---

Conventional dietary advice to preserve cognitive health is the same as that for cutting cardiovascular risk: cut your fat and saturated fat, eat less meat, and include plenty of whole grains. Obviously, this is completely contrary to the dietary principles we follow in the Infinite Health Program. We now know that advice to reduce fat and cholesterol intake was based on flawed data, including observational epidemiological data that cannot be used to generate cause-effect associations, as well as misinterpretations and even misrepresentations by people and industries with hidden dietary agendas. Yet the same errors are still applied to prevent cognitive decline.

Despite conventional low-fat advice, the bulk of studies examining cognitive health and diet demonstrate tha low-fat diets do not have a positive impact on slowing or preventing cognitive decline. But there are still many landmines and misinterpretations in dietary thinking applied to brain health. The numerous varied facets to diet also mean that many concepts are still evolving. Let’s try to distill some real answers out of this hodgepodge.

Nutritional thinking is flawed - and it's no different with brain health

---

Nutritional studies have traditionally and systematically been plagued by flawed logic and incorrect conclusions, leading to decades of poor dietary outcomes, obesity, type 2 diabetes, and over-reliance on prescription medications. One rule worth keeping in mind is the logical error made by the following:

Replace something unhealthy (e.g., white flour products) with something less unhealthy (e.g., whole grains). If there is an apparent health benefit, such as reduced cardiovascular disease, reduced colorectal cancer, reduced type 2 diabetes (which is true), we can NOT then conclude that lots of whole grains must therefore be good. (The effects of no grains should have been explored next. This has indeed been examined in numerous studies, though conducted under various labels such as low-carb, ketogenic, Atkins’, etc.)

For this reason, we cannot conclude: If a standard American diet that includes soft drinks, polyunsaturates, refined grains, etc. is replaced by, say, a Mediterranean diet, and there is an apparent health benefit, then a Mediterranean diet is therefore the ideal diet for health. Instead, the conclusion should be that the Mediterranean style of diet is less harmful, perhaps more beneficial, than a standard American diet, but it does not prove that it is therefore the ideal diet. You will find this logical fallacy used over and over again in nutritional thinking.

Low-fat diet for brain health?

---

Low-fat dietary advice remains the cornerstone of most conventional advice to prevent dementia. This is based on the suspicion and the observational studies demonstrating that higher levels of serum cholesterol contribute to dementia risk; any strategy to reduce serum cholesterol must therefore also reduce risk for dementia. For this reason, there was initial enthusiasm for the use of statin cholesterol drugs to reduce dementia risk.

Observational studies did indeed suggest a reduction in dementia and cognitive decline with statin drugs. Three prospective clinicals trials—Alzheimer’s Disease Cholesterol-Lowering Treatment trial (atorvastatin 80 mg vs. placebo; n = 67), the Lipitor’s Effect in Alzheimer’s Dementia (atorvastatin 80 mg vs placebo; n = 640) , and the Cholesterol-lowering Agent to Slow the Progression of Alzheimer’s Disease trial (simvastatin 40 mg vs. placebo; n = 406)—all failed to demonstrate any delay/improvement in cognition despite the preliminary observational findings. Summarized here:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4409455/
https://www.ncbi.nlm.nih.gov/pubmed/26727124

Most studies examining serum total cholesterol as a predictor of dementia also fail to factor in apo E4. Because apo E4 is clearly and consistently associated with higher total and LDL cholesterol values and dementia, it should come as no surprise that higher serum cholesterol is a predictor of dementia. Unfortunately, this association is overlooked in the majority of studies.

Other methodological limitations plague these studies, including failure to fractionate lipoproteins (thereby tracking flawed and imprecise measures such as total and Friedewald LDL), fail to track measures of oxidation or inflammation, fail to track tissue measures of fatty acid consumption/metabolism (e.g., RBC fatty acid composition).

The evidence supporting an association between saturated fat intake and cognitive decline are uneven and observational, with no prospective studies to confirm the association.

