How Not to Diet: The Groundbreaking Science of Healthy, Permanent Weight Loss

Overall score

Scientific accuracy

Reference accuracy

Healthfulness

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

Book published in 2019

Published by Flatiron Books

First Edition, Hardcover

Review posted February 1, 2024

Primary reviewer: Morgan Pfiffner

Peer reviewer: Hilary Bethancourt and Brooke Nissim-Sabat

Table of contents

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

Claim 1

An anti-inflammatory diet helps with weight loss

Supporting quote(s) and page number(s)

Page 101: “Pro-inflammatory diets are also associated with obesity, especially abdominal obesity”

Page 102: “But there is a way inflammation seems to play a cause-and-effect role in the obesity epidemic: inflammation in our brains”

Page 107: “By choosing to eat more anti-inflammatory foods and fewer pro-inflammatory foods, we may be able to both prevent and treat the damage to the appetite-regulating apparatus in our brains that can lead to—and sustain—obesity”

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 it is not supported by current evidence. The book makes the case that inflammation promotes weight gain by impairing the function of the hypothalamus. This certainly could be true — it’s well-established that the hypothalamus plays a central role in regulating body weight, and dysfunction of the hypothalamus caused by overt damage (from a brain tumor, for example) can cause obesity. Furthermore, studies on animals regularly find that hypothalamic inflammation is associated with (and often precedes) the weight gain accompanying unhealthy diets. However, there is little evidence supporting the suggestion that a certain way of eating can support weight loss by reducing hypothalamic inflammation.

When it comes to how diet impacts inflammation, the author writes “It’s easy to tell if a food is pro-inflammatory or anti-inflammatory: Feed it to people, and see what happens to their levels of C-reactive protein and other markers of inflammation” (page 100). However, these inflammatory markers don’t actually appear to predict weight gain. Although markers of inflammation are associated with obesity, one prospective cohort study conducted in Finland, no major marker of inflammation predicted weight gain or obesity development during 7 years of follow-up. This suggests that inflammation might be a result rather than a cause of obesity. 

Likewise, in a 2021 meta-analysis of 60 studies, changes in C-reactive protein (a commonly measured marker of inflammation) did not predict subsequent weight loss. Additionally, one Mendelian randomization study found that genetically increased levels of various inflammatory markers and mediators had no effect on people’s BMI. Finally, in people with autoimmune diseases, medications that block either interleukin-6 or TNF-a (both of which promote inflammation) actually cause weight gain, rather than weight loss. Overall, this research does not support the claim an anti-inflammatory diet has promise for helping weight loss, and it suggests the link between diet, inflammation, and weight change is more complicated than presented in How Not to Diet.

There are almost no human studies looking at how changes in inflammation measured in the hypothalamus itself relate to weight changes. One observational study did find a correlation between hypothalamic inflammation and BMI in a group of 34 participants, but it did not establish that the inflammation caused people to gain weight (again, it’s possible that weight gain contributed to higher inflammation).

It’s also worth taking a look at whether the dietary recommendations made in How Not to Diet would actually impact inflammation. The book characterizes whole plant foods as generally being anti-inflammatory relative to animal-sourced foods and processed foods. Evidence appears mostly consistent with this — dietary patterns centered around minimally processed, whole foods with mostly plant-based foods (like the DASH and Mediterranean diets) tend to lower markers of systemic inflammation. However, we felt many specific claims about how food impacts inflammation were not well-supported evidence.

One example of this is fiber. How Not to Diet implies that fiber is “strongly anti-inflammatory” (page 109), but existing research does not provide consistent support for this suggestion. In a 2022 meta-analysis of randomized controlled trials, supplementing with an average of 11 grams of soluble fiber every day for at least 12 weeks led to weight loss but had no effect on the inflammatory marker CRP. Specific types of fiber, of which there are many, might have anti-inflammatory effects (resistant starch appears promising in this regard), but it isn’t well supported to say that fiber in general is anti-inflammatory.

Full-fat dairy products, meanwhile, are implied to be inflammatory. At several points, How Not to Diet states that saturated fat is inflammatory. This claim is often preceded immediately by a comment about dairy foods being a major source of saturated fat, with the obvious conclusion being that high-fat dairy is therefore inflammatory. However, existing research indicates that, by and large, dairy foods aren’t inflammatory (with full-fat and low-fat dairy products not differing much in this regard).

