The Allure of Weight-Loss “Science”

Take your pick? South Beach, Pritikin, Atkins, Zone? Ornish, Dukan, Keto, Mayo? SlimFast, Optifast, Medifast, DASH? So many weight-loss diets! Which ones work? Which is best? What does the science say?The typical assumption, of course, is that science must have some verdict here. Science is all about knowledge and answering our questions. Here, I explore this Sacred Bovine—the conviction that all science is inherently certain and that the Scientific Method implicitly guarantees unambiguous conclusions. Might we accept, by contrast, a role for uncertainty in our view of the nature of science? Not all science (especially “science-in-the-making”), we may find, has ready or clear answers. Thus, we may need to help students learn about the limits of science and to adjust their expectations accordingly.Losing weight would seem a prime occasion to rely on and benefit from science. One should simply hypothesize different diets (such as the list above), test each, and find which among them is effective and/or which is most effective. That would seem simple enough, given the portrait of science normally promoted in school.Yet I have been fascinated over the years watching the cavalcade of diets come and go. Low fats one week, low carbs the next. That might foster a deep cynicism about science (see, e.g., Cassetty, 2019; Garza et al., 2019). Indeed, a Pew poll indicated that many Americans take nutritional recommendations “with a grain of salt,” we might say. Much of this skepticism seems to stem from a desire for certainty. We hear many complaints of apparently seesawing science. But revising beliefs when new evidence appears also seems sensible. As does withholding judgment if there is not sufficient evidence. Perhaps we need to foster more realistic expectations about science itself, and when it can be considered “done”?One might think that doing dietary science is as simple as a basic science fair project, using a controlled experiment. For example, one might recount the story of Jacques Cartier, who helped cure his ship crew of scurvy by having them drink juniper tea (Leland, 2012). Or the later work of James Lind, who tested a variety of beverages against each other, from vinegar to barley water to lemon juice (Brown, 2005). Or Christian Eijkman, who found a remedy for beriberi (thiamine deficiency) in Java in the 1890s by comparing chickens eating polished (white) versus unpolished (red) rice (Allchin, 2012).But nutrition science is challenging. One can easily explore vitamin deficiencies because they involve just one variable. Yet the human body is complex, and the number of nutrients is many. There is no one single factor (certainly, no one nutrient) responsible for body weight. Nor can one ethically divide a large population into two groups and prescribe strict diets for everyone.What’s more, there is a great deal of individual variation in genetics, microbiota, and metabolism. Different people respond differently to eating the same foods. Even the “same” food may vary in content. Nor do we eat any given nutrient in isolation. Moreover, it is hard to construct (or find) sufficiently similar groups to compare fairly. To accommodate the variability—to allow the unpredictable effect of different factors to balance each other—one needs an exceptionally large sample—rarely achieved. There is considerable risk of confounding (correlated, but irrelevant) variables hiding in the two groups (for example, some food choices may reflect other dietary “biases” or lifestyle habits). And this all assumes that participants follow the intended dietary regime or report their diet accurately. The data may easily end up being “noisy”—with no discernible discrete patterns, unless the effects are quite large. Weight loss, in particular, is a long-term problem. One needs sufficiently extended studies to detect significant effects. In summary, the ideal of tightly controlled studies is not often found. Nutritional science is “messy” (Belluz, 2016; Ludwig & Heymsfield, 2019).In addition, because of individual variation, what works for one individual may not work for another. The results of a study reflect an average, not necessarily universal outcomes. One cannot always generalize, even from a large group, to any particular person.In general, we must come to terms with the limits of science, including research on weight loss. Some science remains uncertain. Results may be ambiguous. Outcomes from different studies may not perfectly align. Experts may disagree. The science may simply be incomplete. When students are exposed to a “diet” of nothing but textbook science—the static knowledge tested by time and experience—teachers may need to introduce them to the nature of science-in-the-making. Not all science is ready-made.This does not mean there is no scientific basis for certain nutritional guidelines. But it does mean one needs to be careful. What is a reliable benchmark? The temptation may be to dash online and read the first “scientific” study one encounters. But we would be well advised to be more conservative and to invest our trust instead in the consensus of the relevant experts. That means that one published study alone, even if it is hyped by the media, is insufficient. Even if it is published in a peer-reviewed journal! For reliability, one looks for agreement among numerous studies. One seeks work that has been vetted from a variety of critical perspectives. That raises an interesting question for the outsider, or the non-expert: who speaks for science?Here, we might shift focus from the science of a healthy diet to a diet of healthy science. What does it mean to be an effective consumer of science, and to be well informed? Many students may imagine it is no more difficult than googling some basic question online. Type a few keywords. Hit return. And the whole of scientific expertise—say, about “most effective weight-loss diet”—seems at our disposal.There is a growing movement in some circles to “Do Your Own Research” (DYOR). While teachers generally want to promote intellectual independence among their students, this tendency proves deeply problematic here. Research has now demonstrated that those who adopt the DYOR posture are less likely to reach conclusions consistent with the scientific consensus (Ballantyne & Dunning, 2022). In one study, DYOR-oriented consumers exhibited more COVID-19 misperceptions, and lower levels of trust in science generally (Chinn & Hasell, 2023). In