Observational studies

The Rotterdam Heart Study is often held up as evidence of the cognitive benefits of cutting total fat, saturated fat, and cholesterol, based on the initial report of increased dementia over the initial 2 years of observation. However, longer-term follow-up demonstrated that the association did not hold, and there was no association with dementia. Greater fat and saturated fat intake was associated with greater risk of dementia over 2 years in the Rotterdam Heart Study(n = 5386), lower risk with greater fish intake:

https://www.ncbi.nlm.nih.gov/pubmed/9392577

But longer term follow-up of the same Rotterdam Study cohort demonstrated that the association did not hold and there was no association of total fat, saturated fat, and cholesterol with dementia:

https://www.ncbi.nlm.nih.gov/pubmed/12499483

In the Kuopio Ischaemic Heart Disease Risk Factor Study, 2497 participants were observed over 22 years, with no reduction in cardiovascular risk or cognitive decline with egg and cholesterol consumption, an effect independent of apo E4 genotype:

https://www.ncbi.nlm.nih.gov/pubmed/28052883

The Cardiovascular Risk Factors, Aging and Dementia Study was a peculiar study in which intake of spreads and dairy was tabulated once at the start, then used to predict dementia over 21 years. Other forms of fat intake were not queried in detail. Saturated fat intake was minimally predictive but only in apo E4 participants.

https://www.ncbi.nlm.nih.gov/pubmed/16710090

The Women’s Health Study was an observational, diet questionnaire-based study (n = 6183). Higher saturated fat intake was associated with greater likelihood of cognitive decline, higher monounsaturated fat intake associated with lower likelihood over 4 years. However, higher fat intake was also associated with greater likelihood of smoking, drinking, less exercise, and lower income.

http://onlinelibrary.wiley.com/doi/10.1002/ana.23593/abstract

Conclusion

The bulk of evidence therefore indicates that reducing dietary fat is not an effective strategy for preventing cognitive decline, nor is reducing serum cholesterol with statin drugs, even at high dose. We also know that reducing fat is also a formula for creating problems in other spheres of health, such as weight gain and type 2 diabetes, not to mention provocation of hypertriglyceridemia, fatty liver via de novo lipogenesis, exaggerated postprandial lipoprotein abnormalities, and small LDL particle formation. We do not limit dietary fat intake in the Infinite Health lifestyle, nor is there any reason to believe that cholesterol reduction reduces risk for cognitive decline.

Mediterranean diet: Ideal diet for brain health?

---

Based on pockets of long-lived people with less cardiovascular disease, such as the people on the Greek island of Crete, the notion of a “Mediterranean diet” was born. Unfortunately, much of the interpretation of the details and potential benefits of this lifestyle originated with the observations of Dr. Ancel Keys, the very same Ancel Keys who manipulated data to publish his misleading Seven Countries Study that “proved” that saturated fat was harmful.

There are clearly aspects of the Mediterranean lifestyle that are beneficial such as consumption of locally-grown vegetables, olive oil, and fish (unless, of course, you have been convinced that TMAO is a legitimate marker of cardiovascular risk, below). But there are other aspects of Mediterranean life, such as daily bright sun exposure for vitamin D, a better sense of community and extended family, and travel by foot or bicycle that impact health beyond diet and confound the presumptive diet-cognitive health associations.

Observational studies

The Rush University MIND trial: The diet used in this trial was a hybrid of the Mediterranean diet and DASH for blood pressure reduction. Participants with a higher MIND score had 53% less likelihood of dementia over 4.5 years (n = 923).

The MIND diet score has 15 dietary components including 10 “brain healthy” food groups (green leafy vegetables, other vegetables, nuts, berries, beans, whole grains, fish, poultry, olive oil and wine) and 5 “unhealthy” food groups (red meats, butter and stick margarine, cheese, pastries and sweets, and fried/fast food) Participants following diets consistent with DASH or a Mediterranean eating style also experienced a modest, less marked, reduction in risk for dementia. You can quickly see that this diet and analysis makes some fundamental errors in that foods are lumped together as brain healthy or not based on preconceived notions. The outcome was impressive but it is impossible to tell which components of diet were responsible. And, of course, subscribing to the diet as designed may provide a cognitive benefit but would likely be associated with other health problems because of the inclusion of grains. Nonetheless, one potential lesson to take from the MIND trial is that most of the components of the “brain healthy” regimen, such as vegetables, nuts, berries, and olive oil may indeed provide benefit over, say, margarine or sweets.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4532650/

No cognitive benefit to a Mediterranean style diet over 13 years in cognitively normal people (observational, dietary questionnaire):

http://ajcn.nutrition.org/content/97/2/369.long

Prospective clinical trials

Only one prospective clinical trial of the Mediterranean diet in cognitive health has been performed to date, Predimed. In this case, the comparator was a low-fat diet and the Mediterranean diet proved superior:

Predimed-Navarra Trial: A Mediterranean diet with one liter added olive oil per week and 30 grams nuts per day improved cognitive measures (MMSE and clock drawing) over 6.5 years compared to a low-fat diet (n = 447):

https://www.ncbi.nlm.nih.gov/pubmed/23732551
https://www.ncbi.nlm.nih.gov/pubmed/23670794

Conclusion

Good data generated via the prospective Predimed trail demonstrate the superiority of Mediterranea eating style, augmented by higher olive oil/monounsaturated oleic acid/polyphenol intake and nuts, is superior to a low-fat diet. But, as discussed above, we cannot conclude from this that a Mediterranean diet is therefore the ideal diet to prevent cognitive decline.

