impacting one in 250,000, these cases imply that one in 5,500 sertraline users develops MADD. This is still rare, but it is no longer extremely rare. It implies that sertraline increases the risk of MADD 45-fold.

The publication of this paper rapidly led to similar retrospective analyses in western Sweden, North America, and Australia, bringing the total number of sertraline-induced MADD cases reported last year to thirty, with an additional two cases possibly attributable to the use of venlafaxine (Effexor) and duloxetine (Cymbalta), which are considered SNRIs rather than SSRIs because they inhibit the uptake of norepinephrine alongside that of serotonin.

That this one inquiry led to the immediate discovery of 30 cases of sertraline-induced late-onset MADD in just four of the world’s 197 countries is quite remarkable given that the total number of genetic cases ever reported worldwide is close to 600.

It is even more shocking when you see that, as we cover below, genetic MADD is cured with food-achievable doses of riboflavin, while sertraline-induced MADD seems to require going off sertraline to fix, and some people even then do not even recover but are left permanently or quasi-permanently with weakness, pain, neuropathy, and difficulty breathing.

Three case reports of genetic MADD published between 2012 and 2022 mentioned the patient was on sertraline without any further comment. Most MADD case reports do not even catalog the medications the patient is on at the time of symptom onset.

In severe cases, MADD and other disorders causing fat accumulation in muscle can lead to rhabdomyolysis, often but not always in response to exercise. This is a state of muscle breakdown where muscle proteins spill into the urine and make it darkly colored. Four case reports have been published on sertraline-induced rhabdomyolysis where it was not investigated whether it was driven by MADD. Given the other findings, these are likely severe cases of sertraline-induced MADD.

Given the general ignorance of this phenomenon and the fact that the question has not been asked in 193 out of 197 countries, these reports are surely just the tip of the iceberg.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

This is the eigth installment in our series on the truth about serotonin and SSRIs.

Share

Genetic Causes of MADD

MADD is considered a result of pathogenic mutations in the electron transferring flavoprotein dehydrogenase system, consisting of proteins encoded by the genes ETFA, ETFB, and ETFDH. This system is responsible for delivering electrons to the mitochondrial respiratory chain from the oxidation of all fatty acids; the amino acids lysine and tryptophan; the branched-chain amino acids leucine, isoleucine, and valine; and from dimethylglycine and sarcosine, which can be derived from glycine, choline, or betaine (trimethylglycine or TMG). The respiratory chain then uses these electrons to produce ATP, the primary energy currency of the cell. A backup in this system fundamentally impairs cellular energy production.

The system is dependent on riboflavin in the form of flavin adenine dinucleotide (FAD), so high-dose riboflavin supplementation is often helpful by improving the amount of FAD that can bind to, activate, stabilize the proteins involved and ensure their proper three-dimensional folding. Due to the importance of riboflavin, mutations that hurt riboflavin metabolism (in FLAD1, SLC25A32, SLC52A1, SLC52A2, and SLC52A3 genes) can also cause MADD.

While not often acknowledged as a cause of MADD, impairments in the respiratory chain from the point downstream of the ETF dehydrogenase system, from CoQ10 through complexes III and IV, could cause similar metabolic impairments.

Riboflavin deficiency, while it also has broader effects, is sufficient to cause an acquired form of MADD.

One of the major hallmarks is that an acylcarnitine profile shows many acylcarnitine species broadly elevated. These are carnitine-based detoxification products of intermediates in metabolic pathways that could not be completed. For example, derivatives of fatty acids or amino acids that are not fully oxidized will be detoxified using carnitine and will show up on the acylcarnitine profile.

The Centrality of Food-Level Riboflavin to Genetic MADD

Bizarrely, riboflavin status is often not even measured in this disorder, despite it universally being recognized to often be fixed by riboflavin supplementation. Instead, they just put people on high doses (10-150 milligrams per day in some cohorts, in others 300-400 milligrams or more) of riboflavin without testing whether riboflavin-rich foods or low-dose supplements would have worked.

Out of almost 110 cases at one center, a total of 15 people had their riboflavin tested prior to being treated with riboflavin, and they all had lower riboflavin status than healthy controls.

This raises the question of whether suboptimal riboflavin status is required to cause MADD in genetically susceptible individuals. If so, this contrasts wildly with the perception that high doses are required to fix a problem that could not be fixed with food or low-dose supplements.

Out of 48 patients with long-term followup in this cohort, the average dose used was 68 milligrams per day and the average followup was 10 months, and all 48 were “cured.”

Riboflavin generally cured them within two months. 30 of them stopped taking the riboflavin, but in 17 of them the disorder came back and went away when they resumed riboflavin. Some of them stayed on high doses and some of them transitioned to low doses or intermittent doses and none of this made a difference in the treatment outcome.

The return of the disorder after abandoning riboflavin was triggered by stressors such as physical exhaustion, catching a cold, drinking alcohol, or getting pregnant.

The patient in this case series who proved he only needed small doses of riboflavin to cure himself was one who initially cured himself in two months with 10 milligrams per day. Over the course of nine years, every time he had an attack of muscle weakness, he took five milligrams per day and it disappeared within a week.

Clearly he started from a very deficient position that needed a higher dose and duration to fix, then he progressed extremely slowly back across the threshold of deficiency and could pull himself out of it with a lower dose for a shorter period of time.

As I have covered in my abundant riboflavin materials, you can absorb 37 milligrams of riboflavin at a meal despite needs ranging from 1 to 6 milligrams per day. So these cases where people are cured with 150 milligrams are likely flushing high concentrations of riboflavin through the system to fix the deficiency quickly, then when the deficiency is fixed, the disorder disappears until riboflavin status sufficiently declines to the deficient state again.

Genetic MADD thus appears to be remarkably similar to MTHFR C677T in its riboflavin requirement, where bringing riboflavin from 1.6 milligrams per day to 3.2 milligrams a day fully resolves the function of the enzyme.

While the medical treatment of MADD seems largely incompetent in its failure to assess riboflavin status and use the minimal effective dose, the supplement industry is completely incompetent. If genetic cases of MADD only need 5 milligrams long-term, how is it even conceivable that the average person without a rare genetic disorder needs more? Yet it is nearly impossible to find a supplement of riboflavin less than 100 milligrams. Exceptions include a 36.5 milligram Thorne supplement or a 6.5-milligram liquid supplement, but even the Thorne supplement is quite high and not everyone’s digestive system will tolerate the glycerin in the liquid supplement.

In this cohort, recurrence and severity was not related to the specific gene mutations but was higher in people eating a vegetarian diet, which is generally low in riboflavin compared to an omnivorous diet.

This is because the best food for riboflavin is liver; the second-best foods are kidney, heart, and almonds; the third-best foods are red meat, cheese, eggs, salmon, mushrooms, seaweed, sesame, and wheat germ and bran; the fourth-best foods are milk and most other meats. Thus, you can design a vegetarian diet to be rich in riboflavin if you focus on almonds, mushrooms, seaweed, sesame, and whole wheat, but the average omnivore is going to get more by eating meat, eggs, and dairy food, and you have to eat organs to push your riboflavin up to the top of the range.

The one other risk factor for recurrence aside from vegetarianism was weakness in the muscles used for chewing. This is probably because these people had more difficulty eating food and therefore were more malnourished.

This is a rather shocking indication that the quantity of riboflavin in foods – most people get 1 milligram per day from food and you might hit 6 milligrams if you really design your diet around it – is more powerful a driver of MADD than genetics are.

