What Is the Best Diet for Brain Health? | Dr. Gil Carvalho

Full analysis

Dr. Gil Carvalho, a physician and research scientist who runs the YouTube channel Nutrition Made Simple, walks Being Patient founder Deborah Kan through what the evidence actually supports on diet and brain health. The bottom line: no single named diet wins — what matters is a consistent pattern of food choices, and most popular "brain supplements" have far weaker evidence than their promoters claim.

What was covered

• The brain-protective eating pattern. Carvalho described three consistent features across the research: low ultra-processed food intake; substantial fiber-rich foods (fruits, vegetables, whole grains, legumes); and healthy fats from nuts, seeds, fatty fish, fermented dairy, and extra virgin olive oil. He called this the traditional Mediterranean pattern but noted the Mediterranean, DASH, and MIND diets are "95% the same" — the differences are "almost hair splitting."

• Cooking oils and the seed-oil panic. Carvalho said fears about vegetable/seed oils are "not substantiated in research" — studies consistently show better health outcomes with plant-sourced oils. Extra virgin olive oil does appear to have modest advantages over refined olive oil (more polyphenols, better blood-pressure signal), but even regular refined oils "still seem to be beneficial." Deep frying repeatedly in reused oil (as in fast-food industrial fryers) is the real concern, not occasional home frying.

• Omega-3 supplements (EPA and DHA). DHA and EPA are structural components of neuron cell membranes and affect how brain cells communicate. Randomized trials show a cognitive signal — but primarily in people whose baseline omega-3 intake is low, which describes most people eating a standard Western diet. Carvalho said algae-based omega-3 supplements are a cleaner alternative to fish oil because algae is grown in tanks, not seawater.

• Creatine. Strong evidence supports creatine monohydrate for muscle growth and athletic performance (well-documented in trials up to five years). Kidney damage concerns are not supported for healthy individuals; those with existing chronic kidney disease should be cautious. Cognitive benefits are "not nearly as strong evidence-wise" but do appear in specific situations: sleep deprivation, high stress, and possibly older adults. Carvalho said he would not object to someone 60 or 70 trying creatine for memory gaps.

• MCT oil and ketones. Medium-chain triglycerides (MCTs) allow the liver to produce ketones, an alternative brain fuel to glucose. Carvalho called the cognitive benefit evidence "pretty unclear" and "not well established" — most trials are small with technical problems. Ketogenic approaches have an established role in drug-resistant childhood epilepsy; claims that they treat psychiatric illness or brain cancer are, in his words, "a million miles ahead of the data."

• Multivitamins — the COSMOS trial surprise. A large trial called COSMOS gave older adults a multivitamin-multimineral supplement and observed improvement in general cognitive measures including memory. Carvalho flagged two limitations: the trial cannot identify which of the 20-plus vitamins drove the effect, and it is unclear whether the benefit reflects correcting dietary deficiencies rather than adding something new. He called the result notable given that multivitamins have generally been dismissed by mainstream nutrition science.

Notable claims & predictions

• Carvalho on the Mediterranean vs. DASH vs. MIND diets: "When you look at those diets, they're 95% the same. The difference is really almost hair splitting." The Mediterranean has the most randomized-trial support; the others have less.

• Carvalho on seed-oil fear: "A lot of fear mongering over the so-called seed oils… The truth is that is not substantiated in research. In fact, we consistently see better health results for these vegetable-sourced oils."

• Carvalho on omega-3s and who benefits: "Individuals typically with low intake and low status of EPA and DHA… it's more common that we see in those populations that cognitive benefit of supplementing. Whereas when someone is already getting plenty from their diet, generally you see less of an effect."

• Carvalho on creatine and kidney safety: "For the vast majority of individuals, creatine has not shown to cause kidney damage… for a healthy individual with no existing kidney issues, in general, not a concern." He added the caution applies for up to five years — longer data do not yet exist.

• Carvalho on spotting supplement hype: "When there's one magic bullet, this nutrient solves your problem, and by the way, grab it in my online store… these things are almost certainly snake oil." He named misaligned financial incentives — scientists selling the supplements they promote — as the clearest warning sign.

• Carvalho on what he personally takes: An algae-derived omega-3 supplement and a Mediterranean-style diet. He does not mention taking NAD, ketone supplements, or MCT oil.

Fact check

Claim (Carvalho): The Mediterranean, DASH, and MIND diets are "95% the same." Broadly accurate as a characterization of shared principles. MIND was explicitly designed by researcher Martha Clare Morris as a hybrid of Mediterranean and DASH, with added emphasis on berries and leafy greens. The figure "95%" is illustrative, not a measured statistic — but the substantive point that they overlap heavily is well supported in the nutrition literature. No issue.

Claim (Carvalho): The COSMOS trial showed multivitamin-multimineral supplementation improved cognition in older adults. The COSMOS-Mind sub-study, published in Alzheimer's & Dementia (2022), did find statistically significant improvements in global cognition and memory over three years in older adults taking a daily multivitamin. Carvalho accurately noted the trial cannot isolate which micronutrient drove the effect, and that the mechanism (correcting deficiency vs. true supplementation benefit) is unresolved. The claim is accurate and appropriately caveated.

Claim (Carvalho): Creatine trials run "up to about five years." Consistent with the published literature on creatine safety studies. Accurate.

Claim (Carvalho): Farmed vs. wild-caught fish — "not entirely clear" which is better for health. This is a fair characterization of the current state of evidence. No large randomized trial has settled this; observational data are mixed. Accurate.

Claim (Carvalho): Ketogenic diets have "an established therapeutic role" for drug-resistant childhood epilepsy. Well established and accurate. The ketogenic diet has been a recognized treatment for refractory pediatric epilepsy since the 1990s and is endorsed by neurology guidelines.

Claim (Carvalho): AI tools (ChatGPT, Claude, Google AI) are "90 to 95%" accurate on nutrition topics. This is an opinion offered without citation or methodology, not a measured finding. AI accuracy on nutrition varies considerably by topic, and these tools are known to hallucinate citations and statistics. Treat this as an informal observation, not a tested reliability figure.

No claims rise to the level of clearly false or deliberately misleading. Carvalho consistently flagged uncertainty rather than overstating findings — a notable contrast to the influencer behavior he criticized.

Why this matters for you

• If you eat a standard Western diet and rarely eat fatty fish, omega-3 supplementation (fish oil or algae-based) has the best-supported cognitive case among the supplements discussed. Carvalho's own choice is algae-based for purity reasons — a practical distinction worth knowing if you are comparison-shopping.

• If you are 60 or older and noticing memory gaps, Carvalho said he would not object to trying creatine monohydrate — with the note that existing chronic kidney disease changes the calculation. Worth a conversation with your doctor before starting, especially if kidney function has ever been a concern.

• The COSMOS multivitamin finding is worth watching. Multivitamins have been largely dismissed by mainstream medicine for healthy adults. If your diet is inconsistent or your absorption has declined with age (common), a basic daily multivitamin may be doing more than previously thought. No action required now, but it is a reason not to throw out the bottle.

• The seed-oil panic and many supplement claims circulating on popular health podcasts are not supported by the research Carvalho reviewed. If you have been avoiding canola or other vegetable oils, or spending money on NAD or ketone supplements based on podcast recommendations, the evidence base for those choices is thin to nonexistent at this point. Interesting, but nothing urgent to act on this week — except possibly reviewing what you are spending on