Specific foods/nutrients

---

Note that nearly all the studies are observational and therefore non-conclusive. Nonetheless, they can be used to generate hypotheses. Unfortunately, this does not prevent study authors from proclaiming that associations are cause-effect.

Greater vegetable, but not fruit, intake has been associated with less cognitive decline. Six or more servings per week of green vegetables was associated with the greatest effect:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3393520/
https://www.ncbi.nlm.nih.gov/pubmed/15852398

Berry consumption, as well as overall flavonoid intake (anthocyanidins), was associated with less cognitive decline in the Nurses’ Health Study (n = 16,010):

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3582325/

Greater fish consumption was associated with less dementia over 4 years (n = 815):

https://www.ncbi.nlm.nih.gov/pubmed/12873849

Combined observational data suggesting reduced cognitive decline with 4 servings or more of fish per week compared to 1 serving:

https://www.ncbi.nlm.nih.gov/pubmed/29053784

Cocoa flavonoids not only increase hippocampal blood flow, but also increase dentate gyrus function and reduce reaction time but not memory (prospective, randomized):

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4940121/#SD1

Selenium 280 mcg obtained through one Brazil nut per day prospectively improved cognitive measures in people with mild cognitive impairment over 6 months:

https://www.ncbi.nlm.nih.gov/pubmed/25567069

An observational study in which dietary acrylamide was associated with increased dementia over 4 years in men, but not women:

https://www.ncbi.nlm.nih.gov/pubmed/28743904

An observational study suggesting increased dementia with grain consumption, decreased with dairy in a Japanese cohort:

https://www.ncbi.nlm.nih.gov/pubmed/29251743

ARIC observational data suggesting greater cognitive decline with milk consumption:

https://www.ncbi.nlm.nih.gov/pubmed/29039795

Polyphenol components of extra-virgin olive oil, such as oleocanthal, have beneficial effects in experimental preparations relevant to the neuropathology of dementia:

https://www.ncbi.nlm.nih.gov/pubmed/23414128

The Maine Syracuse Longitudinal Study observational data over 18 years

“Higher Wechsler Adult Intelligence Scale (WAIS) scores at baseline were prospectively associated with higher intakes of vegetables, meats, nuts and legumes, and fish, but inversely associated with consumption of total grains and carbonated soft drinks. After adjustment for sample selection, socioeconomic indicators, lifestyle factors (smoking and physical activity), and cardiovascular risk factors, the relations between higher cognitive performance and greater consumption of vegetables, meat, and fish, and lower consumption of grains remained significant.” The issues that plague the MIND trial also apply here, as foods are semi-arbitrarily lumped together, a byproduct of the observational design.

https://www.ncbi.nlm.nih.gov/pubmed/26878011

TMAO

Recall the headlines from Dr. Stan Hazen’s lab at Cleveland Clinic announcing that meat and fish consumption, rich in choline and carnitine, yields higher serum levels of trimethyl-N-amino oxide, or TMAO, that is associated with increased cardiovascular risk. Limited data also demonstrated that a vegetarian eating style was associated with lower TMAO serum levels. The media jumped on this, declaring that this confirms that red meat and fish are atherogenic and cause heart disease.

Problem: Bowel flora was not factored in, dismissed as a black-box effect. More recent data have demonstrated that TMAO serum levels are dependent on bowel flora composition (though the clinical correlates of dysbiosis and SIBO remain uncharted).

My conclusion:

TMAO is a epiphenomenon in dysbiosis/SIBO that converts the native human diet to a potentially unhealthy situation, but it does not necessarily mean that choline/carnitine are atherogenic, inflammatory, or contribute to cognitive decline. It means that media pronouncements that meat and fish cause heart disease or dementia are unfounded and premature, based on insufficient microbiome assessment:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5390365/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5127123/
https://www.ncbi.nlm.nih.gov/pubmed/27585440

Blood sugar effects

---

This is one among several observational studies demonstrating an association between higher blood sugar/HbA1c and cognitive decline over time (up to 16 years).