another study, mothers chose not to vaccinate their children based on their self-styled critiques of expert research, substituting in part their own personal experience (Carrion, 2017). Puzzling, perhaps?DYOR offers several beguiling notions. The first is the illusion that anyone can think like a scientist and that therefore their personal judgment can trump the experts. Yet specialized knowledge matters. And it is distributed across our culture: from doctors and lawyers to Wi-Fi technicians and bridge welders—and nutritional scientists. We depend on each other’s expertise. And that requires informed trust. Not ordinary trust, related to loyalty or promise-keeping. But epistemic trust: recognizing occasions when someone else knows more than we do. For example, who can determine which diet studies are “significant”? What is the consensus across many years of research? How might the claims be qualified, based on intimate familiarity with the research methods? These all require expert judgment. Not casual interloping by a non-expert. Ironically, an invitation to DYOR supports epistemic hubris: a misleading impression of one’s own authority (Dunning, 2019).The second misstep is succumbing to a set of belief traps. Chief among these is relying on mere plausibility, rather than on solid evidence. That seems to be how our minds tend to work, unchecked. Alas, it is easy to create a rhetorical impression that a weight-loss diet is effective, even with no rigorous investigation to back it up (e.g., Bjarnadottir, 2019; MacKeen, 2020). Our minds also respond more strongly to anecdotes and stories than to numbers and data. Knowing that one should rely on systematic evidence requires training. We need to learn not to place a premium on personal testimony about diets—from chatty books to popular magazines or blogs. Even so, our minds seem predisposed to draw conclusions on shards of evidence. Nobel Prize-winning psychologist Daniel Kahneman (2011) calls it “What you see is all there is” (WYSIATI). We are easy prey to cherry-picked evidence: unrepresentative, but chosen to support a certain claim. Only the expert knows whether the evidence is complete and balanced. So, we need to be savvy consumers of science, discerning which sources of information are truly reliable. We need to inquire—not about the scientific arguments themselves—but about who can vouch for the expert scientific consensus.The belief traps are aggravated by the pervasive image that all science is authoritative and beyond question. We are generally not comfortable living with uncertainty. So, we are susceptible to accepting uncritically even ill-informed appeals to “science.” Purveyors of disinformation use the gap of uncertainty and the blindness of our inexpertise to their advantage.A third major pitfall is our personal motivations and desires. The prospect of weight loss is often fraught with emotion. That makes us yet more vulnerable to all those belief traps. Attractive promises and assurances may easily substitute for rigorous science.Finally, much published nutrition research has been tainted by conflict of interest (e.g., Bero, 2015; Nestle, 2015). The food industry conducts research, hoping to lend credibility to their product, whether sugars, fats, or weight-loss “supplements.” Unfortunately, when research seems unfavorable to the sponsor, it may never get published. Other studies violate preregistered protocols for analysis. Namely, there may be statistical gerrymandering: reconfiguring data to align with desired conclusions. All the hullabaloo from the 1960s about fats leading to fatness? It was based on research funded by the sugar industry. The consumer of weight-loss claims must always be prepared to ask: is this science or sales in disguise?All these tendencies, which lure us into a semblance of science, are well illustrated in the case of the Paleo Diet. The premise is plausible, perhaps. Advocates assume that our human ancestors evolved digestive enzymes and physiology to food sources available at the time. But (the reasoning continues) human diets have since changed. We eat more dairy later in life, for example. With civilization, we came to rely more on cultivated crops, especially grains. Yet (presumably) our genes have not had time to adapt. So, lactose intolerance and gluten allergies seem to plague us. The solution, apparently, is to eat like a hunter-gatherer, with plentiful meats and fresh fruits and vegetables. Do we mistake appealing to our desires as justification for our food preferences?Unfortunately, many of the assumptions of the Paleo Diet proved wrong (Zuk, 2013). Once one does the historical research, one finds that there was no uniform (standard) ancestral human diet. Various populations adopted a diversity of diets in different localities. Further, natural selection can indeed be effective on the scale of human history. For example, the regulation of the lactase gene has shifted, and it may function later in life. Ultimately, the Paleo Diet seems little more than a rationalization of a diet that someone found attractive. It is not trustworthy science.A story with the trappings of science is not the same as good science. To be well informed, we need to rely on the consensus of the relevant experts, not our own intuitions, or judgments based on a haphazard encounter with selected evidence and explanations.Attention to diet and its implications for excess weight certainly seems immensely important. Harvard researchers anticipate that “close to 50 percent of the population will be obese by 2030 and 25 percent will have severe obesity” (Cassetty, 2019). What can the student do? What can be believed?NGSS Science and Engineering Practice 8 asks students to find scientific information and evaluate its reliability. That might involve, first, learning who speaks for science. We should rely on reputable institutional sources that can faithfully convey the expert consensus. By consulting such sources—such as the Mayo Clinic or the National Academy of Science—one quickly finds that dietary recommendations are very general and cautious (Mayo Clinic, 2024). Many different weight-loss diets seem to work—but for different people and in different contexts with different conditions. Appreciating the uncertainty of some science seems—especially amidst the deluge of misinformation in the media and encouragement to “DYOR”—an essential foundation for consuming science, and thus for understanding how science can inform our dietary choices.