The Obesity Code: Unlocking the Secrets of 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 2016

Published by Greystone Books

First Edition, Paperback

Review posted November 4, 2019

Primary reviewer: Seth Yoder

Peer reviewer: Stephan Guyenet

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

Reducing calorie intake does not lead to weight loss

Supporting quote(s) and page number(s)

Page 41: “The key assumption of the theory that reducing caloric intake leads to weight loss is false, since decreased caloric intake inevitably leads to decreased caloric expenditure. This sequence has been proven time and again.”

Page 39: “The body must, at some point, reduce its caloric expenditure to meet the lower caloric intake. It just so happens that this adaptation occurs almost immediately and persists long term.”

Page 29: “A 30 percent reduction in caloric intake results in a 30 percent decrease in caloric expenditure. The end result is minimal weight loss.”

Page 41: “The portion-control caloric-reduction diet is virtually guaranteed to fail. Eating less does not result in lasting weight loss.”

Page 43: “Caloric reduction as the primary means of weight loss has disappointed repeatedly.”

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

0 out of 4

This claim received a score of 0 out of 4, indicating that it is opposed by current evidence. One caveat to this score is that the book is somewhat inconsistent on this issue and it is difficult to pin down what it’s actually claiming. Despite confidently asserting that reducing calorie intake does not cause significant weight loss, at other times TOC appears to acknowledge that reducing calorie intake does just that. For example, on pages 36-37, the book discusses the Minnesota Starvation Experiment, in which a diet of 1,570 calories per day caused a group of volunteers to lose one quarter of their body weight over a six-month period.

The Minnesota Starvation Experiment was conducted in the 1940s, but countless other studies have confirmed its findings that reducing calorie intake causes weight and fat loss. The most rigorous of these are so-called “metabolic ward” studies, in which volunteers are housed in a research facility where researchers have absolute control over every morsel that crosses the volunteers’ lips. There are even metabolic ward studies that have compared high-carbohydrate reduced-calorie diets to low-carbohydrate reduced-calorie diets. These studies unanimously report that reducing calorie intake causes weight and fat loss, and that the degree of fat loss is proportional to the calorie deficit. They typically also report that the carbohydrate vs. fat content of the diet makes little difference when calorie intake is held constant, despite substantial differences in insulin secretion.

Nevertheless, these findings don’t rule out the possibility that reducing calorie intake is simply unsustainable. This isn’t quite what we think TOC is claiming (based on the quotes above), but one could argue that while calorie restriction does cause weight loss, most people can’t maintain it so it’s not effective in real life. This is a version of the argument that we believe has some validity. Reducing calorie intake does cause weight loss, and that weight loss can persist for several years if people receive intensive support, as demonstrated by the large, 2.8-year Diabetes Prevention Program trial. Participants in the lifestyle intervention group lost an average of 12.3 pounds through a combination of calorie restriction and exercise (importantly, the intervention also reduced their risk of developing diabetes by 58 percent).

Yet over longer follow-up periods, the amount of weight loss shrank, and after ten years it evaporated altogether. This could be due to the fact that the intervention officially ended years earlier, so the volunteers were no longer necessarily trying anymore. But it nevertheless shows that calorie restriction does not necessarily lead to durable weight loss over the very long term. However, it’s critical to note that no other weight loss strategy, including those recommended in TOC, has been shown to produce weight loss over such a long period of time.

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 references tend to undermine the claim. In most cases the references undermine the claim, while in others they simply do not support it.

On page 36, TOC cites a study published in 1919 to support its contention that calorie restriction does not cause weight loss. As part of this study, groups of men were given calorie-restricted diets and many outcomes were measured, including metabolic rate. TOC suggests that their metabolic rates declined to such a degree that it balanced out the reduction in calorie intake, implying that the reduction in calorie intake would not result in weight loss. However, TOC omits the most relevant finding of the study: the men did in fact lose weight. Of one group, the researchers wrote: “The men were all given a diet containing approximately 1,400 net calories. The resultant fall in body-weight is remarkably uniform…” (pg. 225).

TOC cites the Minnesota Starvation Experiment on pages 36-37, conducted in the 1940s. As previously described, in this study a group of men lost a quarter of their body weight eating a low-calorie diet for six months.

TOC cites the Women’s Health Initiative Dietary Modification Trial, which was a massive controlled study of the impact of diet changes on disease risk. The book states that “Daily calories dropped from 1788 to 1446 a day—a reduction of 342 calories per day for over seven years.” Yet according to TOC, weight loss was only four pounds at one year, and “by the end of the study, there was no significant difference between the two groups.” The study itself reports that weight loss at one year was five pounds (2.2 kg), and there was still a significant effect by the end of the study, although it was small (one pound). More importantly, self-reported calorie intake, as this study reported, is notoriously inaccurate. The authors themselves acknowledge this in the paper, stating “the [food frequency questionnaire] is not sufficiently reliable to allow evaluation of the potential effects of changes in caloric intake.” Well-validated models of human energy needs demonstrate that neither 1,788 nor 1,446 calories per day are plausible intake figures for this group of women, given their body weights and weight changes. The most likely explanation for the findings is not that reduced calorie intake failed to produce the expected weight loss, but that an inaccurate measure of calorie intake produced inaccurate results.