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4447796/

Glycemic index

Higher glycemic index foods (carbs, sugars) were associated with greater beta-amyloid accumulation by PET and cognitive decline:

https://www.ncbi.nlm.nih.gov/pubmed/29070566

Calorie intake

This cross-sectional observational study suggests that higher calorie intake is associated with increased risk for mild cognitive impairment, 5-fold greater risk in people who are apo E4+. But is it higher calorie intake per se or some aspect of diet that is causal, e.g., greater sugar/soft drink intake, greater grain intake and thereby exposure to gliadin protein-derived opioid peptides, or some genetic susceptibility to these effects that is associated with dementia risk?

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3679320/

A useful summary of much of the nutritional evidence from Canevelli et al 2016 at:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4808873/

Low-carb and ketogenic diets

---

For years, we have been warned that carbohydrate-restricted diets are dangerous because of an (imaginary, never documented) obligatory need for carbohydrates by brain tissue. For this reason, studies have looked at whether low-carb diets were inferior to low-fat diets with respect to cognitive performance:

Low-fat vs. very low-carb diet in type 2 diabetics (n = 115) had no effect on cognition over 2 years:

https://www.ncbi.nlm.nih.gov/pubmed/27876096

Low-carb vs. high-carb diets were associated with no difference in cognitive performance in obese participants (n = 47):

https://www.ncbi.nlm.nih.gov/pubmed/24061557

Conclusion

In prospective clinical studies, carbohydrate limitation does not impair cognitive ability.

Taking a low-carb diet even further, emerging data are looking at the advantages of ketogenic diets, especially since serum beta-hydroxybutyrate is proving to be an alternative energy source for the brain, perhaps of heightened importance when insulin resistance is present and glucose availability is impaired. Experimental evidence suggests amyloid-beta and tau pathology reduction, as well as protection against oxidative stress and mitochondrial dysfunction, summarized here:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4101992/

There are two small short-term (6-12 weeks) studies examining the effects of ketogenic diets on cognitive function:

Clinical studies

Very low-carb diet (<20 grams per day) “ketogenic” (though no serum or urine ketones were monitored) vs. low-fat/high-carb group (n = 23) demonstrated improved cognitive measures over 6 weeks with the ketogenic diet:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3116949/

A small clinical trial of a ketogenic diet (n = 10) over 3 months conducted at the University of Kansas. Among the findings:

1) The degree of ketosis achieved was modest with beta-hydroxybutyrate serum levels of only 0.5 mmol/L, a level easily exceeded by any of us trying to achieve ketosis. (A beta-hydroxybutyrate level as high as 0.4 mmol/L can be achieved via MCT oil alone.) This was achieved via carb limitation and MCT oil supplementation.

2) Participants showed improved psychometric measures including a 4.1 point improvement in ADAS-cog and 0.8 point improvement in MMSE score.

3) The 1.5-3.0 tablespoons of MCT oil per day yielded diarrhea in the majority of participants.

http://www.trci.alzdem.com/article/S2352-8737(17)30070-7/pdf

There are not yet any clinical data that document brain/hippocampal changes by volumetric imaging, nor long-term data.

This last study also reinforces the notion that increased serum levels of beta-hydroxybutyrate, whether achieved physiologically or taken exogenously, yields at least a nootropic effect. Ketogenic diets are, by necessity, high-fat (not high-protein, as increased protein intake negates ketosis). Yes, high-fat intake is inflammatory, but only in the setting of dysbiosis/increased intestinal permeability/lack of Lactobacillus and Bifidobacteria, factors we address aggressively.

Additional thoughts on ketosis and ketones

Be aware that many conventional thinkers still believe that saturated fat/high fat intake is the cause for dementia and that, especially in apo E4(+) people, a low-fat diet is desired to reduce cardiovascular risk (which it most definitely does not, typically yielding explosive quantities of small LDL particles, high triglycerides, low HDL, higher blood sugars, and insulin resistance. Insulin resistance is, of course, one of the fundamental etiologies of cognitive decline).

