Grain Brain: The Surprising Truth About Wheat, Carbs, and Sugar– Your Brain’s Silent Killers

Overall score

Scientific accuracy

Reference accuracy

Healthfulness

How hard would it be to apply the book's advice? Very difficult

Book published in 2013

Published by Little, Brown and Company

First Edition, Hardcover

Review posted May 9, 2018

Primary reviewer: Stephan Guyenet

Peer reviewer: Kevin Klatt

Table of contents

1. Introduction
2. Scientific accuracy
3. Reference accuracy
4. Healthfulness
5. Most unusual claim
6. Other
7. Conclusion
8. Updates

The claim that carbohydrate increases the risk of dementia relies on the argument that eating carbohydrate causes insulin resistance and diabetes, which in turn promotes dementia.  It’s true that insulin resistance, diabetes, and the elevated blood sugar levels that accompany them probably increase the risk of dementia, and Grain Brain accurately cites research supporting this.  However, despite the prevalence of this idea in popular writing, there is very little evidence that carbohydrate per se causes insulin resistance, and Grain Brain cites no evidence to support this claim.  In contrast to Grain Brain’s claim that low-fat diets cause insulin resistance and diabetes, a large scientific trial showed conclusively that a calorie-controlled low-fat high-carbohydrate diet, plus exercise, reduces the risk of developing diabetes by more than half.  There is currently no such evidence supporting the high-fat Grain Brain program, although I look forward to seeing the idea tested scientifically.  The evidence overall suggests that the proportion of fat and carbohydrate in the diet has little impact on insulin resistance independent of weight loss, except when eating a very-low-carbohydrate, very-high-fat diet similar to what Grain Brain recommends, at which point insulin resistance seems to actually increase (note: this study directly measured insulin resistance, rather than using indirect and often misleading measures of insulin resistance, such as HOMA-IR, that are more common in low-carbohydrate diet studies).

The question is complicated by the fact that very-low-carbohydrate diets cause weight loss, and weight loss increases insulin sensitivity.  So it’s likely that very-low-carbohydrate diets can indirectly increase insulin sensitivity, possibly including the Grain Brain diet.  But this isn’t the claim that Grain Brain makes.  What rescues the score somewhat is the fact that in one passage, Grain Brain refers to “overconsumption of hyper-processed foods filled with refined sugars” (page 28), implicating overconsumption in general, which is probably more closely linked with insulin resistance than carbohydrate intake per se.

The second claim is that high levels of blood cholesterol are healthy for the brain, including the so-called “bad” LDL cholesterol that causes cardiovascular disease.  This is not only probably incorrect, it’s also dangerous. Grain Brain cites evidence that high blood cholesterol late in life is associated with better brain health and greater longevity.  What it fails to consider is that dementia, like cardiovascular disease, is a condition that develops progressively over a long period of time.  If we want to understand what causes it, we have to look at what’s happening in the years before it’s diagnosed.  To get a clearer picture of this, I located a meta-analysis (study of studies) that examined the association between blood cholesterol levels and dementia in both midlife and late life.  Meta-analyses give us an overview of the whole research literature and help protect against cherry-picking. This study found that high blood cholesterol in midlife is strongly associated with the later development of Alzheimer’s disease, and high blood cholesterol in late life has no association with risk.  This shouldn’t be surprising.  Overwhelming evidence suggests that high levels of LDL cholesterol damage the blood vessels, leading to cardiovascular disease and inflammation (a fact that Grain Brain denies), and cardiovascular disease is probably harmful to the brain.

Furthermore, the fact that some studies report that people with high cholesterol late in life have lower dementia risk doesn’t necessarily mean that high cholesterol causes the lower risk of dementia.  Low cholesterol late in life can actually be a symptom of pre-existing illness such as cancer, meaning that higher cholesterol might simply be a marker of not being sick.  By encouraging people to stop worrying about high cholesterol and stop taking cholesterol-lowering drugs, Grain Brain may be increasing their risk of cardiovascular disease, and perhaps also their risk of dementia.