How Not to Diet also refers to green tea (along with black tea) as “The most anti-inflammatory beverage” (page 108), but a 2022 meta-analysis of randomized controlled trials found green tea extract had no clear effect on the inflammatory marker CRP.

Some of these inconsistencies between claims and evidence may have been because How Not to Diet frequently cites something called the Dietary Inflammatory Index (DII), a model that purports to show whether and to what extent various foods and food components are inflammatory or anti-inflammatory. However, the methodology used to design the DII is far from perfect: it was partially based on petri dishstudies and animal experiments; it did not weight studies according to their sample sizes; and it did not weight effect sizes (in other words, small changes and large changes in inflammation were treated equally). Given these issues, the DII might not be a very reliable model for inferring the effect of specific foods on inflammation.

Importantly, CRP (which was often relied on to infer whether foods were inflammatory or anti-inflammatory) is a blood marker of inflammation that’s produced by the liver, and it’s not clear how much that tells us about inflammation in the brain. So the impact of diet on CRP and other blood markers of inflammation may or may not actually be relevant to the type of inflammation that could matter for body weight.

In summary, there is evidence that the overall whole foods, plant-rich dietary pattern recommended in How Not to Diet may help reduce inflammation. Yet many of the individual dietary recommendations for reducing inflammation (e.g., eliminating dairy foods, increasing fiber, or drinking green tea) are not well-supported by evidence. Importantly, while eating an anti-inflammatory diet may promote improved health and weight loss through various mechanisms (e.g., reduced intake of calorie-dense, hyperpalatable processed foods and increased intake of nutrient-rich high fiber foods (see Claim 3), there is still limited evidence that reducing inflammation is a mechanism through which such a diet might prevent weight gain or cause weight loss.

Claim 2

Certain diet-derived pollutants (“obesogens”) promote weight gain and obesity

Supporting quote(s) and page number(s)

Page 108–109: “The notion that we are being exposed to obesogenic pollutants—that is, obesity-generating chemicals—went from mere speculation in an alternative medicine journal in 2002 to strong scientific plausibility within a decade”

Page 109: “One clue that pollutants may be playing some role (in the obesity epidemic) is that our pets are also getting fatter”

Page 111: “DDT, the banned pesticide now found mostly in meat, particularly fish, is a “presumed human obesogen” responsible for perhaps thousands of annual childhood obesity cases. Meanwhile, the number attributable to the plastics chemical bisphenol A (BPA) may be in the tens of thousands”

Page 113: “Phthalates are another class of plastics compounds associated with weight gain. A European consensus panel of obesogen experts have a 40–69 percent probability of phthalate exposure causing more than fifty thousand cases of obesity annually in older women”

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 it is poorly supported by current evidence. While we shouldn’t rule out the possibility that certain chemical contaminants in foods might promote weight gain in some contexts, evidence is generally lacking that the chemicals cited in How Not to Diet promote weight gain in the amounts found in food.

The book discusses at least six examples of food-derived “obesogens”, a term for chemicals that allegedly promote obesity. The first example is organotins, a class of tin-based chemicals How Not to Diet says are found in seafood, especially “Halibut, swordfish, and canned tuna” (page 111). While multiple studies have indeed confirmed that seafood can be a major source of organotins, whether this actually matters for obesity is another story. The only human study linking organotins to weight gain involved 110 newborn boys, finding that the infants born to mothers with higher placental levels of TBT (a major organotin) gained slightly more weight in the first 3 months after birth. However, this did not translate into a difference in ponderal index (a measure of weight relative to height) when infants were measured 18 months after birth.

Furthermore, available evidence suggests eating fish doesn’t cause weight gain and in fact may even promote weight loss, contrary to what would be expected if the organotins in seafood had a substantial impact on weight gain. In a network meta-analysis of randomized controlled trials, it was found that eating fish resulted in an average weight loss of 1.85 pounds (0.84 kg). And while this is a very small difference in the grand scheme of things, no other food group (including whole grains, nuts, legumes, or vegetables) resulted in more weight loss. That said, one large meta-analysis of cohort studies found a high seafood intake (>3 times per week) during pregnancy was associated with a higher risk of childhood obesity. So it’s possible the effect of fish intake differs when it comes to in utero exposure, but more research is needed.