Glucose hypometabolism is indeed present in various brain regions in cognitive decline:

https://www.ncbi.nlm.nih.gov/pubmed/28709938

But ketone metabolism remains normal in cortical tissue:

https://www.ncbi.nlm.nih.gov/pubmed/26766547

MCTs 20 grams yielded cognitive benefits in non-demented elderly adults over 3 hours of consumption of a single meal in a small study (n = 19)

https://www.ncbi.nlm.nih.gov/pubmed/27568199

A proprietary study (Accera) demonstrating improved cognition in apo E4(-) but not apo E4(+) subjects with 10 grams twice-per-day of an MCT oil preparation yielding serum betahydroxybutyrate levels as high as 0.4 mmol/L.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2731764/

A case study of early dementia in an apo E4 individual responding initially to high-dose MCTs (165 ml of a 4:3 mixture of MCTs and coconut oil, divided into 3-4 doses per day) which yielded improved cognitive function initially with betahydroxybutyrate serum levels of up to 0.4 mmol/L. Deterioration over time prompted addition of a ketone monoester 28.7 grams per day (not the ketone salts commercially available to us) that yielded further improvement. Betahydroxybutyrate serum levels as high as 7 mmol/L were obtained with peak levels lasting 3-4 hours.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4300286/

Caffeine may add to the ketogenic effects of MCTs/exogenous ketones:

https://www.ncbi.nlm.nih.gov/pubmed/28177691

(I have personally been using an MCT oil powder in my coffee every morning, as it is an excellent coffee creamer with perfect mouthfeel. It yields 3-4 hours of mental focus and clarity. Introducing MCTs as a coffee creamer (or oil added to various foods) may be an easy way to insert MCTs into the daily habit while potentially obtaining the synergies between the modest betahydroxybutyrate serum rise of MCTs with that of caffeine.

Note the much higher doses of MCTs discussed, e.g., 20-30 grams per dose, not the 1-3 grams quoted in ReCODE.

Are there factors in coconut and coconut oil besides the 15% MCTs that potentially benefit health and cognition?

https://www.ncbi.nlm.nih.gov/pubmed/25997382

Potential problems, however

1) A ketogenic diet is difficult to maintain, particularly as preference for sweet foods increases as cognitive decline progresses,

2) Long-term ketogenic diets, as practiced by children who remain ketogenic for years to suppress intractable seizures, have 10-100 fold greater risk for calcium oxalate and urate kidney stones (increased serum uric acid; dysbiosis from failure to cultivate bowel flora, as ketogenic diets mistakenly leave out all prebiotic fiber sources). Children on ketogenic diets also experienced stunted growth, typically falling in the 10th percentile for growth, suggesting there may be something fundamentally wrong with prolonged, sustained ketosis. There are also sporadic reports of selenium deficiency, cardiomyopathy, and sudden cardiac death. Some of the latter complications may be at least in part due to the use of a corn oil-based oil supplement that was used to increase fat consumption in ketogenic kids.

3) Some of the ketone salts available to us commercially are toxic with potential for hyperkalemia and toxic intakes of calcium. I shall review the available ketone salts and future direction of ketone supplements in future. Should we choose to use a ketone salt, we shall have to choose carefully and not dose current products more than twice per day to avoid mineral excess. I suspect that the real answer will come with availability of the D-enantiomer without the biologically inert L-form that will be more effective at much lower doses, thereby reducing potential for the toxic potassium/calcium issue.

4) Dysbiosis and small intestinal bacterial overgrowth are virtually assured--As a diet essentially free of carbs is also free of prebiotic fibers. This means that constipation, metabolic distortions (higher blood sugar and insulin, higher blood pressure, higher triglycerides), diverticular disease, and risk for colon cancer, as well as mental/emotional/sleep consequences develop unless we purposefully and consistently provide prebiotic fibers, as well as address SIBO.

Conclusion

Preliminary human clinical studies demonstrate modestly improved cognitive measures over 6-12 weeks of ketosis achieved either physiologically and/or through increased serum beta-hydroxybutyrate levels. While experimental data suggest that dementia pathology may be reduced, no “hard” correlates of improved brain health have yet been demonstrated in humans. However, given that ketosis is the only dietary approach that has yielded measurable cognitive benefits and has, in other studies, been shown to reduce insulin resistance and endogenous glycation, it seems reasonable to make a very low-carb diet with intermittent ketosis our theme for preservation of cognitive health. Because 24-7 ketosis is not advisable, intermittent ketosis augmented with MCTs and perhaps exogenous ketones seems a reasonable compromise, coupled with efforts to maintain healthy bowel flora via prebiotic fiber intake.

Important

Related to nutrition, of course, is the status of bowel flora that may have even further implications for cognitive health. This will be considered in detail separately.

---

Goto Forum discussion.