That is, genetics load the gun and your diet pulls the trigger, with the trigger being very sensitive to the difference between 1 and 6 milligrams of riboflavin.

Unfortunately, riboflavin does not straightforwardly cure sertraline-induced MADD, where the operant variable is that the patients are poisoning their mitochondria with a pill their doctor gave them.

Sertraline-Induced MADD in Southeastern Sweden

In the cohort of sertraline-induced MADD in southeastern Sweden, nine patients presented with constant muscle weakness, usually in the legs more than the arms. Six of them had neck weakness, and in three of them their necks were so weak they physically couldn’t hold their head up.

Six had various manifestations of neuropathy, with numbness, pain, or sensory disturbances in the hands, feet, and lower legs. Three had cramps, two had episodic vomiting, and one became severely tired after fasting and was found to be suffering from acidosis.

All of them had accumulations of fat in their muscles. Two had fatty liver. One had low muscle CoQ10 content.

The authors state that none had altered respiratory chain activity, but they do not describe the details of which parts of the respiratory chain were tested nor do they show the data, so this has to be treated skeptically. For example they could have been testing something like oxygen consumption in response to malate, glutamate, pyruvate, or succinate, which would qualify as testing respiratory chain activity, but would shine no light on the nature of an impairment in the ETF dehydrogenase system.

None of them had classical signs of riboflavin deficiency. However, consistent with the nutritional neglect that broadly characterizes medicine across the board, none of them had their riboflavin tested.

Two of them had been on cholesterol-lowering statins, which made their condition worse. When they stopped the statins, they partially improved. This is straightforward since statins are well established mitochondrial toxins.

While several of them were taking folic acid, vitamin B12, vitamin D, proton pump inhibitors, thyroid meds, anti-convulsants, steroids, or ACE inhibitors, none of the vitamins were used by more than four people and none of the drugs were used by more than two people. None of them were taking drugs expected to cause myopathy as a side effect.

Two of them had a clear genetic causation to the MADD and the other seven were all on sertraline.

The onset of symptoms occurred between five months and eight years after starting sertraline. Three patients had case histories showing a dose response: one developed symptoms after increasing the dose from 50 to 75 milligrams and the other two after increasing the dose from 100 to 150 milligrams.

Three of them had been on Effexor, Paxil, or Lexapro prior to Zoloft without having developed muscle problems.

All patients received 150-300 milligrams per day of riboflavin, while two patients also received two grams per day of carnitine, and one also received 1050 milligrams per day of CoQ10. Four fully recovered, four got better, and one showed no improvement and currently suffers from trouble breathing, sensory disturbances, and pain and weakness in the muscles.

All of the patients showed normalization of the fat content of their muscles despite one remaining symptomatic.

Seven of the patients had their acylcarnitines go back up without symptoms returning, implying that the metabolic disorder was reactivating despite continued riboflavin treatment.

Once the retrospective analysis raised suspicion of sertraline, sertraline was discontinued. There were five patients who had elevated acylcarnitines at the time of sertraline discontinuation and in all five cases the acylcarnitines dramatically improved after stopping the sertraline:

The return of elevated acylcarnitines while on riboflavin could imply a few things:

• Sertraline was causing more and more damage over time, and eventually its damage outweighed the benefit of the riboflavin.

• The riboflavin itself was a double-edged sword. On the one hand, it improved the ability to complete the oxidation of fatty acids. On the other hand, excessive long-term high-dose riboflavin-loading encouraged the body to increase the oxidation of fatty acids, thereby putting more load into the disrupted pathway.

• While symptoms may not have returned, the return of the elevated acylcarnitines was probably a leading indicator that metabolic disruption was returning and that symptoms would have come in the future if the problem had remained unsolved.

The authors unfortunately did not explicitly comment on whether sertraline withdrawal made the one remaining symptomatic patient better, worse, or no different. They imply that the patient was still symptomatic at the time the report was written, indicating that sertraline had caused permanent or quasi-permanent damage, or that something beyond sertraline discontinuation was required to fix it.

It is also unclear whether “no clinical improvement” means this patient remained entirely in the same state or had actually gotten worse on the original riboflavin-based treatment.

Sertraline-Induced MADD in Western Sweden

The group from western Sweden found 17 individuals diagnosed with a lipid storage myopathy at the neuromuscular center in Gothenburg between 2015 and 2023. Two had a clear genetic diagnosis. Of the remaining 15, 12 agreed to further investigation, during which it was revealed that 11 of them were on sertraline at the time of diagnosis. These 11 were then further studied.

They all had weakness in the muscles closest to the center of the body, while most also had weakness in their extremities and about half had weakness in their neck. Most had muscle pain and sensory disturbances, while three had trouble swallowing.

Muscle biopsy showed fat accumulation in their muscles. They measured mitochondrial respiratory chain protein expression, but not enzymatic activity. There are dozens of proteins in the respiratory chain complexes, so they measured a representative protein from each complex. They also stained the tissues using antibodies to complexes I, II, and IV but not complex III. Overall these results suggested a profound defect in complex I, with mild and variable deficits of complexes II and IV and no evidence of a deficit in the more limited measurements of complex III.

Their mitochondria were abnormally small and numerous, with small numbers of them being abnormally big, and some appearing round with abnormal structures inside them.

They had no evidence of harm to mitochondrial DNA (mtDNA), though they had almost double the average number of mtDNA copies, suggesting their mitochondria were duplicating their DNA more than normal.

They then studied the total breadth of protein abundance (the “proteome”) of the muscles and found that nearly 2000 proteins were expressed differently in these patients compared to healthy controls. Proteins involved in helping the respiratory chain complexes assemble were upregulated, while the actual proteins that make up the respiratory chain (the subunits) tended to be downregulated. Complex I subunits were strongly downregulated, complexes II and IV more weakly so, and complexes III and V (V is ATP synthase) remained the same or were upregulated.

Most enzymes involved in fatty acid oxidation were upregulated, but ETFDH, which encodes ETF dehydrogenase and funnels electrons to the CoQ10 pool of the respiratory chain, was downregulated.

These last two changes alone would explain elevated acylcarnitines, since fatty acid oxidation being upregulated means there is more input into the pathway whereas ETFDH being downregulated means the pathway cannot run to completion. Fatty acids that start getting oxidized but can’t finish getting oxidized are the main driver of acylcarnitine accumulation.

Citrate synthase was upregulated, probably reflecting greater numbers of mitochondria. Together with doubled mtDNA content, this suggests increased mitochondrial biogenesis.

These patients had no evidence of genetic variants that could alter sertraline metabolism.

The authors did not measure riboflavin status, nor did they report treating these individuals with riboflavin or with sertraline withdrawal.

Sertraline-Induced MADD in North America

The North American group reported two cases of sertraline-associated MADD.

The first was a 26-year-old woman diagnosed with depression, profound fatigue, “psychogenic non-epileptic seizures,” subclinical hypothyroidism, and heart palpitations. Her acylcarnitine profile appeared as MADD.

She was counseled to take riboflavin and carnitine, avoid fasting, and to increase caloric intake when ill. This did not normalize her acylcarnitine profile.

However, going off sertraline did normalize her acylcarnitine profile, and going back on sertraline made it abnormal again:

There were two measurements showing it normalized while off sertraline rather than just one, so this is especially compelling evidence that the sertraline was elevating the acylcarnitines and that random variation was not creating this as an illusion.