Another obesogen cited in How Not to Diet is DDT, a pesticide now banned in the United States and restricted in many other regions. How Not to Diet claims DDT is found “mostly in meat, particularly fish” (page 111), which appears to be mostly true in many regions of the world. When it comes to DDT exposure and weight gain, much of the research involves infants and children. In the CHAMACOS cohort, boys (but not girls) born to mothers with higher serum levels of DDT were more likely to become obese during childhood. On the other hand, a similar study found no clear relationship between umbilical cord DDT levels and the risk of childhood obesity. Finally, while some studies have reported a correlation between prenatal exposure to DDE (a major DDT metabolite) and a higher BMI in children, one of the longest studies on the topic found no clear association between prenatal exposure to DDE and BMI either during childhood or by the ages of 20–22. 

As far as research in adults, one prospective cohort study of 90 people found that having higher blood levels of DDE, but not DDT, was associated with a higher BMI when participants were assessed 18 years later. However, given the small sample size of the study, it’s hardly definitive. Cross-sectional studies, meanwhile, have had decidedly mixed findings when it comes to the relationship between DDE or DDT and body weight. And as covered already, despite seafood being a potential source of DDT, eating fish doesn’t seem to promote weight gain.

How Not to Diet then focuses heavily on Bisphenol A (BPA), a chemical used to make certain types of plastic. According to How Not to Diet, “ninety percent of exposure” to BPA is from “our diets”, including canned and packaged food (page 112). One meta-analysis of observational studies found that higher levels of BPA (measured in urine) were associated with a higher risk of obesity, abdominal obesity, and overweight. Unfortunately, there are very few prospective studies on BPA, meaning it’s hard to say if BPA precedes the development of obesity, weakening the case for a cause-and-effect relationship between high BPA levels and fat gain. It’s possible higher BPA levels are simply an indicator of consuming more canned foods and drinks (like soda), since those are major sources of BPA. It seems fair to say existing research on BPA is concerning, but not definitive.

Next, How Not to Diet makes the case that phthalates, a group of chemicals utilized in plastic production, might be another obesogen. Phthalates have been studied for their potential to cause obesity, but current evidence (mostly lower-quality observational studies) appears inconsistent. And one meta-analysis of prospective cohort studies found no association between prenatal exposure to DEHP — one of the main diet-derived phthalates — and body fat levels in children.

How Not to Diet also suggests that polycyclic aromatic hydrocarbons (PAHs) are a possible culprit in weight gain, stating that, in addition to having carcinogenic properties, PAHs “may be obesogenic as well” (page 115). There is some evidence linking PAHs to increased body weight, with a 2023 meta-analysis finding that a couple of specific PAHs are associated with a higher risk of obesity. Still, this was based on a handful of cross-sectional studies, which are a weaker form of observational evidence than prospective studies. 

It’s also interesting to note that the two PAHs most clearly associated with obesity in that meta-analysis were naphthalene and phenanthrene, and the main routes of exposure to these chemicals seems to be airborne contamination and occupational exposure, rather than from the diet. In fact, one of the few studies in which PAHs were associated with obesity was looking at PAH exposure from ambient air, rather than food. More research is therefore needed to determine if diet-derived PAH exposure specifically is associated with obesity.

Finally, How Not to Diet suggests certain pesticides found in conventional agriculture may promote obesity, making organic plant foods preferable (page 118). The strongest evidence referenced to support this suggestion is a prospective cohort study from 2017, which found that people who reported eating more organic food were less likely to become overweight or obese over the next several years. A few other, albeit non-prospective studies, likewise found that eating organic food is associated with a lower risk of obesity. While this evidence is intriguing, it’s hard to be certain that eating less nonorganic food is actually responsible for the effect. After all, consumers who favor and can afford organic food may be able to eat higher quality diets overall and be more health conscious in general. 