Since so far we’ve only considered observational evidence that doesn’t do a very good job of teasing out cause-effect relationships, I looked at two other sources of evidence.  First, a study that examined whether people with cholesterol-increasing genes also have a lower risk of Alzheimer’s disease.  This study reported that people with genetically high cholesterol are not protected from Alzheimer’s disease, nor is their risk increased.  This contrasts with the research described in the previous paragraph, but still undermines the claims in Grain Brain.  Second, I examined the impact of cholesterol-lowering statin drugs on dementia risk.  Two meta-analyses suggest that statins have no impact on dementia risk, either when administered before the onset of dementia, or after.  This research isn’t yet definitive, but the best evidence we currently have isn’t consistent with the claim in Grain Brain.

Let’s move on to claim 3, that gluten found in wheat and some other grains is a major cause of many brain conditions including dementia, depression, attention deficit hyperactivity disorder (ADHD), schizophrenia, anxiety, epilepsy, movement disorders, and migraine.  On page 32, the book states that “gluten sensitivity represents one of the greatest and most under-recognized health threat [sic] to humanity”. Grain Brain lists several references to support its extraordinary claims about the impact of gluten on brain diseases.  One key study, cited on page 52, is particularly telling.  It was published in 1999 in the prestigious medical journal The Lancet by Marios Hadjivassiliou and colleagues.  In the paper, the researchers compare rates of anti-gluten antibodies in circulation (supposedly a sign that a person’s immune system reacts to gluten—we’ll revisit this in a moment) in three groups of people: 1) people with brain illnesses whose cause hasn’t been identified despite extensive testing, 2) people who have been diagnosed a variety of known brain illnesses, and 3) people without brain illness.  They reported that group 1— the people with brain illness of unknown cause— had higher levels of anti-gluten antibodies in their blood than the other three groups. That’s it. They didn’t show that gluten had caused their conditions, and they didn’t show that removing gluten improved their conditions.

It gets worse.  Group 2— the group with known brain illnesses—had the lowest rate of gluten antibodies of the three groups.  What diseases were represented in this group? They included epilepsy, movement disorders, and dementia. In other words, this reference actually undermines the claim that there’s a link between gluten sensitivity and these three disorders, one of which is central to the book’s narrative.

Another problem with this study is that it relies in part on a blood marker called immunoglobulin G (IgG) to measure gluten sensitivity, and IgG isn’t an established marker of food sensitivity.  Grain Brain repeatedly recommends getting tested for gluten sensitivity using a test called the Cyrex Array 3, which also relies on IgG.

Other references cited by Grain Brain to support the impact of gluten on common brain disorders are similarly unconvincing.  Together, these studies suggest that there may be a small subset of people with specific types of brain illnesses whose disease responds to a gluten-free diet.  This is certainly interesting but does not support the sweeping claims in Grain Brain.  These studies don’t rule out the possibility that gluten could be involved in causing common brain disorders like dementia, they just don’t demonstrate that it is involved.

Claim 1

Summary of claim

Eating carbohydrate increases the risk of dementia.

Supporting quote(s) and page number(s)

Page 6: “I believe that the shift that has occurred over the past century—from high-fat, low-carb to today’s low-fat, high-carb diet, fundamentally consisting of grains and other damaging carbohydrates—is the origin of many of our modern scourges linked to the brain, including chronic headaches, insomnia, anxiety, depression, epilepsy, movement disorders, schizophrenia, attention deficit hyperactivity disorder (ADHD), and those senior moments that quite likely herald serious cognitive decline and full-blown, irreversible, untreatable, and incurable brain disease.”

Pages 28-29: “But when cells are constantly exposed to high levels of insulin as a result of persistent intake of glucose (much of which is caused by an overconsumption of hyper-processed foods filled with refined sugars that spike insulin levels beyond a healthy limit), our cells adapt by reducing the number of receptors on their surfaces to respond to insulin…  The pancreas then responds by pumping out more insulin. So higher levels of insulin become needed for blood sugar to go into the cells. This creates a cyclical problem that culminates in type 2 diabetes. People with diabetes have high blood sugar levels because their body cannot transport sugar into cells, where it can be stored safely for energy… Like a shard of glass, the toxic sugar inflicts a lot of damage, leading to blindness, infections, nerve damage, heart disease, and, yes, Alzheimer’s disease.”