When combined with the data from southeastern Sweden, where five people all had their acylcarnitines normalize when they went off sertraline, we have two different angles – multiple measurements in one person, single measurements in multiple people – arguing against random variation and in favor of a real effect of sertraline.

After three months of staying on riboflavin and carnitine while eliminating sertraline, the first North American case reported “mild improvement in fatigue.”

The second case was a 61-year-old woman with chronic fatigue, exercise intolerance, weakness in the muscles closest to her trunk, and excessive tightness in her legs.

After a few months of these symptoms, she went to the ER after fainting and having difficulty swallowing and opening her mouth.

The ER found her in metabolic acidosis with rapidly increasing creatine kinase, a MADD-like acylcarnitine profile, and normal urinary organic acids.

Muscle biopsy showed fat accumulation and excessive proliferation of mitochondria with abnormal structures inside them.

Analysis of her respiratory chain activity showed isolated complex II deficiency.

Further investigation revealed she was on sertraline and had increased her dose of sertraline to 100 milligrams per day shortly before her ER visit. She was advised to discontinue it, and this made her creatine kinase rapidly return to normal.

She was further advised to take riboflavin and to moderate fat and protein.

After two years of taking riboflavin but avoiding sertraline, a repeat muscle biopsy showed her complex II activity had returned to normal and she no longer had fat accumulation.

Like the Swedish groups, the North American groups did not measure anyone’s riboflavin status.

It is very odd that they do not report anything about clinical improvement of the second case. Did her chronic fatigue, exercise intolerance, muscle weakness, and tightness improve as her muscle biopsy and respiratory chain activity normalized? Did she have any more episodes of fainting or difficulty swallowing or opening her mouth?

The silence on this question suggests she did not clinically improve, as was seen in one riboflavin-treated patient in the southeastern Swedish paper.

The first case had only “mild improvement” in fatigue, suggesting she was resistant to both riboflavin and sertraline discontinuation.

Sertraline-Induced MADD in Australia

In the Australian cohort, five of the eighteen cases of adult-onset MADD had traditional diagnostic criteria fulfilled for genetic MADD. Of the thirteen remaining, ten were on sertraline. Of the remaining three, one was on venlafaxine (Effexor) and one was on duloxetine (Cymbalta), both of which are SNRIs rather than SSRIs, which means they also inhibit the synaptic reuptake of norepinephrine alongside serotonin. Eight of the ten sertraline-associated cases had muscle biopsies performed, and all of them showed abnormal fat accumulation in the muscle. The patient on venlafaxine had a muscle biopsy and it showed mild inflammation but not fat accumulation.

Only eight of the eighteen adults in the combined adult-onset MADD cohort had riboflavin tested prior to starting riboflavin. Of them, half had low or borderline low riboflavin before starting supplementation. Of those who had normal riboflavin status, one was not prescribed riboflavin and the other three were. They did not respond to riboflavin at all. Among the four who had suboptimal status, three responded well to riboflavin treatment. While baseline riboflavin status was a critical determinant of the response to riboflavin, the genetic analysis had no ability to predict the response to riboflavin.

This again underscores that riboflavin “treatment” in MADD is overwhelmingly about optimizing riboflavin status and not on “treating” a highly specific genetic requirement for supraphysiological riboflavin concentrations.

All three patients who responded well to riboflavin had genetic MADD.

The anti-depressant associated MADD had a unique and different response to riboflavin. Riboflavin normalized the acylcarnitine profile in three and improved it in two, while failing to do anything in the others. The one patient taking venlafaxine had the acylcarnitine profile do nothing at first, but normalized six months after starting riboflavin. All of the patients whose acylcarnitine profile normalized or improved after riboflavin supplementation had them subsequently worsen several months later.

This is exactly what was observed in the southeastern Swedish cohort: riboflavin initially improved the acylcarnitine profile, then it worsened. The only difference is the Australian study did not take the patients off sertraline, so it was not able to show the renormalization of the profile after discontinuing the drug.

Of the four adult patients with suboptimal riboflavin status, three were on sertraline. Riboflavin induced complete recovery only in the one patient who was not on sertraline. In the three on sertraline, one had no response and the other two appeared to improve biochemically at first but then reverted back to their initial biochemical profile without improving clinically.

This is inconsistent with the tendency to improve clinically in the southeastern Swedish cohort and it emphasizes that there is a subset of people with sertraline-induced MADD who do not get better on riboflavin.

Unfortunately, the Australian paper did not report the doses of riboflavin used, and left us with the cryptic statement that they are now “trialing higher dose riboflavin in these patients” to see if it can indeed overcome the effect of sertraline. They speculated without any evidence that sertraline may interfere with the binding of FAD to ETF or ETFDH and therefore it may respond to supraphysiological riboflavin doses.

It is bizarre that they are not “trialling sertraline withdrawal” as the Australian medical establishment, unlike the North American or Swedish establishments, evidently consider it important to maintain their patients on the mitochondrial toxin they were originally prescribed.

The Bottom Line

The signal for sertraline here is very powerful. Other SSRIs and the SNRIs do not seem compellingly associated with MADD. The two Australian cases on SNRIs are counterbalanced by the three Swedish cases where other drugs used prior to sertraline did not cause MADD in the same individuals in whom sertraline did cause MADD. The North American case where two measurements off sertraline produced a fully normalized acylcarnitine profile in one individual while going back on sertraline caused it to resume its MADD-like state alongside five Swedish cases where going off sertraline all made an abnormal acylcarnitine profile normalize offers clearer evidence that this is causation rather than a simple correlation.

The fact that sertraline-induced MADD is 45 times more common than genetic MADD, that it is harder to fix, and that 30 cases were found almost immediately in four out of 197 countries while only about 600 total genetic cases worldwide have ever been reported suggests this is the tip of the iceberg and that this is a rare but powerful and dangerous effect of sertraline.

Indeed, how many people with fatigue, weakness, and muscle pain even get an acylcarnitine profile tested?

If this only occurs in one in 5500 people, or even if it is 100 times more common than this and it is one out of 55 people, this still implies it is a genetic idiosyncrasy creating a predisposition for it to occur. Within some relatively uncommon genetic constraints, however, it becomes a predictable dose-dependent effect of sertraline.

Several heterozygous mutations in the ETFDH system, riboflavin transport, or fatty acid oxidation had been potentially tied to some of these cases, but the defining feature of sertraline was that there was no basis for a classical genetic diagnosis involving two pathogenic mutations in the same enzyme, and even the Australian paper which cast the broadest net for genetic investigation could only find a single heterozygous mutation in three out of ten cases.

That doesn’t mean genetics aren’t involved. It suggests instead that a much broader net needs to be cast to find the relevant genetic predisposition.

Not a single one of these papers mentioned hypoxia or the sigma-1 receptor, showing that pharmacology, psychiatry, and the specialists in inborn errors of metabolism are fully united in completely misunderstanding the role of serotonin in the body and the intracellular impacts of SSRIs.

None of these cases will ever be solved until these disciplines exit the intellectual stranglehold in which SSRIs are seen as “psychiatric drugs” rather than “mitochondrial drugs.”

The question arises whether this can be explained by any of the known effects of sertraline.

The increased mitochondrial biogenesis implied in the western Swedish paper is consistent with the ability of extracellular serotonin to activate this process, though it could also be a secondary compensation for poorly functioning mitochondria.