How Not to Diet singles out the pesticide atrazine as a likely mediator of the effect of nonorganic food on obesity risk, and there is indeed some research in rodents and at least one study of industrially-exposed workers suggesting this chemical can promote obesity with high exposures. But whether atrazine is readily found in nonorganic food in large enough amounts to promote obesity is not well-supported by evidence For one thing, atrazine is almost always undetectable or in very low concentrations in food (the US and EU limit the amount of atrazine allowed in food). Furthermore, with the exception of industrially exposed workers, there appear to be no published studies investigating the association between atrazine exposure in humans and weight gain or obesity.

In addition, the human evidence suggesting that these chemicals could cause weight gain is mostly observational, meaning that it’s hard to know whether they’re causing the problem, or just innocent bystanders.

In summary, the existing research tends to provide only weak and inconclusive evidence that the chemicals and pesticides mentioned in the How Not to Diet are found in high enough concentrations in our food supply to be a major culprit in weight gain. While it would be nice to always eat organic produce and avoid exposure to these so-called obesogens when possible, it remains unclear that avoiding fish, all canned foods, and non-organic foods will affect weight loss. 

Claim 3

A diet high in fiber-rich foods can help with weight loss

Supporting quote(s) and page number(s)

Page 121: “In fact, telling people to increase their intakes of fiber-rich foods may actually be one of the single most effective pieces of advice for weight loss”

Page 122: “Fiber-rich foods require more chewing, slowing down eating rate, which itself can improve satiety”

Page 123: “…fiber, which never gets absorbed, can act as a carrier to dilute and even eliminate calories out the other end”

Page 124: “It’s not what you eat but what you absorb, so you can lose more weight on a high-fiber diet eating the exact same number of calories simply because some of those calories get trapped and never make it into your system. Those on a Standard American Diet lost about 5 percent of the calories they eat into their waste, but a higher-fiber diet can double that”

Page 124: “…if Americans just reached the minimum recommended fiber intake, that might decrease calorie absorption by more than one hundred calories a day”

Page 126: “Those with more fiber in their systems burned more calories—even in their sleep”

Page 129: “…eat fiber-rich foods, though, and our good gut flora take the fiber we eat and churn out molecules that calm our cravings”

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

4 out of 4

This claim received a score of 4, meaning that it is strongly supported by current evidence. Eating a diet high in fiber-rich foods does appear to help with weight loss, an effect attributable to the fiber itself and possibly to common characteristics of fiber-rich foods (e.g., lower calorie density). We did feel the size of the effect was overstated in How Not to Diet, just not to an extent that warranted a lower score for the claim.

A large number of trials have found that weight loss occurs when people take fiber supplements. As an example of this, in one meta-analysis of 27 randomized controlled trials that lasted at least 12 weeks, supplementing with soluble fiber (from things like inulin and flaxseed powder) led people to lose an average of 2.75 pounds. Certainly a small effect on its own, but it did appear that higher doses tended to result in more weight loss.

It’s not yet entirely clear in the research why eating more fiber can cause weight loss, but a few compelling explanations exist. For one thing, fiber inhibits the absorption of macronutrients (like fat and carbohydrates), resulting in fewer calories being absorbed. Fiber may also slow stomach emptying, potentially sending greater fullness signals to the brain, promoting satiety. And last (although perhaps least well-established), fermentable fibers can be converted into short-chain fatty acids (SCFAs) by gut microbes, and some of these SFCAs might increase satiety and weight loss (on the other hand, SCFAs provide calories, so their net effect is unclear).

How Not to Diet specifically recommends increasing fiber-rich foods (as opposed to just taking fiber supplements). This advice could result in other changes to the diet that promote weight loss independently of fiber. Most notably, efforts to increase intake of fiber-rich foods could lead to reduced consumption of ultraprocessed foods (UPFs), which are usually lower in fiber and more calorically-dense. UPFs are a subcategory of processed food typically defined as foods made mostly of substances derived from food (flour, sugar, oil, etc.) as well as additives (like flavorings), with minimal amounts of unprocessed food. A few examples of UPFs are soda, potato chips, and most commercially available candy and pastries, foods which have been linked to weight gain. Not all UPFs are created equal (tofu and store-bought whole grain bread are also UPFs), but many UPFs seem to promote weight gain and obesity, with one clinical trial finding this effect is not entirely explained by differences in fiber content.