Criterion 1.1. How well is the claim supported by current evidence?

2 out of 4

This claim received a score of 2, indicating that it’s weakly supported by current evidence.  Although there is little evidence to support Grain Brain’s claim that eating carbohydrate per se increases the risk of insulin resistance and therefore dementia, the fact that one passage on page 28 implicates overconsumption in general makes the claim more compelling.

Criterion 1.2. Are the references cited in the book to support the claim convincing?

1 out of 4

This claim received a score of 1, indicating that the references cited by Grain Brain to support this claim are not convincing.   This is because the book does not cite evidence that carbohydrate causes insulin resistance, and this claim is not supported by the research literature.

Criterion 1.3. How well does the strength of the claim line up with the strength of the evidence?

1 out of 4

This claim received a score of 1, indicating that the strength of the claim in Grain Brain is substantially overstated relative to the evidence.  This is because the evidence that carbohydrate causes dementia appears very weak, yet it’s presented in Grain Brain as very strong.  It narrowly avoided receiving a score of 0 because of the passage on page 28 implicating overconsumption in general.

Overall (average) score for claim 1

1.3 out of 4

Claim 2

Summary of claim

High levels of blood cholesterol are healthy for the brain.

Supporting quote(s) and page number(s)

Page 42: “…when cholesterol levels are low, the brain simply doesn’t work well; individuals with low cholesterol are at much greater risk for dementia and other neurological problems.”

Page 77: “…when cholesterol levels are low, the brain simply doesn’t work well, and individuals are at a significantly increased risk for neurological problems as a consequence.”

Page 93: “Our understanding of how cholesterol is critical for brain health has brought me and many others in my field to believe that stains—the blockbuster drugs prescribed to millions of Americans to lower cholesterol—may cause or exacerbate brain disease.”

Criterion 1.1. How well is the claim supported by current evidence?

0 out of 4

This claim received a score of 0, indicating that it’s opposed by current evidence.  This is because high LDL cholesterol damages the cardiovascular system, which is important for brain health.  High cholesterol in midlife is associated with a higher risk of dementia.  Also, genetically elevated cholesterol levels are not associated with lower Alzheimer’s disease risk, and cholesterol-lowering statin drugs do not appear to cause higher risk (according to current evidence, which is not definitive).  Grain Brain repeatedly claims that blood cholesterol is taken up by the brain, where it performs vital functions, but in fact the brain makes its own cholesterol rather than taking it up from the circulation.

Criterion 1.2. Are the references cited in the book to support the claim convincing?

2 out of 4

This claim received a score of 2, indicating that the references cited to support it in Grain Brain are only weakly convincing.  The book accurately cites several references suggesting that higher late-life cholesterol is associated with lower dementia risk.  These references don’t establish that high cholesterol actually causes lower dementia risk, which is what the book claims.  Furthermore, they don’t appear to represent the evidence as a whole.

Criterion 1.3. How well does the strength of the claim line up with the strength of the evidence?

0 out of 4

This claim received a score of 0, indicating that Grain Brain greatly overstates the strength of the evidence pertaining to it.  The evidence it cites is observational, meaning that it doesn’t provide convincing evidence that high cholesterol actually causes a reduction in dementia risk.  This is one of several possible explanations for the association, and not the most likely.

Overall (average) score for claim 2

0.7 out of 4

Claim 3

Summary of claim

Gluten is a major cause of many brain-related conditions including dementia, depression, attention deficit hyperactivity disorder, schizophrenia, anxiety, epilepsy, movement disorders, and migraine.