The observations about the respiratory chain patterns in the western Swedish and North American papers seem to conflict with one another or imply clinical heterogeneity, but they cannot be directly compared due to fundamental differences in methodology. Nevertheless, they are both consistent with variations of the hypoxia response, which is centrally impacted by SSRIs.

Sertraline stands out as a powerful sigma-1 activator, a close second to fluvoxamine. Unfortunately, no studies have directly compared the rate of sertraline and fluvoxamine uptake into cells. It could be that while fluvoxamine is a more powerful activator of the sigma-1 receptor, sertraline enters cells much better than fluvoxamine and that this is enough to more than make up for the gap.

In general, SSRIs that strongly activate the sigma-1 receptor are expected to be better for mitochondrial function than those that activate it poorly or trivially.

However, there could be a genetic mutation found in these cases that fundamentally changes the importance of chronic sigma-1 activation. In fact, it could even alter the way sertraline binds to it without altering the way fluvoxamine binds to it. Since no one investigating this appreciates the importance of sigma-1 activation to sertraline’s impact on mitochondrial function, no one has looked at it.

Alternatively, SSRIs differ in how they structurally interact with the serotonin transporter, so there could be a mutation that makes sertraline especially potent in blocking the entry of serotonin into the mitochondria where it is needed to synthesize melatonin. Or perhaps it even makes it less likely for sertraline to dissociate from the extracellular serotonin transporter and therefore enter the cell in the first place.

It is not clear to what degree the elevated acylcarnitines reflect the centrally causal event in metabolic dysfunction or reflect impaired mitochondrial respiration as centrally causal. As noted in the beginning, impairments in the respiratory chain can elevate acylcarnitines. While the western Swedish paper also found elevated fatty acid oxidation and suppressed ETF dehydrogenase, it can’t be ruled out that this is secondary to hypoxia response activation, which can have variable impacts on fatty acid oxidation depending on the intensity and duration of hypoxia.

Thus, there is much more to learn about this, but it appears specifically tied to sertraline, clearly indicated by broadly elevated acylcarnitines, and sometimes but not always responsive to riboflavin supplementation and sertraline discontinuation.

Sertraline discontinuation can be complicated by the fact that SSRI withdrawal can cause mitochondrial dysfunction. However, I suspect that this would not occur in the case of sertraline-induced MADD. Neither of my clients who suffered catastrophic effects from SSRI withdrawal had anything that looked like this phenomenon, despite one who was on sertraline for years, and while clinical improvement was reported for some but not all sertraline-induced MADD patients, none of the papers reported any problems during withdrawal.

The ideal way to address riboflavin would be to first measure blood levels. Especially if they are low, focus should be placed on getting more riboflavin from food. A “riboflavin flush” with a high dose for two weeks might be valuable as a way to fix the deficiency much faster, but it should be followed by going back to food-level riboflavin. Supplemental riboflavin can be titrated up from 6.5 milligrams to 37 milligrams to 100 milligrams using liquid riboflavin, the Thorne product, and a standard 100-milligram product. You can also take a 100-milligram capsule and weigh it out into four 25-milligram capsules using empty capsules and a milligram scale to give you more control over the dose. In such a titration you want to slowly find the minimum dose that is needed to produce the maximally effective result. Going higher than this should be safe in most cases but should only be done if increasing the dose continues to generate big dividends, or if lower doses do nothing and you are temporarily exploring whether higher doses might work. Your aim should be to titrate up to the point of diminishing returns and then stabilize there or slightly pull back to the dose that gave the most recent very large benefit. Starting with the “flush,” going back to baseline, and then engaging in a slow upward titration allows you to more precisely define the dose you need without being fooled by it taking months to fix a deficiency.

The indications in two of these papers that variants in the hypoxia response could be creating blocks in the respiratory chain raise the possibility that some of the patients who never improved on riboflavin and sertraline withdrawal may have needed to fix their respiratory chain.

My Mitome test uses an interpretive algorithm that I invented that detects these variants of the hypoxia response using pattern analysis of the respiratory chain enzyme activity and generates a protocol to address each variant. This test does not diagnose or treat disease, but it can help with biochemical optimization where such bottlenecks in energy production are suspected.

We have now seen that both SSRI use and SSRI withdrawal can cause mitochondrial dysfunction, with poorly understood genetic idiosyncrasies determining which of these outcomes any given person lands on.

Playing with SSRIs is playing with fire.

The risk is you set your mitochondria on fire.

In the next installment, we look at vastly superior approach to mental health that can avoid these catastrophic results altogether.

Read it here:

Here we will examine the concept that both SSRI withdrawal and SSRI switching can cause mitochondrial dysfunction.

This is the seventh installment in our series on the truth about serotonin and SSRIs and builds on the information in the last installment, SSRIs Are Mitochondrial Drugs.

This is educational in nature and not medical or dietetic advice. See terms for additional and more complete disclaimers.

The first individual had no issues being on Paxil long-term nor going off Paxil for the first time, but when he went on Paxil a second time he developed tachycardia and muscle cramps and spasms during sleep. Switching to Lexapro made these worse. Withdrawing from Lexapro then added to these problems a whole list of new problems: sexual dysfunction, peripheral neuropathy, shortness of breath, anxiety, muscle cramping, numbness, and excessive urination.

The second individual developed sexual dysfunction on Zoloft and chronic fatigue in the wake of Zoloft withdrawal in his 20s. At its worst point, he spent a year not being able to stand in the shower and a month needing to be bathed by another person and needing a portable bedside toilet.

Despite these problems onsetting during the withdrawal from Zoloft, Prozac immediately caused extreme spaciness and a low dose of Celexa caused restlessness. He is not confident that Zoloft would not have caused the same problems at that time because he felt hypersensitive to everything. However, the hypothesis that switching SSRIs did indeed worsen his situation should be considered, because SSRIs have different impacts on mitochondrial function.

Both people received individualized protocols based not on generalized information about SSRI withdrawal, but based on mitochondrial testing, whole genome sequencing, and comprehensive biochemical data.

The second just received his protocol so has no results to report yet.

The first received his protocol in November of last year. Despite “having gone to see dozens of specialists who could not tell me what was going on,” six weeks into his protocol he reported “the first time I felt normal in over two years.”

At the six-month mark, he wrote “six months later, I have regained quality of life and am able to do everything I used to be able to.”

After 8.5 months, he wrote “I've still been pretty amazed how effective your protocol has been. On the whole I feel like a whole different person from a year ago. It's wild. I've also gained close to 15 pounds in lean muscle mass since January. Making gains in the gym daily and I think I can end the year with another 10 pounds of mass. Been crushing it at work and also traveling and feel comfortable in my body for the first time in 2-3 years. Excited to see what continued progress I can make!”

The Mitome results of the two individuals were very different. Importantly, neither of them had Mitome tested in the immediate wake of SSRI withdrawal, so in each case the results reflect the combination of genetics, nutritional status, and prior experience, including the catastrophic effects of SSRIs and being off them for several years. The first showed a pattern associated with deficient methylation expected to suppress the hypoxia response, while the second showed a pattern associated with the impact of an excessive hypoxia response.

Whole genome sequencing suggested the first individual’s results were caused by a heterozygous defect in SLC25A26, which prevents methyl groups from being transported into the mitochondria, while the second individual’s results were caused by a heterozygous gain-of-function polymorphism in a receptor for TNF-alpha that creates a chronically increased inflammatory tone, which primes cells to easily activate a hypoxia response that suppresses mitochondrial respiration.