Still, it’s hard to know how much adopting a high-fiber diet will necessarily lead to improved dietary quality, reductions in UPFs, and/or reductions in energy density.  It can be hard to predict how this advice would play out in reality.

In several US-based clinical trials, instructing participants to consume a completely plant-based (i.e. vegan) diet resulted in increased fiber intake and more weight loss than a Mediterranean diet, a healthy omnivorous diet, or their usual (not high-quality) diet. Yet because many other aspects of the diet also changed among those who ate the vegan diet (e.g., reduced fat intake), we don’t know what amount (if any) of weight loss can actually be attributed to the increased fiber intake (though it seems reasonable to think it contributed). Not all higher-fiber plant-based diets show such a weight loss advantage, however. In fact, two clinical trials found that the plant-based Ornish diet failed to produce greater weight loss than the low-carbohydrate Atkins diet, even though fiber intake was quite a bit higher with the Ornish diet. One explanation could be that the Atkins diet groups ate more protein, and protein seems to support weight loss.

Still, because studies of plant-based diets are unable to isolate the effects of fiber from the effects of other major dietary changes (e.g., excluding animal products), it is helpful to look at studies that focused more specifically on having people eat more fiber-rich foods without simultaneously eliminating animal products.One randomized controlled trial found that advice to increase fiber intake resulted in similar weight loss compared to advice to follow a healthy diet and achieve at least a 500 calorie deficit. The increased fiber diet group lost less weight after 12 months (–4.6 pounds vs. –6 pounds), but the difference wasn’t statistically significant. It seems noteworthy that advice to simply increase fiber intake led to nearly comparable weight loss as a more complex dietary overhaul and intentional calorie restriction.

Two other randomized controlled trials compared diets that were high in fiber or protein. In the first trial, a high fiber diet (achieved by eating beans at every meal and emphasizing whole grains, fruits, and vegetables in the diet) produced similar weight loss compared to a diet that was higher in protein (protein was around 28% of calories vs. 19% of calories with the fiber diet). In the second trial, a high fiber diet (39 grams of fiber per day, achieved by eating whole grains and, to a lesser extent, legumes) led to fat loss, but not as much as eating a higher protein diet (featuring 28% of calories from protein and 24 grams of fiber). These studies suggest increasing fiber can help with weight loss, but perhaps not any better than can increasing protein.

Legumes, including beans and lentils, are especially high in fiber compared to other foods. In a meta-analysis of randomized controlled trials published in 2016, it was found that eating legumes (usually about 1 serving per day) resulted in an average weight loss of 0.64 pounds when participants weren’t trying to lose weight and an average weight loss of 3.8 pounds when the diet was calorically restricted.

In summary, a higher intake of dietary fiber appears to at least help support weight loss. While the effect of increasing fiber may not be large, this could be said of most singular dietary recommendations, so it shouldn’t be considered a significant mark against the claim.

Overall (average) score for scientific accuracy

2 out of 4

Reference 1

Reference

Reference 1740. Campbell’s® low sodium soups. Ready to serve low sodium chicken broth. Campbells.com

Associated quote(s) and page number(s)

Page 192: “Even “low sodium” Campbell’s soup may exceed by more than 250 percent my one-to-one sodium per-calorie ratio recommendation I detailed in How not to Die.”

Criterion 2.1. Does the reference support the claim?

2 out of 4

This reference received a score of 2, indicating that it provides weak support for the claim. Unfortunately, the link provided is broken and resulted in a 404 error. Based on the reference title the website link was to Campbell’s ready to serve low sodium chicken broth, which, according to nutrition information we located online, does contain the high sodium-to-calorie ratio cited in the book. However, at 30 calories per serving, this is arguably more a function of chicken broth being very low in calories. 

This “high” sodium to calorie ratio also appears to be the exception, rather than the norm. We looked up the nutrition information for some Campbell’s low sodium soups, finding that the low sodium varieties of their tomato soup, chicken noodle soup, cream of mushroom soup, and vegetable soup did not come close to having 250% more sodium (per mg) per calorie, instead being about 55% lower, 15% lower, 75% lower, and 50% higher, respectively.

Reference 2

Reference

Reference 1193. Southgate and Durnin. Br J Nutr. 24(2)517–535. 1970

Associated quote(s) and page number(s)

Page 123: “Fiber doesn’t just trap sugars; it can act as a fat- and starch-blocker too”

Criterion 2.1. Does the reference support the claim?