Supporting quote(s) and page number(s)

Page 32: “…all the latest science points to the bane of gluten in triggering not just dementia but epilepsy, headaches, depression, schizophrenia, ADHD, and even decreased libido…”

Page 67 displays a table titled “Signs of Gluten Sensitivity” that lists more than 28 disorders including depression, attention deficit hyperactivity disorder, schizophrenia, anxiety, and migraine.

Criterion 1.1. How well is the claim supported by current evidence?

1 out of 4

This claim received a score of 1, indicating that the current scientific literature offers very little evidence one way or the other.

Criterion 1.2. Are the references cited in the book to support the claim convincing?

0 out of 4

This claim received a score of 0, indicating that the evidence cited in Grain Brain partially undermines it.  One of the key studies cited in the book suggests that dementia, movement disorders, and epilepsy are not typically associated with gluten sensitivity.  Other references cited by Grain Brain to support the impact of gluten on common brain disorders are unconvincing because they don’t establish that it’s involved in more than a small proportion of cases of uncommon brain illnesses.

Criterion 1.3. How well does the strength of the claim line up with the strength of the evidence?

0 out of 4

This claim received a score of 0, indicating that the strength of the claim is greatly overstated relative to the evidence.  I was unable to find evidence supporting the claim that gluten causes common brain disorders such as dementia, yet Grain Brain makes this claim strongly.

Overall (average) score for claim 3

0.3 out of 4

Overall (average) score for scientific accuracy

0.8 out of 4

Reference accuracy deteriorated when Grain Brain attempted to support its more questionable claims.  On pages 96 and 97, the book argues that cholesterol-lowering statin drugs may increase the risk of Alzheimer’s disease.  To support this, it cites a webpage that contains a non-peer-reviewed article by a scientist who isn’t an Alzheimer’s disease researcher.  This webpage makes a theoretical argument via various mechanisms that low-fat diets and statin drugs might increase the risk of Alzheimer’s disease, yet it cites no evidence that low-fat diets or statins actually increase risk.  Current evidence suggests that natural variations in blood cholesterol levels due to genetics have no impact on dementia risk, and statin therapy also has no impact that has been detected in studies to date.  Further research may change this picture but that’s the most pertinent evidence we have right now.

Reference 1

Reference

Chapter 4, reference 19.  E.L. de Hollander et al.  Int J Epidemiol 41:805. 2012.

Associated quote(s) and page number(s)

Page 120: “…waist circumference is often a measurement of ‘health’, as it predicts future health challenges and mortality; the higher your waist circumference, the higher your risk for disease and death.”

Criterion 2.1. Does the reference support the claim?

4 out of 4

This reference received a score of 4, indicating that it offers strong support for the claim. The reference suggests that high waist circumference is associated with higher cancer, cardiovascular disease, and all-cause mortality risk.  It also reports higher risk at the lowest level of waist circumference but I didn’t judge this to be sufficient to remove a point, particularly since the overall argument being made in that paragraph of Grain Brain is reasonable.

Reference 2

Reference

Chapter 5, reference 10.  M.P. Mattson et al.  NeuroRX 1:111 2004.

Associated quote(s) and page number(s)

Page 134: “Further evidence supporting the role of calorie restriction in strengthening the brain and providing more resistance to degenerative disease comes from Dr. Mark P. Mattson at the National Institute on Aging, who reported, [quote from the citation]”

Criterion 2.1. Does the reference support the claim?

4 out of 4

This reference received a score of 4, indicating that it offers strong support for the claim.  The source contains the text quoted in Grain Brain and it is presented in the correct context.

Reference 3

Reference

Chapter 2, reference 11.  Marios Hadjivassiliou. Lancet Neurology 9:318. 2010.

Associated quote(s) and page number(s)

Page 56: Discusses the early history of doctors linking brain symptoms with celiac disease.

Criterion 2.1. Does the reference support the claim?

4 out of 4

This reference received a score of 4, indicating that it offers strong support for the claim.  The statements in Grain Brain are consistent with the reference and its citations, which discuss the early history of doctors linking brain conditions to celiac disease, a gut disorder caused by gluten.