The second case is rather straightforward to explain: the hypoxia response is antagonized by sertraline (Zoloft) but exacerbated by fluoxetine (Prozac) and citalopram (Celexa). Thus, while the second individual remains deeply uncertain whether these different SSRIs had genuinely different impacts on him at the time they were tried, the response is consistent with the experimental evidence suggesting these different SSRIs have disparate impacts on the hypoxia response.

The first case is less straightforward to explain because paroxetine (Paxil) appeared to be harmless upon first use and withdrawal but the negative reaction to the second use implies that yo-yoing on paroxetine can be particularly harmful. This yo-yoing effect could be seen as a disorganized response to the conflicting positive and negative effects of paroxetine on mitochondrial function:

• Paroxetine has the least activation of the sigma-1 receptor out of any SSRI, which suggests its ability to increase extracellular serotonin will lead to a high ratio of mitochondrial biogenesis to mitophagy, and this is consistent with rodent studies showing it increases markers of mitochondrial density.

• While the hypoxia response usually opposes mitochondrial biogenesis, the unnatural level of extracellular serotonin achieved with SSRIs will activate both processes. Paroxetine’s lack of sigma-1 activation makes it deplete brain serotonin instead of increasing brain serotonin like the SSRIs that are powerful sigma-1 activators, which means it will increase the anti-respiration aspects of the hypoxia response without supporting the pro-respiration aspects mediated by intracellular serotonin and melatonin.

• This individual was probably hyper-tolerant of the hypoxia response promotion due to his idiosyncratic mitochondrial pattern expected to suppress that response. So, the initial experience of paroxetine may have been simply balancing out the hypoxia response in his particular case, while its promotion of mitochondrial biogenesis may have been quite useful.

• However, withdrawing from paroxetine could have caused a rebound decrease in mitochondrial density that created a greater vulnerability to the toxic actions of paroxetine that became evident upon reusing the drug. This could include the imbalanced hypoxia response, direct inhibition of mitochondrial serotonin receptors, and direct inhibition of respiratory chain complexes. While he may have initially tolerated the imbalanced hypoxia response, this tolerance could have been lost as a result of the rebound decrease in mitochondrial biogenesis.

Lexapro (escitalopram) shares with paroxetine (Paxil) the increase in extracellular serotonin and the promotion of respiration-impairing aspects of the hypoxia response. However, it is a ten times more powerful sigma-1 activator. It is nowhere near as powerful as fluovaxamine or sertraline, but it is much more powerful than paroxetine. This would be expected to balance mitochondrial biogenesis with mitophagy and to balance the hypoxia response with the preservation of respiration caused by intracellular serotonin and melatonin. However, it is also the most mitochondrially toxic of the SSRIs when compared head-to-head using isolated mitochondria. It is especially toxic toward complex IV compared to other SSRIs, and his complex IV was already hurt by his impairment in methyl group transport.

If the paroxetine yo-yoing effect was indeed caused by a rebound drop in mitochondrial density, then escitalopram’s stimulation of mitophagy via the sigma-1 receptor may have aggravated this with a further drop in mitochondrial density.

One thing we have to understand here is that, while the regulation of mitochondrial biogenesis is ordinarily stimulated by an energetic deficit, which signals a need for more mitochondria, the actual process of building new mitochondria is incredibly energy-intensive. It is anabolic. It cannot occur without energy no matter what the regulation says should be happening.

Given all that, this is the likely sequence of events:

• The yo-yoing of paroxetine caused a rebound drop in mitochondrial density after the first withdrawal that flipped the response to a second use in favor of vulnerability to its mitochondrial toxicity.

• This then caused energetic dysfunction that prevented the second use of paroxetine from rebuilding the lost mitochondria.

• This then allowed the relative balance favoring of mitophagy by escitalopram to worsen this.

• The lower mitochondrial content of tissues primed them to respond poorly to the higher mitochondrial toxicity of escitalopram.

The astounding benefits he obtained from his protocol were NOT based on trying to fix the catastrophic destruction induced by SSRI use, withdrawal, and switching, but rather on a coherent theory of his biochemical idiosyncrasies at the time I worked on his case informed by insights into which were genetic in nature based on the analysis of his whole genome sequencing. The lion’s share of the benefits were deducible from Mitome alone, though whole genome sequencing and nutritional/biochemical analysis enhanced the interpretation.

How much of this is generalizable across SSRI withdrawal? Let’s take a look at the literature.

In This Article:

• SSRI Discontinuation Syndrome

• Post-SSRI Sexual Dysfunction

• Prozac Withdrawal Causes Mitochondrial Dysfunction in Mice

• Near Infrared Light Reverses Genital Anesthesia

• Mitochondrial Dysfunction Tanks Testosterone

• Mitochondrial Dysfunction Causes Neurological Dysfunction

• How Common Is Mitochondrial Dysfunction as a Mediator of SSRI

• Discontinuation Syndrome and PSSD?

Share This Article!

Share

SSRI Discontinuation Syndrome

SSRI discontinuation syndrome occurs in over half of people who stop SSRIs according to observational studies, and in about twenty percent of people in randomized controlled trials. Even using the smaller number from the trials, there are millions of people worldwide who are undergoing cessation of SSRIs at any given moment, suggesting there are always hundreds of thousands of people experiencing SSRI discontinuation syndrome in any given moment.

This can involve irritability, anxiety, mood problems, crying, dread, suicidal ideation, insomnia, nightmares, excessive dreaming, lethargy, fatigue, headache, tremor, sweating, anorexia, flu-like symptoms, nausea, vomiting, diarrhea, pain, numbness, tingling, feeling like something is crawling on the skin, electric shocks running through the brain or body, rushing noises, visual traces (seeing something persist when it is no longer there, or seeing moving objects leaving illusory streaks of light behind them, etc), dizziness, light-headedness, vertigo, confusion, difficulty concentrating, amnesia, genital hypersensitivity, and premature ejaculation.

The sexual side effects of SSRI discontinuation are not considered to include the genital numbness and erectile dysfunction characteristic of post-SSRI sexual dysfunction (PSSD), but as we will consider below PSSD can indeed become a long-lasting SSRI side effect that doesn’t start until after SSRIs are stopped, meaning that a subset of PSSD is indeed a subset of SSRI discontinuation syndrome.

SSRI discontinuation syndrome tends to require more than one week and perhaps six to eight weeks of SSRI treatment as a pre-requisite. It occurs more often in men than women, younger individuals than older, in those that stop SSRIs abruptly instead of tapering, and more often with Paxil (paroxetine) than Prozac (fluoxetine) or Lexapro (escitalopram). It can appear as early as 24 hours after the last dose, usually resolves in two to three weeks, but can last longer, even months or years.

Analysis of internet forums shows that “protracted withdrawal syndrome” can last anywhere from five months to almost fourteen years. The average person reporting such protraction of their discontinuation symptoms was using SSRIs for eight years prior to withdrawal.

A survey of people in the UK-based National Health Services found that whether withdrawal problems occurred increased progressively as the length on the SSRI went up: 64% for those using them less than six months; 86% for those using them from 6 months two years; and 96% of those who had used them for more than two years.

38% reported trying to stop but being unable to do so.

Those reporting “severe” withdrawal symptoms were much more likely to have used SSRIs for a longer period of time: 7% of of those using them less than six months; 15% of those using them between six months and two years; and 25% of those using them for longer than two years.

Withdrawal symptoms lasting longer than a year was reported by 7% of those who had been on SSRIs less than six months and 12% of those who had been on them for more than two years.