3 out of 4

This reference received a score of 3, indicating that it offers moderate support for the claim. This reference is specifically applied to the claim that fiber can block the absorption of fat, with the reference being a 1970 publication describing experiments in which adding fiber (termed “unavailable carbohydrates” in the paper) via fruits, vegetables, and whole grain bread to the diet increased fecal excretion of ingested fat, although there was said to be some inconsistency (i.e., not every participant experienced an increase in fat excretion when eating more fiber).

Reference 3

Reference

Reference 666. Wharton and Serodio. Curr Cardiol Rep. 17(5):35. 2015

Associated quote(s) and page number(s)

Page 69: “To date, most weight-loss drugs, despite their initial approval, have been pulled from the market for unseen side effects that turned them into a public threat.”

Criterion 2.1. Does the reference support the claim?

3 out of 4

This reference received a score of 3, indicating that it offers moderate support for the claim. The provided reference states that “Since the 1930s, the majority of anti-obesity medications that were once approved have been withdrawn from the market due to adverse effects and threat to public health”. However, the paper does not actually provide any research or statistics to validate this claim. 

This statement could plausibly be true (at least, at the time it was written) — there have been a number of medications that were approved for the purposes of weight loss, only for this approval to be reversed due to apparent health risks. Notable examples include lorcaserin, sibutramine, rimonabant, and phentermine with fenfluramine (phen-fen). But this has not been the fate of all medications approved for weight loss — orlistat, naltrexone/bupropion, phentermine/topiramate, liraglutide, and semaglutide remain FDA approved (although semaglutide was approved after the book’s publication).

Reference 4

Reference

Reference 3617. Richard et al. Clin Toxicol (Phila). 56(1):15-24. 2018

Associated quote(s) and page number(s)

Page 409: “…cannabinoid hyperemesis syndrome, which is believed to be caused by long-term heavy use of high-potency cannabis.”

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 is a review covering cannabinoid hyperemesis syndrome (CHS), a rare but accepted medical condition involving episodes of nausea and vomiting. In addition to chronic use of high-potency cannabis, CHS can also be caused by synthetic cannabinoids.

Reference 5

Reference

Reference 1417.  Zhi et al. Clin Pharmacol Ther 56(1):82-85. 1994

Associated quote(s) and page number(s)

Page 153: “Orlistats inhibits the enzyme in your intestines that digests fat and can effectively block the absorption of up to 30 percent of the fat you eat.”

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 study found that, on average, people taking orlistat excreted 27.1% more ingested fat in their feces, with the highest dose of orlistat reaching the claimed 30% threshold.

Reference 6

Reference

Reference 121. Finch and Loehlin. Behav Genet. 28(2):101-106. 1998

Associated quote(s) and page number(s)

Page 21–22: “Since all the eggs in an infant daughter’s ovaries are already preformed before birth….”

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. 

Reference 7

Reference

Reference 13. Malik, Willet, and Hu. JAMA 316(6):583–584. 2016

Associated quote(s) and page number(s)

Page 7: “As the chair of Harvard’s nutrition department put it, this “central argument” from the industry is that the “overconsumption of calories from carrots would be no different from overconsumption of calories from soda.”

Criterion 2.1. Does the reference support the claim?

4 out of 4

This reference received a score of 4, indicating that it’s an accurate summation of statements made in an article coauthored by Frank Hu, the chair of Harvard’s Nutrition department, in discussing industry representatives’ opposition to including grams of added sugar on the food label.

Reference 8

Reference

Reference 303. Kant et al. Gut 60(7):893–901. 2011

Associated quote(s) and page number(s)

Page 39: “The rearrangement of anatomy involved in one of the most common surgeries—gastric bypass—is thought to increase bile acid exposure along the intestinal lining. This causes sustained pro-inflammatory changes even years after the procedure, which are thought to be responsible for the increased cancer risk.”

Criterion 2.1. Does the reference support the claim?

1 out of 4

This reference received a score of 1, indicating that it offers little support for the claim. Here How Not to Diet claims that people who undergo gastric bypass are at an increased risk of rectal cancer due to greater exposure to bile acids, which promote inflammation in the rectum.