Reference 4

Reference

Chapter 6, reference 51.  D. Ferraro and G. Di Trapani.  J Headache Pain 9:147. 2008.

Associated quote(s) and page number(s)

Page 173: “Topiramate, which is used to treat epilepsy, comes with awful side effects that would alarm any parent and be distressing to a child.  They include weight loss, anorexia, abdominal pain, difficulty concentrating, sedation, and paresthesia (the feeling of ‘pins and needles,’ or of a limb ‘falling asleep’).”

Criterion 2.1. Does the reference support the claim?

4 out of 4

This reference received a score of 4, indicating that it offers strong support for the claim.  The quote in Grain Brain is taken almost verbatim from the study’s abstract, and further reading in the full text suggests that it is well supported.  The side effects attributed to topiramate in Grain Brain occur and they are quite common.

Reference 5

Reference

Chapter 2, reference 20.  Davis. Wheat Belly.  Chapter 1, p. 17.

Associated quote(s) and page number(s)

Page 64: “What’s most interesting about these exorphins, and further confirms their impact on the brain, is that we know they can be stopped by opiate-blocking drugs like naloxone and naltrexone—the same drugs used to reverse the actions of opiate drugs such as heroin, morphine, and oxycodone.  Dr. William Davis describes this phenomenon well in his book Wheat Belly: [quote from Wheat Belly]”

Criterion 2.1. Does the reference support the claim?

1 out of 4

This reference receives a score of 1, indicating that it doesn’t convincingly support the claim.  Although Wheat Belly is cited accurately, the factual claim being made in that passage is not convincingly supported by Wheat Belly and is probably incorrect.  Grain Brain claims that partially digested proteins in wheat cross the gut, enter the circulation, and cross the blood-brain barrier in sufficiently high concentration to act on opioid receptors in the brain similarly to heroin or morphine.  To my knowledge, this has never been demonstrated in a live animal and is highly unlikely, given how impermeable the blood-brain barrier is to foreign proteins and how little of these proteins is likely to cross into circulation from the digestive tract.  The fact that opioid-blocking drugs like naloxone can induce withdrawal-like symptoms under certain conditions does not support the theory, since the brain produces its own opioids when we eat food (such as wheat-flour-containing goodies like cookies, cake, and pizza).  For example, naloxone can cause withdrawal-like effects in rats accustomed to drinking sugar water, even though sugar water doesn’t contain opioid-like proteins.  In other words, proteins from wheat don’t have to get into the brain to activate opioid pathways—local opioid release in the brain is something that happens naturally with many foods, whether or not they contain gluten.  That said, it is possible that the opioid-like proteins in gluten may have harmful effects locally in the gut, but this idea isn’t discussed in Grain Brain.

Reference 6

Reference

Epilogue, reference 1.  The World Health Organization (http://www.who.int/chp/chronic_disease_report/media/factsheet1.pdf).  This reference is a link to a PDF document from the WHO that is no longer available.  I wasn’t able to find it elsewhere in a quick Google search. I randomly selected a new reference: Chapter 8, reference 13.  N.T. Lautenschlager et al. JAMA 300:1027. 2008.

Associated quote(s) and page number(s)

Page 201: “In a recent report in the Journal of the American Medical Association, professor Nicola Lautenschlager of the university of Western Australia found that elderly individuals engaged in regular physical exercise for a twenty-four week period had an 1,800 percent improvement on measures of memory, language ability, attention, and other important cognitive functions compared to a control group.”

Criterion 2.1. Does the reference support the claim?

1 out of 4

This reference receives a score of 1, indicating that it doesn’t convincingly support the claim.  Although the cited study does suggest that exercise improves cognitive function in older adults at risk for Alzheimer’s disease, the paper describes it as a “modest improvement”, far from the “1,800 percent improvement” claimed in Grain Brain.  This would indicate an 18-fold increase in cognitive function, a feat that has never been remotely approached by any intervention.  It’s unclear where the 1,800 percent figure in Grain Brain comes from, but it doesn’t appear in the paper.