Another example of a reference undermining the claim can be found on page 42, on which TOC states:

“Caloric reduction as the primary means of weight loss has disappointed repeatedly. Reviews of the literature by the U.S. Department of Agriculture highlight this failure.”

In this case, a 2001 review of the relevant trials of broadly popular diets (such as vegetarian, low-fat, low-carb, USDA Food Guide Pyramid, etc.) was cited to support this. The report concludes:

“Review of the literature suggests that weight loss is independent of diet composition. Energy restriction is the key variable associated with weight reduction in the short term.”

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

0 out of 4

This item received a score of 0 out of 4, indicating that the strength of the claim is greatly overstated. The claim is refuted by a large volume of compelling evidence, including several studies cited in TOC itself, which we reviewed above.

Overall (average) score for claim 1

0.0 out of 4

Claim 2

Elevated levels of insulin are the primary cause of obesity.

Supporting quote(s) and page number(s)

Page 134: “In practical terms, this means that not only does insulin causes obesity, [sic] but also that insulin causes type 2 diabetes. The common root cause of both diseases is high, persistent insulin levels.”

Page 147: “Insulin causes adult obesity. Insulin causes newborn obesity. Insulin causes infant obesity. Insulin causes childhood obesity.”

Page 148: “The origins of both infant obesity and adult obesity are the same: insulin.”

Page 134: “That high insulin levels cause both obesity and type 2 diabetes has profound implications. The treatment for both is to lower insulin levels […]”

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

2 out of 4

This item received a score of 2, indicating that the claim is weakly supported by current evidence.

The idea put forward in TOC is that insulin is the primary driver of weight gain. A diet high in carbohydrates and with high eating frequency, the theory contends, will lead to high insulin levels resulting in weight gain. This model of obesity is similar to the carbohydrate-insulin model proposed by obesity researcher David Ludwig, MD, PhD. Within TOC the model is referred to as the “hormonal obesity theory.”

The hypothesis that elevated insulin levels are the primary cause of obesity does have advocates in the obesity research community, but it is a minority position that most of the community does not believe is well supported by evidence. We briefly review some of this evidence here, but we acknowledge that research on this is ongoing.

If we want to evaluate this hypothesis, the first question to ask is whether people with higher insulin levels gain more weight over time than people with lower insulin levels in the general population. If elevated insulin is the primary cause of obesity, we should see a clear and consistent pattern whereby people with higher insulin gain more weight over time than people with lower insulin. If we do not see a clear and consistent pattern, it’s hard to understand how elevated insulin could be the primary cause of common obesity.

Since researchers often measure insulin levels and body weight, there is a large body of evidence we can draw from. The only review paper we are aware of that has collected this evidence was published by Hivert and colleagues in 2007. Among 22 studies that measured insulin levels at baseline and observed weight change over time, there is no clear pattern suggesting that people with higher insulin gain more weight than people with lower insulin. Five studies found that higher insulin levels were associated with greater weight gain, eight found that higher insulin levels were associated with less weight gain, and nine found no association. Given these results, we find it hard to understand how elevated insulin could be the primary cause of obesity, although they don’t rule out a smaller role for insulin.

Overfeeding studies are another pertinent source of evidence. In these studies, researchers feed volunteers high-calorie diets and observe how much weight and fat they gain over time. If it is true that diets high in carbohydrate lead to an increase in circulating insulin and therefore an inevitable increase in fat storage and weight gain, then we would expect that overfeeding people with a high-carbohydrate (higher-insulin) diet would lead to more weight gain than overfeeding with the same number of calories of a high-fat (lower-insulin) diet. However, the only two tightly controlled studies that have done this both reported that high-carbohydrate and high-fat diets produce about the same amount of fat gain when calorie intake is the same.

While the carbohydrate-insulin model still has its proponents, it has been considerably weakened by scientific findings over the last few years, which have failed to find a fat loss advantage for low-carbohydrate diets when calorie intake is held constant under tightly controlled conditions.

That said, the insulin model is not completely lacking in scientific support, and TOC focuses on the supportive evidence. For example, beginning on page 78, the book discusses the observation that people with diabetes tend to gain weight when they initiate insulin therapy. This is certainly true, but requires context. Fat tissue does require a certain amount of insulin signal to function properly, and when insulin signaling is severely impaired (as in uncontrolled diabetes), it is unable to store fat effectively. At the same time, people with uncontrolled diabetes excrete up to several hundred calories per day via glucose in their urine. For these reasons, they tend to lose weight and fat. Injecting insulin corrects these defects, resulting in weight gain. Yet this scenario appears to have little relevance to people without diabetes, as demonstrated by the lack of correlation between insulin level and weight gain in the general population (discussed above).

TOC also walks the reader through a number of drugs that impact body weight (pages 82-85), arguing that they all change body weight in the same way: by altering insulin secretion. This one-factor explanation for the body weight effects of diverse drug classes is not persuasive overall, although it may be correct in some cases. For example, people with type 2 diabetes often lose weight due to an insufficient insulin signal resulting in dysfunctional fat tissue and blood sugar loss via the urine, so it makes sense that at least some drugs that restore insulin signaling in people with type 2 diabetes will promote weight gain. This may be the case with sulfonylureas, discussed in TOC.