The reference provided found that people who underwent gastric bypass experienced a delayed yet significant increase in macrophage migration inhibitory factor (MIF), a cytokine considered both proinflammatory and a potential cancer promoter. 

However, this study did not find that bile acids explain the increase in MIF seen in the study. Specifically, investigators performed an in vitro study, incubating colorectal cancer cells with a major bile acid called deoxycholic acid (DCA), ultimately reporting no effect of DCA on MIF levels.

More importantly, though, the central claim here is that gastric bypass actually increases the risk of rectal cancer, which is not well supported by available research; several large observational studies have found no apparent change in the risk of rectal cancer following bariatric surgery (How Not to Diet even cites a cohort study to support this premise, but the study found it was banding procedures, not gastric bypass, that were linked to a higher risk of colorectal cancer mortality).

Reference 9

Reference

Reference 1704. Ledikwe et al. Am J Clin Nutr 83(6):1362–1368. 2006

Associated quote(s) and page number(s)

Page 186: “On average, those who eat low-energy-density diets consume hundreds of fewer calories, yet they eat significantly more food, about three quarters of a pound more per day.”

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 study cited looked at 7,356 adults in the United States and found that people eating a low energy density diet (defined as <1.7 and <1.6 calories per gram of food for men and women, respectively) ate fewer calories and more food by weight compared to people eating high-energy density diets (defined as >2.1 and >2.0 calories per gram of food for men and women, respectively).

Reference 10

Reference

Reference 3852.  Rosinger et al. Am J Clin Nutr. 104(6):1554–1561. 2016

Associated quote(s) and page number(s)

Page 437: “Spot-checking urine from nearly ten thousand men and women, researchers found that nearly half the obese individuals were walking around underhydrated compared to fewer than one in three individuals who were normal weight or lighter.”

Criterion 2.1. Does the reference support the claim?

3 out of 4

This reference received a score of 3, indicating that it provides moderate support for the claim. How Not to Diet cites a study as part of the claim that people with obesity (a BMI ≥30) are more likely to be underhydrated than people with a normal body weight or below (a BMI <25), but there are a few small issues with this.

For one thing, the referenced study used urine osmolarity to assess hydration status. Urine osmolarity can be a reflection of hydration status, but it’s not an exact measure of this metric, since factors other than body water status can influence urine osmolarity (including protein intake, which the study did not adjust for).

Furthermore, there is some disagreement as to what urine osmolarity cutoff should be used as an indicator of underhydration. The referenced study used ≥831 mOsm/kg – [3.4 × (age − 20 y)], but one research group has proposed that this value should be as low as >500 mOsm/kg. Applying this latter cutoff to the referenced study would reduce (perhaps substantially) how many people were classified as underhydrated.

With this all in mind, this study does suggest that people with obesity are more likely to be less hydrated, even if the exact numbers of those who are “underhydrated” might vary if different methods were used for measuring and defining underhydration.

Overall (average) score for reference accuracy

3.2 out of 4

Summary of the health-related intervention promoted in the book

How Not to Diet recommends a diet based around whole plant food, like whole grains, legumes, nuts, fruits, and vegetables. The book also recommends avoiding or limiting animal foods, refined sugar, refined grains, salt, and added fats and oils. Beyond this, How Not to Diet contains numerous additional tips people can consider implementing, including intermittent fasting, eating with limited distractions, eating so food stays in the mouth longer, drinking cold water on an empty stomach, incorporating vinegar into the diet, eating spicy foods, and drinking green tea.

Condition targeted by the book, if applicable

Obesity and overweight.

Apparent target audience of the book

The target audience of How Not to Diet is anyone hoping to lose weight in a healthy manner.

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

4 out of 4

The intervention received a score of 4, as there is good evidence that the dietary recommendations can result in weight loss. A number of studies have tested the effect of plant-based diets similar to the one recommended in How Not to Diet, with this research generally finding that such diets have weight-reducing effects.

This was demonstrated in one meta-analysis of 11 randomized controlled trials, in which following a (typically low-fat) vegan diet led to a weight loss of 6 pounds when compared to a variety of control diets and 16.3 pounds when compared to not dieting. Notably, in all but one of the studies, the vegan diet group was not advised to limit their calorie intake, suggesting that vegan diets can help reduce weight without having to actively count calories. 