Reference 7

Reference

Chapter 10, reference 2.  This reference cites the egg industry website www.incredibleegg.org and a 2006 article in Science News.

Associated quote(s) and page number(s)

Page 228: “Scores of studies have confirmed the value of eggs, which are quite possibly the world’s most perfect food; the yolk is the most nutritious part”.  This quote is in the context of a passage suggesting that eggs don’t raise blood cholesterol and don’t harm cardiovascular health.

Criterion 2.1. Does the reference support the claim?

2 out of 4

This reference receives a score of 2, indicating that it offers weak support to the claim.  At the time I checked it (Feb 2018), the egg industry website cited in Grain Brain made vague health claims about eggs but didn’t cite studies that support the book’s claims.  The Science News article describes a study in which eating eggs raised cholesterol, both “bad” LDL and “good” HDL.  In passing, it also describes an observational study suggesting that one egg per day is not associated with higher cardiovascular disease risk.  Together, this doesn’t offer very much support for the claims made in Grain Brain, particularly that eggs are the “world’s most perfect food”.  I do think it’s reasonable to suggest that eggs can be part of a healthy diet but the claims in Grain Brain go beyond this.

Reference 8

Reference

Chapter 4, reference 1.  R.H. Lustig et al. Nature 482:27. 2012.  This reference is to the epigraph on page 103 and is not associated with specific claims made in Grain Brain.  I randomly selected a new reference: Chapter 3, reference 31.  This is an article by the computational biologist Stephanie Seneff, “APOE-4: The clue to why low-fat diet and statins may cause Alzheimer’s”.

Associated quote(s) and page number(s)

Page 96-97 make several arguments based on this reference that suggest statins may increase Alzheimer’s disease risk.

Criterion 2.1. Does the reference support the claim?

1 out of 4

This reference receives a score of 1, indicating that it does not convincingly support the claim.  The cited reference does make the claims repeated in Grain Brain, but they aren’t convincing.  The reference isn’t a scientific study but rather a non-peer-reviewed article by a scientist who doesn’t conduct primary research on Alzheimer’s disease.  The article argues that the biology of blood cholesterol and brain health suggest that statins and low-fat diets should increase Alzheimer’s disease risk.  Besides being unpersuasive from a biological standpoint, the arguments don’t cite evidence that statins and low-fat diets actually do increase Alzheimer’s disease risk.  Current evidence suggests that natural variations in blood cholesterol levels due to genetics have no impact on dementia risk, and statin therapy also has no impact that has been detected in studies to date.  Further research may change this picture but that’s the best evidence we have right now.

Reference 9

Reference

Chapter 2, reference 3. https://www.healthspringholistic.com/

Associated quote(s) and page number(s)

Page 51: “As many as one in four people are vulnerable to [celiac disease] due to genetics alone; people of northern European ancestry are particularly susceptible.  What’s more, people can carry genes that code for mild versions of gluten intolerance, giving rise to a wide spectrum of gluten sensitivity. Celiac disease doesn’t just harm the gut.  Once the genes for this disease are triggered, sensitivity to gluten is a lifelong condition that can affect the skin and mucous membranes, as well as cause blisters in the mouth.”

Criterion 2.1. Does the reference support the claim?

2 out of 4

This reference receives a score of 2, indicating that it offers weak support for the claim.  The claims in this part of Grain Brain appear to come from a blog post titled “Gluten: What it is and what you need to know”.  This blog post contains no references to support its claims and appears on the website of an organization called the Healthspring Holistic Center.  This organization describes itself as a “progressive holistic center with a unique approach to healthcare”, including “some of the latest healing, detoxifying, and wellness modalities”.  That said, some of the claims made in the blog post and repeated in Grain Brain appear accurate, which is why this reference didn’t receive a lower score.

Reference 10

Reference

Chapter 5, reference 11.  H.C. Hendrie et al. JAMA 285:739. 2001.