However, we run into problems when we try to apply the same logic to other drugs. Weight gain itself increases insulin levels, so observing that insulin and body weight both increase when a person takes a drug– as TOC does– is not compelling evidence that the increase in insulin caused the weight gain. For example, TOC argues that the antipsychotic drug olanzapine increases insulin levels and causes weight gain, concluding that the former effect is responsible for the latter (pg. 84: “As insulin rises, so does weight”). Aside from the issue that correlation doesn’t necessarily imply causation, the evidence is not very consistent that olanzapine increases insulin secretion prior to weight gain. In rats, olanzapine actually reduces insulin secretion initially, and only after weight gain occurs does insulin increase. Furthermore, a recent study in rodents attributes much of the weight gain effect of this drug to its actions on serotonin receptors in the brain, which are known to impact body weight. Although current evidence is not definitive, it does not point in the direction of increased insulin secretion being the primary cause of weight gain in people who take olanzapine.

Liraglutide is one of the most effective weight loss drugs on the US market, and it actually increases meal-related insulin secretion (which is why it was originally developed as a diabetes drug). Its sister compound semaglutide also increases meal-related insulin secretion and is even more effective for weight loss; in a recent randomized controlled trial, people taking the highest dose lost a remarkable 14 percent of their body weight over a one-year period (33 lbs). It’s worth noting that 1) the weight loss effect of these drugs may have nothing to do with insulin secretion and is more likely mediated by the brain, and 2) since these drugs reduce food intake and body fat mass, they may eventually reduce insulin levels, just like other means of weight loss.

TOC argues that SGLT2 inhibitors, which cause people to lose several hundred calories per day of blood sugar through the urine, cause weight loss because they reduce insulin secretion. A more likely explanation is that forcing the body to excrete hundreds of calories per day causes weight loss, and insulin levels decline as a result. This is consistent with the observation that weight loss by calorie restriction, exercise, or surgery reduces insulin levels. Since the extra sugar is coming out of the bloodstream, the pancreas has to secrete less insulin to control blood sugar, and this probably also contributes to the reduction in insulin levels.

Overall, we think the evidence suggests that insulin secretion probably isn’t the primary determinant of body weight, but we acknowledge that research is ongoing and some of the arguments in TOC do support the claim. For these reasons, we judge the claim as “weakly supported by current evidence”.

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

2 out of 4

This item received a score of 2 indicating the references are weakly convincing. There are some reasonable references to support narrow parts of the claim, but when these references are further explored the mechanisms are not in line with the mechanisms proposed by TOC.

On pages 81-82, as further evidence of the hormonal theory of obesity a case study is mentioned of a woman with a tumor that would spontaneously release insulin leading to bouts of very low blood sugar:

“She had gained 25 pounds over the year prior to her diagnosis. Increased caloric intake did not account for the weight gain. Reduced caloric intake did not account for the weight loss. The defining element was insulin: its rise and fall corresponded to the rise and fall in weight.”

However, to say that increased caloric intake did not account for the weight gain is unsupported by the case study. Calorie intake was not measured and only briefly mentioned: the authors noted that when the woman would suffer these bouts of low blood sugar, the symptoms were relieved by food or juice. Very low blood sugar is known to promote hunger and weight gain, most commonly in people who inject too much insulin. This would suggest that caloric intake could have played a role. Furthermore, although clinically low blood sugar levels (hypoglycemia) can certainly cause hunger and weight gain, this is uncommon in the general population and so the relevance of this situation to common obesity is questionable. Moreover, case studies are considered among the weakest forms of evidence available.

There is a section in TOC titled Evidence from the Pima People that the “hormonal theory of obesity” helps to explain. On page 142 and 143 some claims are made about the Pima:

“The Pima Indians of the American southwest have the highest rates of diabetes and obesity in North America. An estimated 50 percent of Pima adults are obese, and of those, 95 percent have diabetes.”

The reference for these figures and others about the Pima are taken from a web article published on the NIDDK website. It doesn’t appear to be there anymore, but the Wayback machine has it archived. The figures on obesity and diabetes are reversed in the article, which states “one-half of adult Pima Indians have diabetes and 95% of those with diabetes are overweight.”

How did this obesity and diabetes come about in this population? The idea that it might be a transition to a sedentary lifestyle is dismissed in TOC because “Native American tribes developed obesity in the 1920s, decades before widespread use of cars” (page 143).The answer, according to TOC, is “Simple enough. The same thing drives obesity in the Pima Indians as in everyone else: highly refined carbohydrates” (143). But many other facts are omitted to provide this uncomplicated narrative. The article states that the Pima’s tradition of irrigation and agriculture was disrupted by American farmers so they had to rely on lard, sugar, and white flour from the government. According to the article, “The greatest changes have occurred in consumption of fat. In the 1890s, the traditional Pima Indian diet consisted of only about 15 percent fat and was high in starch and fiber, but currently almost 40 percent of the calories in the Pima diet is derived from fat.” There is also mention of genetic risk factors and the fact that the Pima in Arizona are heavier than another Pima group living in Mexico, which is much more physically active. It seems rather unlikely that refined carbohydrate is the only explanation for high rates of obesity and diabetes among the Pima, although it could have contributed.