Observational studies also find that people following vegan diets tend to have lower BMIs, suggesting that even outside of an intervention context, vegan diets might be helpful for weight loss (though recognizing that people who choose vegan diets may also be implementing other lifestyle changes that help prevent weight gain).

There are likely a number of reasons why people often lose weight on vegan and plant-based diets, including (but not limited to):

– An increase in fiber consumption (covered in claim 3)
– A reduction in processed food consumption (briefly covered in claim 3)
– A reduction in energy density (covered in detail in our review of the book The Volumetrics Diet)

These factors likely lead to weight loss by producing spontaneous reductions in calorie intake (and to a lesser extent, calorie excretion, in the case of fiber).

Finally, it’s worth mentioning that a major factor when it comes to the ability of a diet to produce weight loss is adherence, and some may find the diet too restrictive to follow, making it less effective long-term. That said, it isn’t clear that — on average — this diet is necessarily more difficult to follow than other diets capable of producing weight loss to a similar degree (see relevant studies here, here, and here, although this is not always the case). 

Plus, How Not to Diet also contains a number of behavior-related suggestions that may help improve dietary adherence and weight loss, like changing contextual clues related to eating (e.g., removing certain foods from your house), using implementation intentions (in which you pre-plan actions to deal with future situations that make it challenging to stick to your diet), and practicing certain self-monitoring actions (e.g., weighing yourself regularly). 

Furthermore, it may not be necessary to implement all recommendations in this book to see weight loss benefits. Many of the diet strategies suggested in this book (e.g., increasing intake of whole, fiber-rich foods and eliminating sugar and certain highly processed foods) can be implemented while remaining on an omnivorous diet. And it appears quite possible to achieve clinically meaningful amounts of weight loss without eliminating all animal products from the diet (especially if the animal products are high in protein, as covered in claim 3).

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

4 out of 4

The intervention received a score of 4, indicating that it is likely to greatly improve general health. Compared to a typical Western eating pattern, the diet recommended in How Not to Diet seems to have a number of benefits. For one, prospective cohort studies have collectively found that following a vegan diet is associated with a lower risk of ischaemic heart disease. This evidence is observational, reducing confidence in a cause-effect relationship, but in clinical trials vegan or nearly vegan diets (often with an emphasis on minimally processed food) tend to lower LDL levels (a major risk factor for CVD), making a reduction in heart disease risk appear fairly plausible.

Vegan diets with an emphasis on whole plant foods also seem to have some benefits to blood sugar control. Randomized controlled trials indicate that such diets can improve insulin resistance in people with obesity and lower HbA1c in people with type 2 diabetes (effects likely caused — at least in part — by the weight loss occurring on such diets). Vegan diets have also been linked to a lower risk of cancer, although the evidence is somewhat limited.

The diet recommended in How Not to Diet may not be free of downsides, though. Studies tend to find that people on vegetarian or vegan diets have lower bone mineral density. Consistent with this, following a vegan diet is associated with a higher risk of breaking a bone. 

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

1 out of 4

The diet received a score of 1, indicating that it is likely somewhat nutritionally inadequate. How Not to Diet advocates for a plant-based diet and gives reasons to avoid nearly every animal food (nonfat dairy might be an exception, although it isn’t explicitly recommended either). As a result, people following the diet are likely to exclude all animal foods, which could result in an inadequate intake of some essential nutrients.

The main nutrient of concern is vitamin B12. People who follow vegetarian and vegan diets are at a high risk of vitamin B12 deficiency. This risk can be mitigated by obtaining sufficient vitamin B12 from supplements and/or foods fortified with vitamin B12. Thankfully, How Not to Diet mentions this as an issue, stating that “B12 supplements or B12 fortified foods are critically important for anyone adopting a plant-based diet” (page 273).

A few other nutrients seem more likely to be low on a vegan diet, including zinc, vitamin D, selenium, and iron (with iron inadequacy primarily a function of poorer bioavailability). The diet may also result in inadequate bodily levels of long-chain omega-3s, although this is controversial as the actual health consequences are difficult to pin down.

Overall (average) score for healthfulness

3 out of 4