Associated quote(s) and page number(s)

Page 134: “Many people believe Alzheimer’s disease is something you “get” from your DNA, but this particular study told a different story.  It was shown that the incidence of Alzheimer’s disease among Nigerian immigrants living in the United States was increased compared to their relatives who remained in Nigeria.  Genetically, the Nigerians who moved to America were the same as their relatives who remained in Nigeria. All that changed was their environment—specifically, their caloric intake.  The research clearly focused on the detrimental effects that a higher caloric consumption has on brain health.”

Criterion 2.1. Does the reference support the claim?

2 out of 4

This reference receives a score of 2, indicating that it offers weak support for the claim.  The reference is an interesting study suggesting that African-Americans have a much higher risk of developing dementia than similarly-aged people in Nigeria.  In contrast to what Grain Brain states, the African-Americans in the study were not “Nigerian immigrants” who were genetically the same as “their relatives who remained in Nigeria”—they were a cross-section of elderly African-Americans of unspecified African ancestry living in the Indianapolis area.  Furthermore, the study doesn’t specifically pinpoint excess calorie intake as the cause of the difference in dementia risk. In fact, the words “calorie” and “energy” don’t appear in the article. That said, it seems reasonable to speculate that differences in calorie intake could contribute to the difference in dementia risk, which is why this reference didn’t receive a lower score.

Overall (average) score for reference accuracy

2.5 out of 4

Grain Brain appears to be targeted to a general audience, and it has a goal of “dramatically reducing your risk of debilitating brain disease in the future” (page 9), including dementia and a long list of other brain disorders.  It received an overall healthfulness score of 3.3, indicating that it may be slightly healthier than the typical diet and lifestyle (although this is highly uncertain). Below, I’ll explain why.The first criterion is whether the intervention (the Grain Brain program) is likely to improve the target condition (brain disorders) relative to typical diet and lifestyle patterns.  I assigned this a score of 3, indicating that the effects of the Grain Brain program on the conditions in question are probably slightly beneficial.  I wasn’t able to identify convincing evidence that the diet and lifestyle program described in Grain Brain would either decrease or increase the risk of developing the brain conditions listed in the book.  We currently know very little about the long-term impacts of very-low-carbohydrate, very-high-fat ketogenic diets on the human brain and body.  Contrary to what Grain Brain claims, this is probably not how most of our ancestors ate (more on this later) so we can’t rely on historical or evolutionary precedent to argue that this type of diet should be healthier for the brain.  That said, there is very limited evidence that a ketogenic diet may be helpful for people who have already developed Alzheimer’s disease or Parkinson’s disease, and exercise does seem to improve cognitive health in people at risk of dementia, so my best guess is that the Grain Brain program could be slightly beneficial on balance.  This remains highly uncertain, and neither scientific evidence nor historical precedent rule out the possibility that it could harm brain health in the long run.

The second criterion is whether the intervention is likely to improve general health relative to typical diet and lifestyle patterns.  I assigned this a score of 3, indicating that it’s likely to slightly improve general health. I didn’t assign this score because I’m confident it will slightly improve health, but rather because this is my best guess based on very limited evidence.  What little evidence we have on Grain Brain-like diets ranges from concerning to encouraging, and all the long-term data we have comes from animal models.  Depending on the study, in mice a ketogenic diet either lengthens lifespan or has no impact on it.  It tends to promote leanness in mice while also increasing fat deposition and inflammation in the liver.  In humans, it causes fat loss in most people who carry excess weight, which is a good sign.  And a 2018 study suggests it may be helpful for managing blood sugar levels and weight in people with diabetes.  Human studies suggest the diet appears safe for up to a year, but we simply don’t know what happens when a person eats such an extreme diet for multiple years and decades.  Due to its abundant vegetable content, regular physical activity, and the fact that it will probably help people shed excess weight, my best guess is that the Grain Brain program is healthier than the way most people eat and live.  But we won’t really know if that’s true until science catches up to the claims.