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

1 out of 4

This item received a score of 1 indicating that the strength of the claim is substantially overstated. Although most of the evidence that is cited is from reputable sources, these sources do not provide robust evidence to support the claim. Rather, bits of data from case studies, cross-sectional analyses, and other less-compelling forms of evidence are used to draw conclusions about insulin and its role in obesity that are not well supported by the overall scientific literature.

One area of the book that is intended to provide support for the hormonal theory of obesity (a.k.a. the carbohydrate-insulin model) is on pages 147 and 148 in a section titled It’s Insulin. Two observational studies are mentioned in this section: one notes that maternal weight gain is associated with increased birthweight and the other notes that obesity in childhood predicts obesity in young adulthood. Neither study measured or controlled for insulin levels, yet TOC remarks that “The sad but inescapable conclusion is that we are now passing on our obesity to our children. Why? Because we are now marinating our children in insulin starting in the womb, they develop more severe obesity sooner than ever before.” It is unclear what evidence leads TOC to conclude that insulin is responsible for the transmission of obesity between generations.

TOC also reaches facile conclusions about the impact of sugar on insulin secretion, type 2 diabetes, and obesity. One example is on pages 158-159 where TOC discusses Chinese data on sugar consumption. TOC cites the website of a sugar broker, which suggests that sugar consumption in China is increasing by 5 percent per year. Then the book cites two news articles that both draw from the same JAMA article on the prevalence of diabetes in China. Using these figures, TOC concludes the following:

“Why sugar is so fattening? [sic] Sugar is sometimes considered “empty calories,” containing few nutrients. It is also thought to make food more “palatable” and “rewarding,” causing overconsumption and obesity. But perhaps the fattening effect of sugar is due to its nature as a highly refined carbohydrate. It stimulates the production of insulin, which causes weight gain.”

The references provided do not support the claim about insulin, and we believe it is substantially overstated.

Overall (average) score for claim 2

1.7 out of 4

Claim 3

Intermittent fasting is particularly effective for long-term fat loss.

Supporting quote(s) and page number(s)

Page 235: “Long-term weight loss is really a two-step process. Two major factors maintain our insulin at a high level. The first is the foods that we eat—which are what we usually change when we go on a diet. But we fail to address the other factor: the long-term problem of insulin resistance. This problem is one of meal timing.”

Page 236: “Instead of searching for some exotic, never-seen-before diet miracle to help us break insulin resistance, let’s instead focus on a tried-and-true ancient healing tradition. Fasting is one of the oldest remedies in human history and has been part of the practice of virtually every culture and religion on earth.”

Page 236: “The answer we are looking for is, in a word, fasting. When we talk about fasting to break insulin resistance and lose weight, we are talking about intermittent fasts of twenty-four to thirty-six hours.”

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

2 out of 4

This item received a score of 2, indicating the claim is weakly supported by current evidence. Fasting of the type advocated in TOC can certainly cause fat loss, but current evidence does not suggest that it is a unique or superior method, and we are unaware of evidence of its long-term effectiveness.

TOC presents intermittent fasting as a particularly effective long-term treatment for obesity and insulin resistance; however, it does not focus on a particular type or duration of fasting. The book discusses types of fasting ranging from simply skipping breakfast to 36-hour fasts performed 2-3 times per week, and, while not recommending any specific fasting pattern, states “Longer fasting periods produce lower insulin levels, greater weight loss and greater blood sugar reduction in diabetics” (page 257).

Several systematic reviews and meta–analyses (studies that provide an overview of the research literature) have investigated intermittent fasting and concluded that it does cause weight loss. Yet the degree of weight loss is approximately the same as standard calorie restriction, which TOC argues is ineffective. One 2018 meta-analysis investigated the differences between intermittent and continuous calorie restriction in weight loss and metabolic factors such as cholesterol and fasting insulin levels. The weight loss results were similar, but intermittent fasting yielded a “slight reduction in fasting insulin concentrations”, with “uncertain” clinical relevance. It’s worth noting that most of these meta-analyses were published after TOC, although some of the trials represented in them were available prior to its publication.

The trials represented in the most recent meta-analysis assigned people to “intermittent energy restriction” vs. “continuous energy restriction”. Typically, the intermittent energy restriction group was assigned to eat about a quarter of their usual calorie intake 1-3 days per week, while the continuous energy restriction group was assigned to consistently reduce their calorie intake by about one quarter each day. This is similar to, but not identical to, what TOC recommends.

One trial, published in 2016, did involve daylong zero-calorie fasts three days per week in adults with obesity, again using continuous calorie restriction as a comparison. This is within the range of intermittent fasting protocols that TOC recommends. Over eight weeks, both groups lost about the same amount of weight (7.1 vs. 8.2 kg) and body fat (3.7 vs. 3.7 kg). There was no significant difference in metabolic rate between groups. The researchers also followed the volunteers for 24 weeks after the diets were complete, and found that there may have been modest differences in fat regain between groups that were either significant or non-significant depending on how they measured body fat (percent fat mass vs. total fat mass). Overall, this study offers little evidence that intermittent fasting is a superior weight loss method vs. continuous calorie restriction.