The most common “non-emotional” problems reported by the majority were headache, derealization, depersonalization, and lightheadedness. 28% reported “brain zaps,” 27% reported muscle cramps, 27% reported nausea or vomiting, 24% reported diarrhea, and 30% reported trouble walking stably.

An in-depth study of “brain zaps” where people were approached through forums on mentalhealthdaily.com found that the onset to the first zap was eight days with fluoxetine and two days with paroxetine, suggesting the big difference in half-life of these two drugs primarily led to a difference in how long it took to experience the zaps.

52% said the “zaps” felt like an electric shock inside the head, 18% said it felt like a vibration, 12% said they could hear the zaps, 4% said they could see the zaps, 7% said it felt like a pop or a twitch, and a small number of people described it with such terms as a strobe without a light, a crackle, their brain skipping a beat or blinking, or as if the video of their mind was buffering.

The most common triggers of brain zaps were moving the eyes or the head from side to side.

The most up-to-date reviews mention NOTHING about mitochondria in their explanations of this syndrome. The common explanation for why Paxil has a higher prevalence of discontinuation syndrome is that is is much more rapidly cleared from the body. This is an extremely unsatisfactory explanation, however, because this should be easily solved with a more drawn-out tapering.

These reviews mention NOTHING about the sigma-1 receptor. In fact, the most recent review on the mechanisms of SSRI discontinuation syndrome is part of an anthology titled “Emerging Neurobiology of Antidepressant Treatments” and is written by one of the editors of the anthology yet the entire anthology is over 300 pages and does not mention sigma-1 receptors even once. Thus, there is no discussion of how Paxil stands out as the weakest activator of the sigma-1 receptor out of all SSRIs, ten times weaker than Lexapro and twenty times weaker than Prozac. This positions it as uniquely powerful in its ability to deplete total brain serotonin, making the sudden absence of its ability to sustain extracellular brain serotonin much harder to deal with.

In rats, fluoxetine produces an initial decrease in brain serotonin lasting one to three days, which then recovers to a slightly higher than baseline level thereafter. Withdrawing from fluoxetine causes a much larger initial drop in brain serotonin that takes seven days to recover from.

This is consistent with fluoxetine being a moderate sigma-1 activator. Initially, a higher extracellular serotonin concentration causes greater activation of serotonin receptors that exert negative feedback on serotonin synthesis. However, its activation of the sigma-1 receptor then causes a broad increase in intracellular serotonin synthesis. The rats then become dependent on the constant sigma-1 activation to counter the constant overactivation of serotonin receptors. When fluoxetine is withdrawn, the drop in sigma-1 activation causes a sudden drop in serotonin synthesis, but then the return to normal extracellular serotonin signaling allows this to even out.

By contrast, treatment of mice for twelve days with paroxetine causes a decrease in brain serotonin and withdrawal leads to a rebound above baseline at least as early as day 2. In some areas of the brain, the rebound is mild and increases further into day 5. In others, it is pronounced, and it levels off by day 5.

The authors of these studies use individual SSRIs as a representative of the drug class and then generalize about the discontinuation syndrome. This leads to fundamentally incoherent conclusions.

For example, the authors of the latter study — done by the same group that wrote the first-cited review, one of whom was an editor of the anthology — suggest that SSRI discontinuation syndrome is driven by rebound overactivation of the serotonin system.

Separately, they claim that paroxetine has a higher rate of discontinuation syndrome than fluoxetine because it leaves the body faster.

If the latter claim is true, paroxetine should do the same thing to brain serotonin during withdrawal as fluoxetine, but should do it faster. Yet they simply have opposite effects: equilibration to fluoxetine leads to a slightly higher brain serotonin level than baseline, and equilibration to paroxetine leads to a considerably lower level than baseline; discontinuation of fluoxetine leads to a big drop in brain serotonin and discontinuation of paroxetine leads to an increase.

This recent paper, like the more than 300-page anthology, makes no mention of sigma-1 receptors. Yet these receptors easily explain the results: fluoxetine maintains serotonin synthesis by activating this receptor and paroxetine fails to do so. Therefore, the negative effect on brain serotonin of activating the extracellular serotonin receptor dominates the effect of paroxetine, so serotonin goes down on the drug and up during withdrawal; the sigma-1 activation caused by fluoxetine is the slightly dominant effect, so serotonin goes slightly up on the drug and down during withdrawal.

Yet they are both associated with SSRI discontinuation syndrome, making rebound increases or decreases in serotonin poor explanations of the syndrome, at least in isolation and treating the syndrome as if it is a single pathological entity.

Another hypothesis to explain the higher rate of discontinuation syndrome with Paxil is that it has anticholinergic activity by binding to and inhibiting acetylcholine receptors, so its withdrawal will cause a rebound excess of acetylcholine activity. But this is a unique function of Paxil among the SSRIs and is similar to the older tricyclic antidepressants, whereas between 20% and a majority of all SSRI users get this syndrome, and the symptoms expected from a cholinergic rebound are limited to a relatively small portion of the discontinuation symptoms. So it could play a minor role in Paxil withdrawal but not an overall major role in the syndrome.

Paxil also has modest inhibition of dopamine and norepinephrine reuptake, so Paxil withdrawal could lower activity of these neurotransmitters, but the criticisms of the anticholinergic hypothesis also apply here.

The most up-to-date reviews admit this syndrome is poorly understood, and recommend tapering SSRIs, but also acknowledge that “present guidelines for managing SSRI discontinuation are based on dose tapering and not a knowledge of the neural mechanisms underlying discontinuation.”

In making this statement, they also betray the fact that they are intellectually trapped into the belief that the mechanism is fundamentally “neural” when the answer lies in accepting that SSRIs are whole-body drugs acting primarily on cells that aren’t neurons, and that what they do to neurons and what they do to other cells fundamentally involves messing with serotonin and melatonin’s impact on mitochondria as well as the SSRIs themselves entering the cells and activating other receptors such as the sigma-1 receptor.

Post-SSRI Sexual Dysfunction

Sexual dysfunction was at first recognized as occurring on SSRIs. Post-SSRI Sexual Dysfunction (PSSD) was recognized after evidence accumulated that this sexual dysfunction could persist after discontinuing SSRIs. The most commonly reported problems in men are genital anesthesia (total loss of any feeling in the genitals), loss of sexual pleasure, weak orgasms, decreased libido, erectile dysfunction, and premature ejaculation.

PSSD is often associated with general anhedonia, apathy, and poor mood.

Genital anesthesia, and, in women, loss of sexual sensitivity in the nipples, are according to some reviews specific problems associated with PSSD that are not reported in other forms of sexual dysfunction. However, genital anesthesia also accompanies similar syndromes that endure after the use of finasteride, which inhibits the conversion of testosterone to the more potent DHT, and isotretinoin, a compound that mimics some aspects of vitamin A but otherwise acts as a vitamin A antagonist. The incidence of PSSD in SSRI users has been estimated to be 0.46%.

Researchers and clinicians dealing with PSSD have speculated that genital anesthesia results from damaged spinal cord neurotransmission; that decreased libido results from impaired dopamine transmission in the brain; that erectile dysfunction could result from SSRIs altering the function of oxytocin or androgens; and that weak orgasms result from decreased oxytocin and beta-endorphin release. They have further speculated that multiple rather than single uses of SSRIs, preexisting disorders of glutamate neurotransmission, MTHFR variants, and abuse of recreational drugs could all increase the risk of PSSD.