The third criterion is whether the diet provides an adequate intake of essential and nonessential nutrients for general health in the target audience.  I assigned this a score of 4, indicating that the diet appears to be more than nutritionally adequate. If the diet advice in Grain Brain is applied faithfully, it should provide sufficient amounts of all essential nutrients, plus a level of nonessential nutrients (e.g., colorful polyphenols and fiber) that is somewhat superior to typical diets.  If a person doesn’t keep up with the high levels of vegetable intake recommended in Grain Brain—which seems likely for many people—the diet could end up being substantially nutritionally inadequate.  This is partially because a high proportion of its calories come from added fats, which are invariably low in essential nutrients.  If this doesn’t sound right to you, I invite you to compare the per-calorie vitamin and mineral content of butter to unrefined foods like beans, vegetables, nuts, meats, eggs, and yogurt.

Summary of the health-related intervention promoted in the book

Very-high-fat, high-saturated-fat, high-cholesterol, very-low-carbohydrate ketogenic diet with abundant vegetables and frequent intake of berries.  Unlimited saturated fat and cholesterol. Also emphasizes regular physical activity, sufficient sleep, and specific supplements.

Condition target by the book, if applicable

Grain Brain targets many brain illnesses but focuses especially on dementias such as Alzheimer’s disease.

Apparent target audience of the book

General audience.

Criterion 3.1. Is the intervention likely to improve the target condition?

1 out of 4

The Grain Brain program receives a score of 1, indicating that it has positive and negative qualities but its overall effects on dementia risk are unknown.  I was unable to find compelling evidence, either in the book or elsewhere, that either supports or refutes the central claim that a very-low-carbohydrate ketogenic gluten-free diet is able to reduce the risk of developing dementia and other brain disorders in humans.  A study in mice suggests that a ketogenic diet may be able to protect cognitive function over time, and a second study suggests that it doesn’t protect cognitive function in a mouse model of Alzheimer’s disease.  That said, there is limited evidence that a ketogenic diet can be useful in some brain disorders once they’re established.  A randomized trial reported that a ketone-generating fat supplement improved cognition in people already suffering from Alzheimer’s disease.  A paper describes a doctor’s experience with seven patients suffering from Parkinson’s disease, suggesting that most of them improved on a ketogenic diet.  Grain Brain also cites a paper showing that exercise can modestly improve cognitive function in older adults at risk of dementia.  For most of the other conditions Grain Brain claims to be able to prevent or treat, I wasn’t able to identify relevant evidence that supports or refutes the book’s claims.  The program’s emphasis on ignoring or even increasing blood cholesterol undercuts its potential benefits, since poor cardiovascular health probably increases dementia risk.

Criterion 3.2. Is the intervention likely to improve general health in the target audience?

2 out of 4

The Grain Brain program receives a score of 2, indicating that it seems possible but highly uncertain that it can improve general health relative to how most people typically eat and live.  I’m not aware of any compelling evidence on the long-term effects of the ketogenic diet in humans. Trials suggest it’s probably safe to eat a ketogenic diet for up to a year.  Studies in rodents suggest that its effect on lifespan are either neutral or beneficial, which is reassuring.  At the same time, in rodents it causes fat deposition and inflammation in the liver, and rodents can’t tell us about the cardiovascular impacts of the diet because they aren’t naturally susceptible to coronary artery disease.  That said, the ketogenic diet does promote weight loss in most people who carry excess fat, and Grain Brain also recommends regular physical activity, sufficient sleep, and a high intake of foods that may be beneficial like vegetables and berries.  Although the long-term impact of the ketogenic diet on general health remains unknown, the overall Grain Brain program seems more likely to do good than harm.

Criterion 3.3. Does the diet portion of the intervention promote an adequate nutrient intake for general health in the target audience?

3 out of 4

The Grain Brain diet receives a score of 3, indicating that it should provide more than adequate levels of essential and nonessential nutrients.  I didn’t identify any obvious vitamin or mineral deficiencies in the Grain Brain diet.  Its emphasis on vegetables and berries should provide greater levels of nonessential nutrients like fiber and colorful polyphenols than what most people eat, as long as readers faithfully adhere to what’s described in the book.

Overall (average) score for healthfulness

2.0 out of 4