Since the most recent meta-analyses were published there have been only a handful of intermittent fasting trials, but the results appear to be comparable to the above reviews. A randomized trial of people with obesity reported similar weight loss in both the intermittent and continuous calorie restriction groups, although the continuous restriction group had higher intakes of fruit, berries, fiber, and vitamin C compared to the intermittent group. Another publication from that same trial showed the intermittent fasting group to have stronger feelings of hunger, but no differences in waist circumference, blood pressure, triglycerides, and HDL cholesterol. One trial randomized adults with obesity between ages 18 and 72 to one of three groups for one year: continuous calorie restriction, a one-week-on-one-week-off intermittent calorie restriction pattern, and a calorie restriction pattern of two days/week of energy restriction. By the end of the study all groups had favorable changes in weight, fasting glucose, HDL cholesterol, and triglycerides, but there were no differences between the groups.

One exception is a trial of men with obesity that reported greater weight loss among the intermittent fasters than those on continuous calorie restricted diets.

Ultimately, any successful weight loss diet will involve calorie restriction. In fact, popular weight loss diets are simply different ways to achieve a calorie deficit. It is the same reason why weight loss surgery is effective at promoting weight loss. Intermittent fasting is a valid way to achieve a calorie deficit and promote weight loss, but it does not appear to stand out vs. other approaches.

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

2 out of 4

This item received a score of 2 indicating the references are weakly convincing. TOC generally cites reputable journals and texts, but focuses on lower-quality evidence that is more consistent with its claims. As previously discussed, the highest-quality evidence available (randomized controlled trials and meta-analyses of them) suggests that fasting of the type recommended in TOC is not uniquely effective for weight loss, and long-term evidence is scarce.

TOC does not cite this evidence, and instead relies on lower-quality evidence. For example, in the last chapter that deals most directly with the subject of fasting, TOC cites several small non-randomized studies from the 1960s to make the case for fasting. These interesting papers present data on small numbers of patients with obesity who lost weight, often with little hunger, using water-only fasts. Two of the papers describe a method that begins with a 4- to 14-day water-only fast, followed by longer periods of intermittent fasting. Often, the researchers mention having worked with a larger number of patients (up to 50), but they do not provide average data for the group as a whole, instead only presenting data for a few patients. It is unclear how representative these patients are of typical results. Although these references do offer some degree of support to the idea that fasting helps with weight control, they do not convincingly establish that fasting is a particularly effective method for long-term weight loss.

TOC also spends several paragraphs discussing case reports, which are one of the least reliable forms of evidence. We certainly do not dispute that not eating causes fat loss, and that the longer one doesn’t eat, the more fat is lost. The studies cited in TOC support this. But a water-only fast can’t be continued indefinitely, and prolonged fasting is not really what TOC recommends for long-term fat loss. The book focuses on intermittent fasting and time-restricted feeding, which can be continued indefinitely.

 

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

2 out of 4

This item received a score of 2, indicating the strength of the claim is moderately overstated.

As discussed above, available evidence shows that intermittent fasting does cause weight loss. However, the evidence also shows that the weight loss from intermittent fasting is comparable to that of standard calorie restriction, with some evidence suggesting fasting insulin is modestly reduced in intermittent fasters compared to those with daily caloric restriction.

It is important to note that the randomized controlled trials investigating intermittent fasting on outcomes like weight loss and fasting insulin are limited. The most recent meta-analysis on the topic included eleven trials, all 8-24 weeks long, with a handful of other trials completed since then. When TOC was published in 2016 the range of trials was narrower, but sufficient to indicate that intermittent fasting is probably not a superior weight loss method. According to one systematic review published around the same time as TOC, eight trials met the authors’ criteria, with most lasting 5-12 weeks and one lasting one year. We conclude that the claim is moderately overstated, because while the evidence does suggest that intermittent fasting can be beneficial, it does not indicate that it is a superior method for long-term weight loss, as TOC suggests.

Overall (average) score for claim 3

2 out of 4

Overall (average) score for scientific accuracy

1.2 out of 4

Reference 1

Reference

Chapter 4, Endnote 2. Public Health England [Internet]. Source data: OEDC. Trends in obesity prevalence. Available from: noo.org.uk/NOO_about_obesity/trends. Accessed 2015 Apr 8.

Associated quote(s) and page number(s)

TOC discusses obesity in the UK and states “Obesity increased relentlessly, even as we sweated to the oldies. Just take a look at Figure 4.1.” (page 49-50)

Criterion 2.1. Does the reference support the claim?

4 out of 4

This item received a score of 4, indicating the reference offers strong support for the claim. While the link is broken as of this writing, a look at the website in 2015 indicates that this exact graph was presented.

 

Reference 2

Reference

Chapter 4, endnote 12. McTiernan A et al. Exercise effect on weight and body fat in men and women. Obesity. 2007 Jun; 15(6):1496–512.