PSSD is consistently described in most reviews not as sexual dysfunction arising after the withdrawal of SSRIs, but rather sexual dysfunction arising often within days or weeks of starting SSRIs, occasionally within 30 minutes of the first dose, and either persisting or worsening after withdrawal of the medication. However, more broad-minded reviews see a more general syndrome associated mainly with use of SSRIs, isotretinoin, and finasteride, and possibly with non-SSRI neuropsychiatric drugs like buprenorphine, ondansetron, quetiapine, lithium, and haloperidol, and note that it can have its onset after the cessation of the drug. Most of the non-SSRI neuropsychiatric drugs had been co-prescribed with SSRIs, however, so the best high-level perspective is that this is a syndrome that can have its onset soon after the initial use, any time during the use, or after the cessation of SSRIs, finasteride, or isotretinoin. The syndrome can last for as little as weeks or months, but has been documented to last, in the longest case, for at least 18 years.

Even very recent reviews have a nearly complete dearth of useful actionable information or genuine insights into the causation of these syndromes. Speculations about the mechanisms are at a level of proximate cause (for example, decreased neurotransmission through the spinal cord driving genital anesthesia), not ultimate cause, and therefore shed little light on what could be done about the problems. Reviews claim that most case reports fail to support any benefit of testosterone replacement therapy, PDE5 inhibitors, dopamine-boosting drugs, or any other drugs that have been tried.

There is, by contrast, support for the use of bupropion and PDE5 inhibitors from subjects with SSRI-induced sexual dysfunction who are still on SSRIs. This includes Viagra improving erectile dysfunction in a meta-analysis of multiple trials; a single, small, double-blind study using bupropion as an adjunct therapy for broader sexual dysfunction; and a small number of case reports using bupropion similarly. The bupropion trial selected only individuals who were presently using SSRIs and had been using them for at least three months. There was a 20% increase in the desire/frequency subscore of the sexual function test, but there was no statistically significant difference in the total score. Notably the baseline scores were 40 out of 70, where the cutoff for sexual dysfunction is 47 for males and 41 for females, indicating only mild sexual dysfunction.

The commonality of this syndrome across exposure to SSRIs, finasteride, and isotretinoin suggests these drugs could share a primary mechanism in common. Given the importance of testosterone to sexual function, one possibility is that they all do what finasteride is designed to do: lower the conversion of testosterone to DHT.

In one study, isotretinoin in the amount of 1 milligram per kilogram of bodyweight for 20 weeks lowered DHT in women but not men; in a second study, 0.75 mg/kg isotretinoin for twelve weeks lowered DHT by 21% in men, and the resulting ratios of metabolites suggested it inhibits 5-alpha-reductase just like finasteride does. However, in vitro, it has been shown to inhibit DHT production through an alternative pathway dependent on 3-alpha-hydroxysteroid dehydrogenase and no in vitro studies have examined whether it directly inhibits 5-alpha-reductase or decreases its expression.

Bizarrely, despite the existence of this syndrome and its establishment in the literature, there are no studies examining the effect of SSRIs on DHT.

Viagra prevents the breakdown of cGMP, a mediator of erections. cGMP synthesis is initiated by nitric oxide. Nitric oxide is also the mediatory of the parasympathetic branch of the sexual nervous system which mediates the totality of arousal. Several serotonin receptors are capable of increasing or decreasing nitric oxide synthesis. Sertraline and escitalopram have both been shown to decrease nitric oxide production. Therefore, SSRIs probably decrease nitric oxide production on average, and Viagra helps to counteract the effects of this on cGMP.

Buproprion has complex actions but one is to increase the activity of dopamine and norepinephrine by inhibiting their reuptake from synapses, so this can easily explain its benefit to sexual function across the board.

However, the fact that these drugs help does not necessarily tell us anything about the root cause of PSSD. Viagra will cause erections in any man, and buproprion improves sexual function even in women who that are not depressed and have never used SSRIs.

Prozac Withdrawal Causes Mitochondrial Dysfunction in Mice

The only study I am aware of that looked at the impact of SSRI withdrawal on mitochondrial function was a study administering Prozac to rats. This one study appears to have generated several papers, including one on respiratory chain complexes and another on citrate synthase, which is a citric acid cycle enzyme that can serve as a useful but imperfect marker of mitochondrial density.

They administered a low and high dose for 28 days.

Two hours after the first dose, the Prozac increased complex I in the hippocampus. On day one, Prozac was beneficial.

Two hours after the last dose, 28 days later, Prozac had no effect at all. After a month, the benefit had worn off.

Twenty-four hours after the last dose, however, the rats withdrawing from the high dose of Prozac had reduced complex II + III activity in the striatum and those withdrawing from both doses had reduced complex IV activity in the hippocampus. During withdrawal, the respiratory chain was harmed.

On day one, Prozac increased citrate synthase in the striatum. In the hippocampus and prefrontal cortex it seemed greater than in controls but it wasn’t statistically significant. Two hours after the last dose citrate synthase seemed lower in the striatum, but it wasn’t statistically significant. 24 hours after the last dose, citrate synthase appeared unchanged across the board. Overall there seemed to be a mild stimulation of mitochondrial biogenesis that wore off over time.

The net result of this is that Prozac caused an acute improvement in mitochondrial function. Over time, the rats became dependent on it. They adjusted to the drug, and then after 24 hours of withdrawal they had new-onset mitochondrial dysfunction.

Near Infrared Light Reverses Genital Anesthesia

Another hint that SSRI withdrawal is mediated by mitochondrial dysfunction is a case report suggesting genital anesthesia can be reversed with near infrared light.

One case report showed that low-powered laser in the near infrared range on the scrotum and spine led to a 40% increase in penile sensitivity.

Near infrared is obtainable from the sun as well as home devices like near infrared saunas or lamps. Infrared light acts in part by feeding photons to complex IV (cytochrome oxidase), which improves its ability to facilitate ATP production. Further, near infrared also structures mitochondrial water to improve its viscosity in a way that facilitates greater production of ATP by ATP synthase.

Mitochondrial Dysfunction Tanks Testosterone

In male rats, inhibiting complex I with the fish poison rotenone depletes testosterone.

According to Chapter 10 of Saudubray, Inborn Metabolic Diseases: Diagnosis and Treatment, mitochondrial disorders can result in infertility and hypogonadism.

This can be explained through several mechanisms.

First, the synthesis of testosterone is anabolic and therefore requires energy input. If you tank your energy production, you do not have energy to make testosterone.

Second, testosterone is itself anabolic. If you tank your energy production, you will tell your hypothalamus that you cannot afford to make testosterone, because if you do, you will increase your synthesis of muscle, which is extremely energetically expensive.

Worse, you could wind up having sex.

For men, you could get someone pregnant, and then be responsible for spending precious energy providing for the anabolism of that new human and its future upbringing.

For women, you get get put on the hook for spending your energy on making an entirely new human and then feeding it milk for one to three years thereafter.

Therefore, energetic deficiency makes it more difficult to make testosterone and makes the brain or sex organs restrain its production.

Mitochondrial Dysfunction Causes Neurological Dysfunction

The brain is only two percent of our bodyweight but consumes twenty percent of its oxygen.

Energy must be invested in the anabolic synthesis of myelin, neurotransmitters, and other brain materials and to maintain neurotransmitter function from each fraction of a second to the next.