Associated quote(s) and page number(s)

“Many other longer-term randomized studies have shown that exercise has minimal or no effect on weight loss. A randomized 2007 study of participants who did aerobics for six days per week over one year found that women reduced their weight, on average, by 3 pounds (approximately 1.4 kilograms); men, by 4 (1.8 kilograms).” (page 54)

Criterion 2.1. Does the reference support the claim?

2 out of 4

This item received a score of 2, indicating weak support for the claim. While those figures are stated in the study, the way the results are framed in TOC is misleading. The first thing to note is that while the average participant was overweight, many of them were actually lean to begin with. Despite this, those in the exercise group lost more weight than those in the control group that did not exercise. Additionally, waist circumference decreased, and total fat mass decreased significantly in both men (-3.0 kg; -6.6 lbs)  and women (-1.9 kg; -4.2 lbs). Moreover, the results showed that those with greater physical activity lost more weight and fat than those performing less physical activity, and those who were overweight and obese lost the greatest amount of weight. In other words, this study suggests that exercise meaningfully reduces weight and fat mass.

Reference 3

Reference

Chapter 8, endnote 23. Bliwise DL. Historical change in the report of daytime fatigue. Sleep. 1996 Jul; 19(6):462–4.

Associated quote(s) and page number(s)

TOC discusses sleep deprivation and states the following: “Shift workers are especially prone to sleep deprivation and often report fewer than five hours of sleep per night.” (page 94)

Criterion 2.1. Does the reference support the claim?

1 out of 4

This item received a score of 1, indicating that the reference does not convincingly support the claim. The cited text from 1996 discusses sleep and how self-reported fatigue has increased since 1930, according to a particular personality survey, the Minnesota Multiphasic Personality Inventory (MMPI). It mentions shiftwork, but only insofar as stating that more Americans likely engage in shiftwork than they used to. This is mentioned only in the introduction and cites other research. It is unclear exactly where the “fewer than five hours of sleep per night” figure comes from; however, a 2007 paper published in Nutrition Research Reviews states the following: “Some studies have used shift workers, for example, who might sleep less than day workers, under 5 h on working days” which also mistakenly cites the Bliwise paper as evidence of the claim.

Reference 4

Reference

Chapter 13, endnote 7. Bergmann RL et al. Secular trends in neonatal macrosomia in Berlin: influences of potential determinants. Paediatr Perinat Epidemiol. 2003 Jul; 17(3):244–9.

Associated quote(s) and page number(s)

“Obesity is considered an energy-balance problem, one of eating too much or exercising too little. Since six-month-olds eat on demand and are often breastfed, it is impossible that they eat too much. Since six-month-olds do not walk, it is impossible that they exercise too little. Similarly, birth weight has increased by as much as half a pound (200 grams) over the last twenty-five years. The newborn cannot eat too much or exercise too little.”

Criterion 2.1. Does the reference support the claim?

0 out of 4

This item receives a score of 0, indicating that the reference undermines the claim. The cited study seems to offer nothing to support the claim of birthweight increasing by half a pound over the last twenty-five years. In fact, it seems to investigate a related but distinctly different topic, which is the prevalence of neonatal macrosomia in Berlin. Neonatal macrosomia is a term used to describe newborns with a birthweight of greater than 4,000 grams (or 8 pounds, 13 ounces). The results of the study indicate that prevalence of neonatal macrosomia increased in Berlin between 1993 and 1999. This item would have only received a score of one had the study itself not contradicted the claim of birthweight increasing by 200 grams. According to the Bergmann paper “The prevalence of a birthweight of ≥4000 g increased significantly from 9.1% to 10.1%, although the mean birthweight remained relatively constant (3325 g in 1993, 3322 g in 1999).” Additionally, following the quote above in TOC, the book goes on to suggest that the reason for the increased weight gain in newborns could be due to pesticides in the food supply and cites a rodent study as evidence. However, the Bergmann paper also investigates the risk factors that are associated with neonatal macrosomia and pesticides are not mentioned; rather, other factors such as maternal age, height, smoking status, and diabetes, were mentioned as the potential risk factors.

Reference 5

Reference

Chapter 15, endnote 6. American Diabetes Association [Internet]. Low calorie sweeteners. Edited 2014 Dec 16. Available from: diabetes.org/food-and-fitness/food/what-can-i-eat/understanding-carbohydrates/artificial-sweeteners. Accessed 2015 Apr 12.

Associated quote(s) and page number(s)

“The American Diabetes Association states on its website, ‘Foods and drinks that use artificial sweeteners are another option that may help curb your cravings for something sweet.’” (page 147)

Criterion 2.1. Does the reference support the claim?

4 out of 4

This item received a score of 4, indicating the reference offers strong support for the claim. The ADA webpage makes that statement verbatim.

Reference 6

Reference

Chapter 16, endnote 12. Fuchs CS et al. Dietary fiber and the risk of colorectal cancer and adenoma in women. N Engl J Med. 1999 Jan 21; 340(3):169–76.

Associated quote(s) and page number(s)

“At first, it was believed that high fiber intake reduced colon cancer. The subsequent studies proved to be a bitter disappointment. The 1999 prospective Nurses’ Health Study followed 88,757 women over sixteen years, and found no significant benefit in reducing colon cancer risk.” (page 181)

Criterion 2.1. Does the reference support the claim?

4 out of 4

This item received a score of 4, indicating the reference offers strong support for the claim. TOC accurately cites the findings of the Nurses’ Health Study, so it is narrowly correct. However, the broader claim that subsequent studies did not support the fiber-cancer link does not appear to be correct. Most (not all) meta–analyses of observational studies, which provide a less-biased overview of the research literature as a whole, report that a higher intake of fiber is associated with a small but significant reduction in colon cancer risk. Since this could have been scored differently depending on whether we’re considering the narrow claim or the broad claim, we decided to give TOC the benefit of the doubt.

Reference 7

Reference

Chapter 19, endnote 46. Aviram M, Eias K. Dietary olive oil reduces low-density lipoprotein uptake by macrophages and decreases the susceptibility of the lipoprotein to undergo lipid peroxidation. Ann Nutr Metab. 1993; 37(2):75–84.

Associated quote(s) and page number(s)

“Olive oil contains large amounts of antioxidants including polyphenols and oleocanthal, which has anti-inflammatory properties. Among its purported benefits are reduced inflammation, lowered cholesterol, decreased blood clotting and reduced blood pressure.” (page 234) The reference in question refers specifically to the claim of lowered cholesterol.

Criterion 2.1. Does the reference support the claim?

3 out of 4

This item received a score of 3, indicating the reference offers moderate support for the claim. It may be more accurate to state that the Aviram study provides evidence that LDL oxidization is reduced along with a reduction in LDL uptake by a cell type called macrophages. The study did demonstrate that ten individuals with normal cholesterol levels had their LDL reduced after two weeks of a diet supplemented with olive oil.

Reference 8

Reference

Chapter 20, endnote 18. Stote KS et al. A controlled trial of reduced meal frequency without caloric restriction in healthy, normal-weight, middle-aged adults. Am J Clin Nutr. 2007 Apr; 85(4):981–8.

Associated quote(s) and page number(s)

“Studies of eating a single meal per day found significantly more fat loss, compared to eating three meals per day, despite the same caloric intake. Significantly, no evidence of muscle loss was found.” (page 243)

Criterion 2.1. Does the reference support the claim?

4 out of 4

This item received a score of 4, indicating the reference offers strong support for the claim. The Stote study is a randomized crossover trial that found those eating one meal per day led to decreased fat mass and no significant loss of fat-free mass (muscle is a large part of fat-free mass). There were also some other interesting findings in the study, such as that those eating one meal per day had greater total cholesterol (including both LDL and HDL), blood pressure, and a greater reported hunger and desire to eat over those eating three meals per day. Those eating three meals had nearly double the cortisol of those eating one meal, felt fuller, and weighed an average of 1.4 kg more. Although the researchers provided all food to participants in the study, which increases compliance and the accuracy of food intake measures, the participants were not supervised during most meals so it is not certain that the calorie intake estimates provided in the paper are accurate. Because of this, we cannot be very confident that calorie intake was actually the same between groups.

Reference 9

Reference

Chapter 20, endnote 24. Drenick E. Prolonged starvation as treatment for severe obesity. JAMA. 1964 Jan 11; 187:100–5.

Associated quote(s) and page number(s)

“While most expect the fasting period to be extremely difficult, clinicians noted the exact opposite. Dr. E. Drenick wrote, ‘The most astonishing aspect of this study was the ease with which prolonged starvation was tolerated.’” (page 245)

Criterion 2.1. Does the reference support the claim?

4 out of 4

This item received a score of 4, indicating the reference offers strong support for the claim. This quote is contained within the Dreneck paper, although there is a bit of inconsequential text that is omitted from the direct quote. The full quote is “The most astonishing aspect of this study, to the patient and to the physician, was the ease with which prolonged starvation was tolerated.”

Reference 10

Reference

Chapter 20, endnote 35. Williams KV et al. The effect of short periods of caloric restriction on weight loss and glycemic control in type 2 diabetes. Diabetes Care. 1998 Jan; 21(1):2–8.

Associated quote(s) and page number(s)

After describing a 2011 study on intermittent fasting, the following is stated: “Further studies confirm that the combination of intermittent fasting with caloric restriction is effective for weight loss.” (page 247)

Criterion 2.1. Does the reference support the claim?

3 out of 4

This item received a score of 3, indicating the reference offers moderate support for the claim. This study was designed with three arms: one group ate a moderately calorie-restricted diet for 20 weeks while the other two groups ate the same moderately-restricted diet, but with periods of a very low-calorie diet. The difference in the last two groups was the varying days on which those very low-calorie diets were consumed. All groups lost weight, but the very low-calorie groups lost more weight than those simply on the moderately-restricted diet because they ended up consuming between 18,000 and 28,000 less calories. So if you assume that the days of very low calorie consumption were equivalent to fasting then the claim is true. But according to this study it is also true that you can lose weight without the fasting, just not as rapidly.

Overall (average) score for reference accuracy

2.9 out of 4

Summary of the health-related intervention promoted in the book

The cornerstone of the dietary recommendations in the book is intermittent fasting. It also includes the following five guidelines:

1. Reduce your consumption of added sugars.
2. Reduced your consumption of refined grains.
3. Moderate your protein intake (in practice, TOC recommends a high protein intake of 20-30 percent of calories).
4. Increase your consumption of natural fats.
5. Increase your consumption of fiber and vinegar.

Condition targeted by the book, if applicable

TOC focuses primarily on the condition of obesity, hence the title; however, it also discusses type 2 diabetes since TOC makes the case that obesity and insulin resistance are connected.

Apparent target audience of the book

The intended audience for this book is not explicitly stated, but it appears to be intended for a general adult audience. Presumably the book is focused on those adults who are obese and are interested in losing weight, as suggested by the text on the cover: “Why you’ve never been able to lose weight, and how that can change now.”

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

3 out of 4

This item receives a score of 3 indicating the intervention is likely to moderately improve the condition. Currently, not many studies on intermittent fasting last 6 months or more, but the available evidence suggests that it does cause weight loss. A 2018 systematic review of the literature included only six studies on intermittent fasting as a treatment for overweight and obesity, and only three of those trials lasted six months or more. The review concluded that an intermittent fasting diet results in greater weight loss than no diet at all and results in equivalent weight loss to continuous calorie restriction. However, the weight loss produced by both methods is rather modest. Since that review was published, other human trials lasting 6 months or more have supported this conclusion. This diet method is also about as effective at improving glycemic control in patients with type 2 diabetes compared to continuous energy restriction.

It’s important to note that these are only a handful of studies, and that none of these last longer than one year.

TOC also recommends reducing intake of added sugars and refined grains and eating a high-protein (20-30 percent of calories) diet. In practice, this will result in a less-processed, higher-protein, lower-carbohydrate diet, which on its own promotes weight loss. It seems likely that combining intermittent fasting with this diet would increase its weight loss effectiveness to some degree, but to our knowledge this remains to be tested scientifically. Individually, none of these approaches have produced impressive and durable weight loss results in scientific studies, but that isn’t a critique of this diet in particular– achieving substantial, lasting weight loss through diet is an ongoing challenge for the research and medical communities.

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

3 out of 4

This item receives a score of 3 indicating the intervention is likely to moderately improve general health.

The American Heart Association published a report in 2017 that examined intermittent fasting and its impact on cardiovascular and metabolic risk. The evidence was mixed, but there seemed to be a trend toward a beneficial effect. According to the report: intermittent fasting lowered triglycerides but had no effect on cholesterol; intermittent fasting reduced insulin resistance and lowered insulin concentrations, but had no effect on glucose levels; and intermittent fasting also lowered blood pressure, but a minimum of 6% body weight had to be lost to see these results. This indicated that the benefit may derive from weight loss itself rather than the diet used to achieve the weight loss. Moreover, the clinical trials included in the report were only a few weeks in duration, with the longest being 24 weeks.

Since that report was published, a one year trial of alternate day fasting was published that compared it to continuous caloric restriction and a control group that did not diet. Interestingly, there were no differences in blood pressure, total cholesterol, fasting plasma glucose, and inflammatory factors between the three groups. However, LDL levels were elevated in the alternate day fasting group compared to the other groups after one year.

With respect to intermittent fasting and a disease like cancer there is even less information available, most of which rely on animal studies and epidemiological data to draw conclusions. A 2014 systematic review and meta-analysis indicates that both general caloric restriction and ketogenic diets are effective at reducing cancer in animals while intermittent fasting was not. A 2016 review article states that some short-term studies of intermittent fasting show a decrease in biomarkers like adipokines and inflammation that may someday lead to cancer, but the evidence is not consistent.

The recommendations related to diet quality, such as eating less added sugar and refined grains, and eating more protein and fiber, should deliver meaningful health benefits for most people. This will be accentuated by the focus on nutrient-dense, unrefined foods found in the meal plans. Aside from the harder-to-quantify benefits of eating more nutrient-dense food, this diet plan will cause weight loss in most people, and that will tend to improve general health. A reasonable analogy for this diet is the “healthy low-carbohydrate diet” of the DIETFITS trial, a whole-food-based low-carbohydrate diet that produced 13 pounds of weight loss at one year. Although DIETFITS participants had the advantage of extensive support from research staff, those who follow the guidelines in TOC will likely lose some amount of excess fat.

An additional element of TOC worth mentioning is exercise. An entire chapter is devoted to making the argument that exercise and physical activity do not cause weight loss. However, Appendix B states that “all types of exercise, including resistance (weights) and cardio, are encouraged” and that ” fasting is an ideal time to exercise.”  So although TOC doesn’t believe exercise contributes to weight loss, it nevertheless encourages it, which we think is a good thing for health.

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

This item receives a score of 3, indicating that the diet is likely more than nutritionally adequate.

The dietary pattern recommended in TOC is not a rigidly restrictive diet. Rather it seems that there are a set of guidelines to apply to any dietary pattern (reduce added sugars, reduce refined carbohydrates, moderate protein, increase natural fats, and increase fiber and vinegar) as long as fasting is incorporated. The net effect of these guidelines will likely be an increase in the diet’s concentration of essential and nonessential nutrients for most people. This is particularly true when we consider the book’s meal plans, which appear to be mostly whole-food-based.

 

Overall (average) score for healthfulness

3.0 out of 4