The release of a neurotransmitter into a synapse, its action on the downstream neuron, its clearance from the synapse, the transmission of an action potential down the neuron, and the trigger for another neurotransmitter to be released at the other end all involves the exquisite control of a symphony of calcium, magnesium, sodium, chloride, and potassium ions that are pumped to one side of a membrane and unleashed to flow back to the other side over and over again, where all the pumps are fueled by the energy contained within the ATP made by the neuronal mitochondria.

According to Chapter 10 of Saudubray, Inborn Metabolic Diseases: Diagnosis and Treatment, mitochondrial disorders can result in the following problems reported during SSRI withdrawal: ataxia, cramps, and peripheral neruopathy.

Muscle contraction occurs when sodium- and potassium-based neural signals cause calcium to be released at the end of a neuron, which causes acetylcholine to be released from the neuron onto the muscle, which then causes calcium to be released inside the muscle cell, which then signals the contraction. Muscle relaxation occurs when ATP puts the calcium back where it belongs. Magnesium is needed to support ATP, explaining the role of magnesium in muscle relaxation. If any of these ions are deficient or are not pumped where they are supposed to go using energy from ATP, you get cramps.

Ataxia can occur from energetic failure in various parts of the nervous system or inner ear responsible for balance and coordination.

The causation of peripheral neuropathy has been reviewed here and here. It is a form of “ectopic signaling,” meaning signaling that is occurring in the wrong place. You can have a physical injury where sprouts of regenerating sensory receptors fire abnormally because they have lost normal connections, or because demyelinated nerves can bind together abnormally and fire abnormally, or can allow their signals to be discharged horizontally, propagating among a broader set of nerves than would normally be impacted.

Physical injury can be seen as a forceful undoing of the work of mitochondrial energy production. This energy is invested in creating order, and when its work is forcefully undone, relative chaos ensues.

If you have primary energetic failure, you get the same result: rather than a physical injury disrupting the work of mitochondrial energy production, the work cannot be done in the first place. Thus, disordered ion flows and neural connections produce ectopic signaling.

The specific abnormal sensations generated during ectopic signaling takes on flavors from the subconscious, from fears, or from the history of past sensations, which is why one person can experience pain and another tingling and another electric shocks.

Some of the problems of SSRI withdrawal are more reminiscent of psychedelics than the common symptoms of mitochondrial dysfunction. For example, visual traces, being able to see or hear brain zaps, and hearing whooshing sounds that aren’t there.

While electric shock-like pain is found in many disorders and can be seen as similar in its causation to peripheral neuropathy, “brain zaps” themselves appear to be a phenomenon relatively isolated to SSRI withdrawal.

Psychedelics act on 5-HT2A receptors to induce psychedelic experiences, but serotonin does not normally induce these experiences even though it activates these receptors. Research suggests that this is because psychedelics but not serotonin indiscriminately enter the cell and bind to intracellular 5-HT2A receptors, located at least partly on the Golgi apparatus, and that they can stay activating these receptors for prolonged periods of time. By contrast, the areas of the brain involved in these perceptual modifications do not substantially express serotonin transporters outside those involved in reuptake of synaptic serotonin under normal conditions. Therefore, serotonin doesn’t usually enter these cells. However, if mice are engineered to highly express serotonin transporters in the relevant neurons the serotonin can enter the cell and it can induce these types of perceptual changes.

As covered in SSRIs are Mitochondrial Drugs, activation of the sigma-1 receptor increases expression of serotonin transporters in the cell membrane. These are normally activated under whole-body or cellular stress of multiple types. The natural ligands are unknown but may include DMT (the chemical from ayahuasca that we also make in small amounts endogenously), choline, sphingolipids, myristic acid, and DHEA-sulfate.

It is likely that in hypoxic stress, monamine oxidase activity declines and DMT rises, which would allow it to activate sigma-1 receptors. However, these conditions also facilitate the movement of serotonin into the mitochondria where it is converted to melatonin rather than acting as a persistent activator of Golgi 5-HT2A receptors. Further, the response is coordinated to normalize energy production under hypoxia as quickly as possible. Nevertheless, hypoxia probably sometimes activates these receptors to cause perceptual abnormalities, since some people can develop psychosis from altitude. Psychological stress, which activates sigma-1 receptors, is also a major contributor to psychotic breaks.

The question arises what on earth serotonin receptors are doing inside the cells of neurons that do not usually express serotonin transporters. Perhaps they are sitting there waiting for LSD or mushrooms to arrive. More likely, various forms of stress activate serotonin entry into the cell, where it does activate those receptors but not usually in a sustained and isolated way that facilitates a psychedelic trip.

SSRI withdrawal could induce sustained activation of these receptors in a few ways we can brainstorm:

• Chronic activation of the sigma-1 receptor may cause a long-lasting increase in serotonin transporters on the membranes of these neurons.

• Chronic deprivation of intracellular serotonin in and of itself causes a long-term feedback loop increasing the serotonin transporter expression on these neurons, primarily to secure sufficient mitochondrial melatonin.

• Mitochondrial dysfunction in the wake of serotonin withdrawal leads to sigma-1 activation, increasing the serotonin transporters on the membranes of these neurons.

The relative specificity of this to SSRI withdrawal could simply result from a bias toward having greater impacts on certain parts of the brain with high concentrations of hypoxia response-promoting serotonin receptors.

The “brain zaps” being associated with side-to-side head and eye movements was suggested by the authors of that study to be related to activation of the vestibulo-ocular reflex, whose job is to stabilize vision by making eye movements compensate for head movements.

This reflex is powerfully activated by serotonin, Prozac, and the “God molecule” psychedelic from toad venom, 5-methoxy-DMT. The commonality of these suggests it is mediated by activation of extracellular serotonin receptors.

Brain zaps are probably a form of “ectopic signaling” that is very similar in nature to peripheral neuropathy, but is a result of specific disruptions to ion gradient maintenance in the neural circuitry of the vestibulo-ocular reflex. Thus, the reflex activates with a lateral eye or head movement and then ions spill into the wrong places and initiate a perceptual abnormality.

This likely reflects mitochondrial dysfunction that has essentially gravitated to an area of the brain where extracellular serotonin is particularly important due to the long-standing use of a mitochondrial drug that is centrally involved in increasing extracellular serotonin.

How Common Is Mitochondrial Dysfunction as a Mediator of SSRI Discontinuation Syndrome and PSSD?

While we cannot answer this with rigorous research, this is because serotonin is almost universally perceived as a “neural” and “psychiatric” molecule that modifies mood. Its central role in the hypoxia response and mitochondrial function and the direct impacts of SSRIs on intracellular sigma-1 receptors and the mitochondrial respiratory chain are entirely ignored in discussions about SSRI discontinuation syndrome.

However, the fact that mainstream psychiatry and pharmacology has no real answers about this, especially for those suffering protracted withdrawal lasting years, indicates that we will learn a lot more when we think outside of these traditional constraints.

Doing so leads to the conclusion that SSRI withdrawal is primarily a problem of mitochondrial dysfunction.

Unlike the one study on withdrawing from Prozac hurting the respiratory chain in mice, I do not know of any studies showing that switching from one SSRI to another causes mitochondrial dysfunction. But given that they have very disparate impacts on mitochondrial dysfunction, it is easy to see how this could happen, and the cases I wrote about at the beginning suggests that switching SSRIs over the course of withdrawal or after yo-yoing can play a role in the energetic dysfunction that develops.

In the next article, we will look at the ability of SSRIs themselves to cause mitochondrial dysfunction, and then we arrive at what can be done about it.

You can read the next article here: