Four Big Fat Lies About Weight and Health
“There are many core ideas within the science of body weight and health and the mediating role played by exercise and diet which are simply assumed to be true.”
—Michael Gard and Jan Wright,
from The Obesity Epidemic:
Science, Morality, and Ideology
We were sitting in my backyard, a handful of women from the neighborhood. We were eating cake, as it happened, a lemon poppy seed cake I’d made that morning, and drinking iced tea, and talking about our lives. And so it was inevitable that the conversation came around to weight—the weight we wanted to lose, the weight we’d gained, the weight other women had lost or gained, or lost and gained. A typical conversation, in other words.
One neighbor mentioned a popular TV show featuring an actress who would be too large to, say, walk the runway. “I can’t stand watching anything she’s in,” the neighbor commented. “I’m afraid she’s going to have a heart attack any minute. She’s just so unhealthy. Doesn’t she know she could drop dead? Doesn’t she know what she’s doing to herself?”
There was no chance the actress would keel over mid-episode, of course, since the show was taped. So what was my neighbor getting at? I puzzled over this for a few minutes and eventually realized that she wasn’t actually worried about the actress’s health. She couldn’t be because in fact she didn’t know anything about the actress’s health. So what she was actually talking about was the actress’s looks, not her health. My neighbor thought the actress was unattractive because of her weight, but she was too politically correct to come out and say that. Critiquing health in the context of weight, on the other hand, is perfectly acceptable in polite company. In fact, it’s almost required in some circles. As health—or at least the perception of health—has become a social and moral imperative, judging other people’s health status has become not just accepted but expected.
In his 1994 book The Death of Humane Medicine and the Rise of Coercive Healthism, Czech doctor Petr Skrabanek defines healthism as a worldview that judges human behaviors by how we think they affect health. Notice that emphasis on belief rather than reality. Behaviors we believe make people healthier—exercise, for instance—carry a sense of moral virtue; we say we’re “being good” when we take the stairs instead of the elevator, eat salad (without dressing!) for lunch, spend an hour at the gym. Behaviors we think make people less healthy are seen as unacceptable. We’re “bad” when we eat a slice of cake or binge-watch Orange Is the New Black. Restaurant owners know this; it’s why they name desserts sinful cheesecake or chocolate decadence, cleverly acknowledging and deflecting the moral judgment around eating them.
A lot of what we believe about weight and health comes from assumptions of healthism, starting with one of our most cherished truths: Fat is unhealthy. It’s a statement that’s so general, so broad, and also so deeply ingrained, it’s hard to pin down precisely what it means. It’s just something everyone knows, right?
It’s hard, in this culture, to question that statement—to even imagine a world where that might not be true. That long-ago day in my therapist’s office, I couldn’t even consider the possibility that maybe being overweight or even obese wasn’t as unhealthy as I thought. I’d spent years worrying about how my weight might affect my health, especially after one doctor sat me down and told me if I were her sister, she’d put me on a diet that very minute and make me lose weight. “If you don’t,” she warned, “you’ll wind up with heart disease or diabetes or high blood pressure, or all three.”
She didn’t have to convince me; I already believed it. I already worried about whether I was (as my grandmother often commented about others) eating myself into an early grave. I already imagined my arteries clogging each time I ate a shred of fat—actually, each time I ate anything. By the time I got home from the doctor’s office that day, panic was making me hyperventilate, which made me think I was having the threatened heart attack right then and there.
Compared with some of the stories I’ve heard from other people since then, her tactics were actually pretty mild. She didn’t, for example, refuse to treat me until I’d lost weight, or write “noncompliant” on my medical chart, or try to sell me a carton of Medifast or a Weight Watchers membership. But she did make it perfectly clear that I would never be healthy until I lost weight. (And she did insist on putting me on statins, which caused extreme muscle pain; I had to stop taking them.) Ironically (but predictably), her well-meant lecture had the opposite effect from what she’d intended. I went into such an anxious tailspin over the next few weeks that I wound up stress-eating and gaining weight.
It took a family catastrophe to make me even start to question what I thought I knew: my then-fourteen-year-old daughter developed anorexia. As I sat hour after hour in silent shock beside her bed in the ICU, the formula I’d lived with all my life without question—fat=bad, thin=good—seemed horribly skewed. Suddenly there was such a thing as too thin, and it was right in front of me, in the matchstick thinness of her arms and legs, the almost pornographic arches of her ribs, the knobs of her vertebrae.
My feelings about food were turned upside down, too. In the months that followed I became an expert in making calorie-dense concoctions, to deliver the nutrition my daughter needed in the smallest possible footprint. Foods like butter and nuts and ice cream, long banned from my kitchen and my diet, now filled the shelves and freezer and featured prominently at each meal. My husband and I ate what she ate to help her get past her fears about fat, to model for her that this was how normal people ate. I knew I couldn’t show any ambivalence about what we were eating; my daughter is extremely empathic, especially when it comes to my feelings. So I had to not just eat what she was eating, not just pretend it was normal, but actually believe it was. I couldn’t do it for my own sake, but I could do it for her.
Over the next year, as my daughter put on weight and wrestled with recovery, I grappled with my own feelings about food and weight in a new way. My former fear foods were now saving my daughter’s life, and, maybe, protecting my younger daughter from the same disease. Each pound my oldest gained represented not a problem to be solved but a victory over the demon that held her hostage. Now when I saw a young woman who wouldn’t be considered thin, my automatic thought was She’s lucky (though, to be clear, it’s entirely possible to be overweight or obese and suffer from anorexia or other eating disorders). Now when I hit the grocery store, I scrutinized labels looking for the highest-calorie foods rather than the lowest. My daughter and I even laughed about it one night, as we noticed the horrified looks from other shoppers observing our quest for more calories.
But maybe the biggest push to change came from watching other people respond to my daughter. Especially middle-aged women. More than once, when she was sick, they literally approached her in the street to praise her beauty, admire her gaunt figure, even ask her—a fourteen-year-old girl who looked like she was dying—for diet tips. Even friends who knew how ill she was commented admiringly on how gorgeous, how svelte she was. It was as if they couldn’t help themselves. And if I hadn’t seen with my own eyes how much my daughter suffered and how ill she truly was, maybe I would have found her thinness glamorous and beautiful, too.
As she put on weight, those comments dried up. To my eyes she looked infinitely better with every pound; her eyes were shining and she actually smiled. To the rest of the world, apparently, only ribs you could count, sunken cheeks, and a hollow look were worth praise.
I learned a lot from my daughter’s anorexia. I learned about the neurobiology of hunger and appetite, how our “choices” around food and eating are influenced by physiology. People with anorexia nearly always have a family history of eating disorders, anxiety, or both. They also tend toward certain personality traits, inborn characteristics like perfectionism that don’t typically change over the course of a person’s lifetime. I learned that, in fact, both my own struggles with food and my daughter’s illness likely derived from our shared genetics and the way our brains are wired rather than from anything either of us had said or done or experienced.
Most tellingly, I learned how pervasive and deeply entrenched our culture’s preference for thinness has become. We all know this, of course; we’ve read countless magazine stories about it, we’ve taken college courses on it, we talk about it with our friends and families. But until you experience it personally on some level, it’s hard to fathom just how automatic it’s become.
For example: one of my best friends grew up tall and thin, never worrying about her weight. In middle age she gradually put on about 30 pounds. Last year she lost the weight, not because she set out to diet but because she changed the way she ate for health reasons. She was shocked at the number and kind of comments that suddenly came her way. “I began feeling that how I look is of supreme importance to other people,” she told me. “And then it struck me that we shouldn’t be focusing on the physical image we project at all.” She’d been listening to me think and talk about the issue for years, but it took her own experience to make her aware of both how common and how destructive such judgments are. “Honestly, until that moment, the subliminal thoughts I had sometimes about myself and others still had an echo of my mother’s idea that being overweight means ‘having no self-control,’” she said. “Now I don’t think that will be true.”
A HUGE PART of being OK with my body was connected with health. Could I be healthy at this weight and at this point in my life? According to the BMI chart, I was mildly obese, a phrase that haunted me. And of course “everybody knows” you can’t be obese and be healthy.
But what does health actually mean? It’s a fuzzy concept, a word so general, so vague, that it’s virtually meaningless. Is health the simple absence of illness? That didn’t feel right; we talk about health as a positive condition, not just the lack of a negative one. Is health the state of feeling good, energetic, happy? Is it a physical state, or does it include mental and emotional well-being, too?
I couldn’t answer any of these questions. Neither could anyone else, it turned out, though plenty of people have tried, starting with the World Health Organization (WHO), arguably the world’s experts on this subject. In 1948, WHO issued a statement defining health as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity.” That seems more aspirational than realistic; by that standard, I don’t know anyone who would qualify as healthy.
Other experts have tried to refine that definition. Alistair Tulloch, a now-retired British MD, posed the question in a 2005 article in the British Journal of General Practice. He suggested that since we live in a world full of accidents, infections, disease, poverty, poor working conditions, and a host of other adversarial forces, health measures our ability to adapt to and function normally in such a hostile environment. It’s an interesting idea, but it suffers from the same vagueness that afflicts the word health: Who’s to say what “normal” functioning is or what “successful adaptation” looks like?
Around the same time Tulloch was grappling with the concept, Swiss pharmacology professor Johannes Bircher took a stab at the problem with this statement: “Health is a dynamic state of well-being characterized by a physical, mental, and social potential, which satisfies the demands of a life commensurate with age, culture, and personal responsibility.”
That covers a lot of bases; too many, really, like the WHO definition. And it doesn’t bring us closer to a common vocabulary. It doesn’t speak to the fact that what’s healthy for me might not be what’s healthy for you because we have not just different physical and mental needs but different expectations, too. For instance, I’ve struggled with a chronic anxiety disorder since I was ten. Thanks to exercise, meditation, and better living through chemistry, I experience way fewer symptoms than I used to. I function a lot better than I did in my twenties, but maybe not as well as someone without an anxiety disorder. Does that make me mentally unhealthy?
I don’t care either way; I don’t need to slap a label on my mental health. But it does matter to me, you, and everyone when, say, we open a magazine or a website and see the headline “Can you really be fat and healthy?” Because the answer depends on what we mean by “healthy.” Are we talking about medical health? Psychological health? Heart health? Nutritional health?
Medical definitions, which often dominate the conversation, tend to focus on measurable characteristics like cholesterol or glucose levels. Depending on age, gender, genetics, and other factors, one person’s ideal range might be too high or too low for someone else. Even medical health is a constantly moving target, one we’re not likely to come to consensus about any time soon.
But let’s get back to the magazine or the website, to the automatic sense of dread we get when we read a headline that asks whether we can be overweight and healthy. Most of us aren’t parsing the literal definition of health; we’re worrying about our health, our weight. And the message we get over and over is that we can’t be healthy—whatever that means—and also be overweight or obese.
So what do we actually know about the relationship between health and weight? I’ve talked to hundreds of experts, looked at more than a thousand studies, and immersed myself in the research over the last five years to investigate this very question. (And there’s a lot of it; I’ve been told by numerous researchers that the easiest way to get a study funded now is to include the word “obesity” in the proposal. Even better, cite “childhood obesity.”) What I’ve found is that much of what we think we know either isn’t true or doesn’t mean what we think it means, starting with these four often-repeated “facts” about weight and health.
1. Americans are getting fatter and fatter—at this rate nearly half of us will be obese by 2030!*
We measure overweight and obesity these days using body mass index, or BMI, a ratio of height to weight. Doctors and scientists like BMI because it’s convenient and noninvasive; you just plug in the measurements and voila! You’ve got an easily quantifiable way to characterize, compare, and contrast. The problem with BMI is that it’s not an accurate measurement or predictor of health, especially for people who are shorter or more muscular than average. It doesn’t take into account the amount of muscle or fat in a body, or how strong (i.e., heavy) a person’s bones are. It says nothing about a person’s future risk of disease or death, and it was never intended to. Its creator, Belgian mathematician Adolphe Quetelet, came up with BMI in the 1830s as a way of looking at trends in populations, not in people. But in the late 1970s, researchers began using it precisely as Quetelet hadn’t intended, to categorize individuals’ weight and health status. It’s been the medical model go-to ever since.
Body Mass Index Chart from Vertex42.com. Reprinted with permission.
Using the BMI categories, the most recent reports from the Centers for Disease Control and Prevention (CDC) classify 34 percent of adult Americans as overweight and another 35 percent as obese. About 2 percent of adults are considered underweight; the rest fall into the “normal” category.
The number of overweight and obese Americans* has certainly risen since the mid-twentieth century, with the biggest jump happening between 1980 and 2000. The first researcher to call attention to the change was Katherine Flegal, an epidemiologist at the CDC’s National Center for Health Statistics in Hyattsville, Maryland, which compiles statistics on everything from fertility to mortality. Flegal, who grew up in Berkeley, California, wears her hair short and spiky, and looks at least ten years younger than seventy, has spent her career analyzing all sorts of medical data. According to her research, between 1960 and 1991 the percentage of Americans who were overweight rose from 25 to 33 percent.
It’s hard to get a precise handle on how those numbers compare with today’s, though, because the definitions changed abruptly between then and now. Before 1998, the BMI chart had only three weight categories: “underweight,” or below 18.5 on the chart, which included 2 percent of Americans; “normal,” from 18.5 to 27.3 (the cutoff for men was higher) on the chart, which applied to 40 percent of Americans; and “overweight,” anything above 27.3 (or 27.8 for men), which covered 58 percent of the population. Those cutoffs were revised downward in 1998 to where they are now,, and a category for “obesity” was added. So comparing pre-1998 BMI statistics to post-1998 stats is like comparing pre-steroids home run records to those made in the age of performance-enhancement drugs. In other words, more or less pointless.
Still, we do know a few things. The average American is in fact heavier (by about twenty pounds) and taller (by about an inch) than we were in 1960. And dire predictions notwithstanding, the rates of overweight and obesity leveled off around 2000. We’re not actually getting heavier and heavier; our collective weight has pretty much plateaued.
Why did our weight go up? Plenty of experts have theories, including what I think of as the Big Three: We eat too much. We eat the wrong kinds of foods. We exercise too little. There’s likely some truth in all those statements (for everyone, not just for those on the heavier end of the weight spectrum). But other factors have contributed to the rise as well: Many of us are poorer than we used to be, and poverty is strongly correlated with both how much you weigh and your likelihood of developing certain diseases, like type 2 diabetes.1 We also live with rising levels of chemical contaminants, and researchers are finding more and clearer correlations between exposure to those contaminants, levels of obesity, and levels of diabetes. The main culprits are the so-called persistent organic pollutants—pesticides, PCBs, and other compounds that build up in our food, water, and bodies2—and endocrine-disrupting chemicals (EDCs) like bisphenol A (also known as BPA). For instance, a 2011 study from researchers at the University of California–Irvine found that early exposure to EDCs, which are found abundantly in plastics, canned food, agricultural fungicides, and elsewhere, made mice fat.3 And a number of studies have confirmed links between the prevalence of diabetes and our exposure to persistent organic pollutants and EDCs.4
More of us take psychotropic medications, too: one in five Americans, and more than a quarter of all American women, according to a 2011 report.5 Drugs treating anxiety, depression, bipolar disorder, personality disorders, psychoses, and other mental-health conditions are known to cause weight gain, especially when taken over a period of time.6 I’ve experienced this myself. The first time I went on an SSRI for anxiety, I gained forty pounds over three years. When I went off the medication, I dropped twenty-five pounds within a month, and the rest soon after. Going back on the drugs was a difficult decision for a lot of reasons, and weight gain was one of them. I remember thinking Which is worse, being fat or being crazy? I chose to take the medication again. This time around I didn’t gain as much, but I still gained some, and I’m guessing that weight will stay with me as long as I take the SSRIs. Which is likely to be for the rest of my life.
Some nutrition experts think the low-fat craze of the 1980s contributed, too. Marion Nestle, a professor of nutrition and food studies at New York University, told Frontline a few years back that the emphasis on cutting fat out of foods led to many Americans eating more carbohydrates, which in turn triggered the weight gain.7 (More on that in Chapter 3.) And new research suggests that our long love affair with artificial sweeteners like aspartame, saccharine, and sucralose contributes to weight gain by interfering with the “good” bacteria in our guts, and thus altering our metabolisms.8
Whatever the causes, the rise in our average weight has translated to small gains for some and precipitous gains for a few. “About 20 percent of the population is much heavier than it was, but the majority of the population isn’t much heavier,” explains Linda Bacon, a researcher and professor of nutrition at City College of San Francisco. She suspects that the percentage of people who have gained significant amounts of weight are more vulnerable to whatever combination of factors caused the shift in the first place. But we’re certainly not all destined to become obese, says Bacon.
Humans aren’t the only species that’s heavier. Animals are, too. And while human-related changes in diet and activity levels might explain some of that increase, especially among pets and zoo animals, it doesn’t explain the changes for lab animals, for instance, whose diet and activity levels are closely monitored and documented. Their higher weights can’t be blamed on overeating, sedentary lifestyles, or any of the other causes often mentioned in connection with humans. A 2010 study published in the Proceedings of the Royal Society theorizes a different set of factors at work, including, potentially, environmental toxins, viruses, and epigenetic factors we don’t yet understand.
2. Obesity can take a decade or more off your life.
After charting the prevalence of overweight and obesity, epidemiologist Katherine Flegal began to wonder what her findings meant for Americans’ health. Were more people going to die earlier because they were heavier? To answer that question, she and her colleagues set out to map the relationship between BMI categories and mortality. They expected to find a linear relationship: the higher a person’s BMI, the greater his or her risk of dying prematurely.
But that’s not what they found. Instead, Flegal and her colleagues discovered what statisticians call a U-shaped curve, with the bottom of the curve—the lowest risk of death—falling around 25 to 26 on the BMI chart, making the risk of early death lowest for those now labeled overweight. People considered “mildly obese” had roughly the same risk of dying as those in the “normal” category. Death rates went up for those on either end of the scale—underweight and severely obese—but not by much.
“The differences we’re talking about overall are pretty tiny,” explains Flegal. (Researcher Reubin Andres, who was director of the US National Institute on Aging back in the 1980s, had suggested a similar U-shaped curve, though his was linked more closely with age: the older you are, the less “risky” it is to be heavier.)
As soon as Flegal’s analysis appeared in the Journal of the American Medical Association, the excrement hit the air conditioning. Other researchers claimed her work was shoddy, that she’d left out important data. They said there was no way her results could be accurate. S. Jay Olshansky, a professor of epidemiology at the University of Illinois–Chicago, responded with a journal article arguing that rising obesity rates would shorten lives by two to five years.
That statistic got a lot of attention and helped establish the idea that, as countless media outlets went on to report, for the first time in history a generation of children would have shorter life-spans than their parents. And that prediction is still floating around today, despite the fact that it’s been utterly and thoroughly debunked. As one of Olshansky’s coauthors, University of Alabama–Birmingham biostatistician David B. Allison, sheepishly told a reporter from Scientific American, “These are just back-of-the-envelope plausible scenarios. We never meant for them to be portrayed as precise.”
This type of end-justifies-the-means truth-bending is common in the world of obesity research. Last year, for instance, the National Obesity Forum, an influential lobbying group in the United Kingdom that works on behalf of a long list of pharmaceutical companies, admitted to lying—actually admitted it—in its latest report. The authors had warned that obesity in Britain was continuing to rise, and that an earlier prediction that half the population would be obese by 2050 was “optimistic and could be exceeded by 2050.”9 In actuality, rates of obesity in the United Kingdom, as in the United States, have plateaued or diminished slightly. The group knowingly misrepresented the facts “to reach a wider public,” confessed spokesman Tam Fry.10
Reports like these, which are grounded in opinion and clear agendas rather than fact, feed an increasingly hostile and confusing public conversation around weight. And it’s hard to understand where they’re coming from. I mean, shouldn’t we be glad to hear that a few extra pounds—or more than a few—might not be so bad for you? Might, in fact, even be good for you under certain circumstances?
Given the ongoing backlash, Katherine Flegal decided to clear up the relationship between weight and mortality once and for all. She and her colleagues spent several years meticulously collecting every study they could find with data on weight and mortality—ninety-seven in all. They pooled the data, broke it down in every possible statistical configuration, and published their meta-analysis in early 2013. The results were exactly the same as results from the earlier analysis: being overweight does not increase a person’s risk of dying prematurely, and being mildly obese increases it only slightly.
Flegal’s painstaking methodology didn’t prevent the uproar this time, either. John Wass, an endocrinologist and vice president of the Royal College of Physicians, told the BBC, “Huge pieces of evidence go against this. Countless other studies point in the other direction,” a claim he failed to substantiate. Tam Fry, the above-mentioned spokesperson for the National Obesity Forum in the United Kingdom, commented, “It’s a horrific message to put out. We shouldn’t take it for granted that we can cancel the gym, that we can eat ourselves to death with black forest gateaux.”
Which, of course, isn’t what Flegal’s meta-analysis says at all. It simply maps correlations—in this case, associations between the risk of dying early and BMI. It doesn’t speculate on how to interpret these correlations or why they might hold true. In fact, Flegal takes pains to point out that she never recommends policy or puts out “messaging”; she simply crunches numbers and presents data.
And in any case, surely there’s some middle ground between stuffing ourselves with cake and starving? “Seek moderation in all things,” wrote Aristotle 2,500 years ago, advice that has stood the test of time. Demonizing cake—or sugar, or fat, or carbs—typically leads us to bounce between depriving ourselves (sugar is evil and must be avoided!) and bingeing (I ate a crumb of cake so I might as well eat the whole thing!).
One of Flegal’s biggest critics has been Walter Willett, a Harvard scientist and nutritionist. After the 2013 meta-analysis came out, Willett told National Public Radio, “This study is really a pile of rubbish, and no one should waste their time reading it.”
A month later, he organized a symposium at Harvard with the sole purpose of attacking Flegal’s work. One of his main critiques was that Flegal’s team hadn’t accounted for the confounding effects of smoking. Since people who smoke tend to both be thinner and die earlier than nonsmokers, researchers have to compensate statistically for those effects. One way to do that is to run the data both with and without smokers. Flegal says she did just that and the numbers were virtually identical. And most were very close to a relative risk of 1, meaning that each group (here, smokers and nonsmokers) had the same probability of experiencing a particular outcome—in this case, dying prematurely.
But it’s also true that if you fiddle with statistics enough, you can make them play any tune you want. When I asked Willett to explain his critique of Flegal’s work, he declined to comment, referring me instead to a 2010 study he worked on that found mortality risks lowest among people with “normal” BMIs. That study, it turns out, deleted not only anyone who had ever smoked but also anyone with a history of cancer or heart disease, ultimately eliminating nearly 80 percent of the deaths in the data. Which no doubt explains why Willett’s findings were so different from Katherine Flegal’s.
In his excellent book Naked Statistics, economics professor Charles Whelan explains why, as Mark Twain once commented, “There are three kinds of lies: lies, damned lies, and statistics.” While statistics are rooted in math, and math is an exact science, Whelan points out that statistics are often used to describe complex, multidimensional phenomena, and they can approach those descriptions from all sorts of perspectives.
To put it another way, most people think of statistics as black-and-white, either–or statements: Either being heavier shortens your life or it doesn’t. But life doesn’t work that way, and neither do statistics, which usually describe coexisting truths—what we might think of as this-and-that statements. In this case, being overweight is associated with a longer life and severe obesity is associated with a shorter life and the relative difference between them is small.
You could spin those truths any number of ways. For instance, you could say that obesity takes time off a person’s life. In the broadest possible sense there might be some truth in that statement. But what it doesn’t convey is that (a) it’s only severe obesity that correlates with a shorter life span, (b) being overweight correlates with a longer life span, (c) the effects of obesity on mortality predictions are low overall, and (d) all of these relationships are correlations only; there is no established cause and effect pattern between weight and mortality. (See #3 below for more on correlations.) So even severe obesity (what’s often labeled “morbid” obesity”) may increase a person’s chance of premature death only slightly. If at all.
That last fact is a function of risk factors, something we hear a lot about but rarely understand. In their 2005 book The Obesity Epidemic: Science, Morality, and Ideology, Australian professors Michael Gard and Jan Wright explain that a risk ratio needs to be large to suggest a strong association between a disease and a given variable. Their example: the risk ratio for lung cancer in middle-aged male smokers compared with nonsmokers is 9 to 10, meaning that smokers die from lung cancer at rates nine or ten times higher than nonsmokers. So smoking is considered a significant risk factor for lung cancer. By contrast, the risk ratio for heart disease in people who are overweight or obese is between 1.1 and 2, meaning their risk of developing heart disease is the same as or only slightly higher than the risk for people with “normal” BMIs.
When Doctors Can’t See Beyond Weight
Ray’s daughter, Pattie, fifty-seven, teaches sociology at a college in Las Vegas. She shared this story after Ray’s death.
My father was a mechanic. He did a lot of physical labor, and he was always a small man. When he was sixty-six, he retired and started gaining weight rapidly. When he tried to eat anything, it would make him feel so full he’d be short of breath. He told the doctor he wasn’t eating but was continuing to gain weight. The doctor out-and-out didn’t believe him. He said, “Here’s a diet. Come back in three months.”
Eventually my dad went to the hospital and they pulled twenty-seven pounds of fluid out of his abdomen. It turned out he had hemochromatosis, where your liver doesn’t process iron well. It’s a simple blood test to diagnose, and once they find it they can treat it and you can live a long life.
But because he was sick for a year before he was diagnosed or treated, my dad got liver disease and died eighteen months later. I feel he died of malpractice, but I couldn’t prove he didn’t overeat while he had twenty-seven pounds of fluid on him. I couldn’t even get an attorney to take the case. Our society is so sure it’s all about calories in, calories burned. The relationship between health and weight is so complex, and we try to make it simple. And that hurts people.
Gard and Wright also point out that we are becoming a society obsessed by risk, and by the fantasy that we have nearly infinite power to control it. “Fundamental to such notions is that by naming the risk it can be managed—that is, uncertainty can be reduced—and by understanding the lines of causality, one can act rationally to avoid it,” they write. If that were true, we’d all live forever. Or at least a lot longer.
Because statistics describe relationships that often can’t accurately be reduced to simple statements, writes Charles Whelan, “That leaves plenty of room for shading the truth.” Which is how researchers using the exact same data can—in good faith or not—come up with such different findings. And why reading the research on weight and health can feel like being trapped in an M. C. Escher drawing, where the walls turn into ceilings and water flows upward and where, no matter which path you follow, you wind up going around and around the same closed loop.
Meanwhile, most researchers accept Katherine Flegal’s findings that overweight and mild obesity confer no added risk of dying early, while underweight and severe obesity add a slight risk. For most people, it seems, weight alone is not strongly linked with mortality.
3. Being fat causes heart disease, stroke, type 2 diabetes, and other serious illnesses.
One of the big challenges in considering weight and health is how to distinguish correlation and causation. Both are ways of talking about relationships among sets of events or variables. Correlation is, essentially, an association; this happened and that happened. The events may be unrelated—for instance, your sister gets married and you get a new job. Or one may cause the other—you give birth to a child and your health insurance premiums go up. Or a third variable may influence both of them: people who have yellow teeth are more likely to develop lung cancer, because smoking correlates with both.
Causation, as the term suggests, refers to a cause-and-effect relationship: the gravitational pull of the moon causes the ocean tides. Without the moon’s force, the tides literally would not flow in and out. In medicine, causation is notoriously difficult to prove, since so many variables contribute to human health. Still, some relationships are considered causal. For instance, the fact that smoking causes lung cancer. This doesn’t mean everyone who smokes will get lung cancer, or that only people who smoke will get lung cancer. It does mean that for a certain percentage of the population, smoking will lead to lung cancer.
Mostly, when we talk about weight and health, we’re talking about risk factors, forces that increase susceptibility to a particular disease or condition. All risk factors are not created equal, though we often talk about them as if they were. For instance, both having yellow teeth and smoking are risk factors for lung cancer, but clearly smoking carries a much higher degree of risk than tooth color, which after all can be caused by all sorts of things. So we need to be careful not to conflate risk factors with causes, unless, like smoking and lung cancer, they actually do cause a disease.
One of the underlying assumptions about health is that if only we do everything right, we’ll be healthy. On some level we seem to believe, however illogically, that if we eat right, exercise enough, take care of ourselves in all the ways we’re supposed to, we’ll live forever. Or maybe till we’re 110. But the truth is we’re all going to die of something someday, and whether that something is accident, murder, disease, or plain old age, we can’t control it. We might think we can, or hope we can, but in the end the best we’re likely to do is take care of ourselves as well as possible and hope for the best. Or at least not the worst.
As Michael Gard and Jan Wright point out, it’s human nature to think we can control our destinies by changing our actions. We want to believe we’re in charge. “I am the master of my fate, I am the captain of my soul,” as Victorian poet William Ernest Henley wrote in 1875.11
I’ve experienced this firsthand every time I pick up a package of Cheerios at the grocery store. The label on the box says “May reduce the risk of heart disease,” and even though I recognize the weasel words (“may reduce the risk” rather than “will prevent you from developing”), and even though I know the FDA warned General Mills to change the wording on the label because it’s misleading, still I get a warm fuzzy feeling every time a big yellow box goes into my cart.
In the case of weight and health, dozens of risk factors can come into play, and few if any have been proven to cause any given outcome. That’s why we usually end up talking about surrogate markers like blood pressure and cholesterol levels, measurements that may be associated with true clinical endpoints like heart attacks and strokes but aren’t in and of themselves diseases.
So what do we actually know about the relationship between weight and disease? We know that obesity has been correlated with heart disease, gallbladder disease, and type 2 diabetes, as well as with risk factors for those diseases. The strongest correlation is found between obesity and type 2 diabetes; one 2014 study found that people who were obese but also “metabolically healthy” (that is, their cholesterol and glucose levels were normal) were four times more likely to develop diabetes than metabolically healthy people who were not obese.12 The question is how to interpret that correlation, since many other factors also correlate with type 2 diabetes. (Repeat after me: correlation does not equal causation.)
One aspect of the weight–health relationship that’s rarely mentioned is the fact that definitions and cutoff points for many illnesses and risk factors have changed quite a bit over the last decade. For instance, diabetes used to be diagnosed when blood sugar levels hit 140 milligrams per deciliter (mg/dl); now the cutoff is 126 mg/dl. Suggested treatments include statins and blood pressure medications.13 New disease categories like “prediabetes” and “prehypertension” have also emerged; a blood sugar level above 100 is now considered “prediabetes” and is often treated aggressively. Similarly, hypertension is diagnosed when blood pressure reaches 140/90, and now “prehypertension” begins at a reading of 120/80. There’s been plenty of controversy over these widening disease categories, but whether they’re valid and helpful or not, they skew the perceived medical relationship between weight and health by shifting millions of people into the “ill” category.
One of the most influential studies on heart disease is the Framingham Heart Study, founded in 1948 to identify some of the risk factors for heart disease and come up with prevention strategies. Researchers identified a group of about five thousand middle-aged men and women, all white, all living in Framingham, Massachusetts. Over the last sixty-some years, the lives and health of those men and women have been studied, measured, analyzed, and recorded. The study now includes a second and third generation of Framingham residents, the children and grandchildren of the original batch, as well as two cohorts that are more diverse, also from Framingham.
A lot of the current thinking on heart disease comes from the Framingham data and two other ongoing studies: the Nurses’ Health Study, which follows about 120,000 middle-aged female nurses from eleven states, and the National Health and Nutrition Examination Surveys (known as NHANES), which studies a sample of about five thousand people from around the country. Katherine Flegal uses the NHANES data, which is considered a strongly representative sample, in her analyses.
These three large studies, as well as smaller studies, show correlations of varying degrees between overweight/obesity and health conditions including heart disease, type 2 diabetes, gallbladder disease, stroke, fatty liver disease, and risk factors for those diseases. The strongest findings link BMI and type 2 diabetes, especially for people younger than age fifty-five with BMIs over 40.
That’s where the story ends for many researchers, doctors, and media outlets: with correlation. The trouble is, we still don’t know much (if anything) about causation. Nor do we know what to do with this information on correlation, how to best apply it practically to improving human health. One underlying assumption is that since being obese correlates with a higher risk of heart disease or diabetes for some people, we should all be trying to lose weight. But the research doesn’t support that, for a variety of reasons (see Chapter 2). And that assumption doesn’t take into account contributing factors that might explain those correlations.
For instance, there’s the chicken-and-egg question. We assume that weight gain comes first and causes diabetes and other illnesses. But what if, as surgeon and nutrition researcher Peter Attia has suggested, weight gain is actually an early symptom of diabetes?14 Or what if weight gain and type 2 diabetes are both caused by an unknown third variable? Pediatric endocrinologist Robert Lustig, who’s written copiously about the evils of sugar, suggests that we actually can’t identify which comes first, weight gain or insulin resistance. “Behavior can alter biochemistry, but biochemistry can also alter behavior,” he wrote in a 2008 editorial in The Journal of Pediatrics. No one truly knows which comes first, the illness or the weight gain. What we’ve got right now is a limited correlation, not out-and-out causation, though if you read the news or talk to your doctor about it you’re unlikely to hear this perspective.
Another factor that rarely makes it into the weight–health research is physical activity. Most researchers don’t even ask people whether or how much they exercise. And according to researchers like Steven Blair, who’s been described in the New York Times as one of the country’s leading experts on the health benefits of exercise, that’s junk science. “I continue to get irritated at the establishment in general focusing on obesity and ignoring activity,” says Blair, a professor of exercise science, epidemiology, and biostatistics at the Arnold School of Public Health at the University of South Carolina.
Blair, who’s in his early seventies, resembles Kris Kringle with less hair and more attitude. In other words—in his words—he’s short and fat. He also runs every day for an hour, making him short, fat, and physically active. In the early 2000s, Blair directed the nonprofit Cooper Institute in Dallas, a research group founded by Kenneth Cooper, MD, who coined the term aerobics. Over the years, Blair has produced or worked on hundreds of studies looking at how, exactly, exercise benefits health. His research has convinced him and others that physical inactivity is a much bigger health problem than overweight or obesity. “How successful can we be in getting sedentary people to become active?” he says. “That’s been the focus of my life. Fitness is a powerful indicator of physical health.”
Blair has done much of his work with Paul McAuley, a professor of health education at Winston-Salem University in North Carolina with a PhD in exercise physiology. McAuley spent a while in the corporate world, setting up a fitness program at Sony Pictures. Eventually he wound up back in academia, a few years after cancer epidemiologist Eugenia Calle published a study15 that’s still cited when researchers talk about how obesity is linked to early death. The key word here is linked; Calle’s study showed a correlation between higher BMI and early death, but it didn’t—it couldn’t by definition—show causation.
Paul McAuley sees two big problems with Calle’s meta-analysis. “They failed to control statistically for a major known confounder, fitness,” he says. “That’s invalid.” Given the strong links between fitness, illness, and mortality (the “you can be fat but fit” concept), leaving out physical activity irreparably muddies the findings. Then, too, says McAuley, in the Calle analysis, higher BMI did not predict early death for African Americans. (Most of the research on weight and health has been done only in white populations.) So whatever’s going on in the relationship between weight and mortality is clearly a lot more complicated than fat=death.
In 1999, Steven Blair published research showing that being physically unfit was as much or more of a risk factor for heart disease and death as diabetes, obesity, and other weight-based risk factors. Since then McAuley, Blair, and others have argued it’s healthier to be fit and fat than unfit and thin, while Walter Willett and other researchers continue to insist that fitness and metabolic status can’t possibly make up for the negative effects of obesity.
No Such Thing as a Fat Athlete
Liv, thirty-nine, is a social worker in Portland, Oregon.
I was a fat baby, a fat kid, a fat teen, and I’ve been a fat adult. Having to navigate the health-care system has been continuously a challenge. I would go to the doctor with a cold and he’d tell me, “You need to go on a diet.” I’d go in with an ear infection and it would be, “You need to go on a diet.”
I’m five seven and weigh 350 pounds. I play in a competitive adult kickball league and it’s a lot of fun. A few summers ago I injured my knee, and when I went in the nurse practitioner said, “You weigh 350 pounds, it’s going to be hard on your body.” I told her it felt like more than that. I bike to work four times a week, eight miles round trip. I swim, I do yoga, I’m pretty in tune with how my body feels. She said, “Just ice, elevate, ibuprofen.”
Eventually, twelve weeks after the injury, they determined I had a torn meniscus. The orthopedist said they wouldn’t do surgery even though that’s the only way to fix it, because they only do surgery on kids and people who actively do sports. I said, “I tore this playing kickball.” I told her how active I am. She said, “We only do this for athletes.” So now I can’t play kickball. I was out all season and I don’t foresee being able to play again. It feels like my kneecap is sliding down my shin. And I feel the real reason they won’t repair it is that I’m a fat person and they don’t think I’m really an athlete or someone who deserves to have the surgery done.
There’s one more issue that riles people in the often contentious world of weight–health research: the so-called obesity paradox. (The name reflects the widespread assumption that obesity must be bad for health, ergo any finding that seems to contradict that assumption is paradoxical.) Study after study has shown that overweight and moderately obese patients with certain chronic ailments live longer and do better than normal-weight patients with the same health problems. And many of those problems are the ones most often blamed on obesity, like heart disease, stroke, and diabetes.
One of the earliest researchers to document the obesity paradox, Carl Lavie, says he had a hard time getting anyone to publish his first paper on heart failure. “People thought, this can’t be true, there’s got to be something wrong with their data,” says Lavie, a cardiologist and professor at the Ochsner Heart and Vascular Institute in New Orleans.
For instance, because type 2 diabetes correlates with higher BMIs, doctors typically recommend that patients diagnosed with diabetes lose weight if they’re heavy. That’s what epidemiologist Mercedes Carnethon had been taught, and she had no reason to question it. She’d read about the obesity paradox in people with heart failure and end-stage kidney disease, but she’d dismissed it, figuring it reflected the fact that thinner people were losing weight because they were close to death rather than the other way around.
Then Carnethon, an associate professor at the Feinberg School of Medicine in Chicago, started hearing about an obesity paradox in diabetes. She didn’t believe it at first, but a look into her own data turned up the same effect. “People who were normal-weight at the time their diabetes was identified went on to have a doubling of mortality compared with those who were overweight or obese,” she says. A deep dive into the research confirmed the pattern, though no one yet understands why. Some recent hypotheses: Maybe thinner people actually develop genetic variants of these chronic illnesses that are deadlier than the usual versions. Maybe heavier people get more aggressive treatment because they’re thought to be at higher risk, and so they wind up with better outcomes. Maybe body fat distribution, rather than weight per se, plays a role.
Or maybe the missing piece of the puzzle is something altogether different. Paul McAuley looked at hundreds of studies documenting the paradox and became convinced (unsurprisingly) that fitness was the key. In 2010, he published an article based on data from Stanford’s Veterans Exercise Testing Study, an ongoing study of more than twelve thousand middle-aged male veterans whose fitness levels were documented through lab testing. McAuley’s conclusion: overweight and obese men lived longer than normal-weight men only if they were fit.
Another factor complicating the relationship between weight and health is poverty. It’s no secret that poor people, at least in the United States, are more likely to be fat than wealthier people. All sorts of factors play into this: lack of access to good food and opportunities for exercise (for instance, many kids can’t play outside because they live in dangerous neighborhoods), lack of time for self-care (especially among people working two or three jobs), and stress, which also independently contributes to illness and premature death. American and Canadian research groups have found strong correlations between being poor and developing type 2 diabetes, especially among African American women, regardless of how much people weigh or their BMIs.16 In fact, stigma and stress probably play a much bigger role than we think. In one 2014 study of nearly ten thousand people, those who were unhappy about their weight—whether they were thin, overweight, or obese—were more likely to go on to develop type 2 diabetes, especially if their dissatisfaction went on for years.17 Findings like these illustrate yet again the complexity of the relationship between weight and health. It just can’t be reduced to a sound bite.
Unfortunately, none of this stops doctors and researchers from recommending weight loss for health reasons, even to people diagnosed with diabetes, heart failure, and other conditions where the paradox operates. I asked Mercedes Carnethon if she still counseled her diabetic patients to diet. “We’d never want to back away from weight-loss recommendations,” Carnethon said in surprise, as if I’d suggested that the world was flat. When I pressed her on why, she said, “Evidence does generally show that being in an overweight range is less healthy.” What about Katherine Flegal’s work, I wondered; that suggests otherwise. There was a long pause, and then she said, “Dietary changes that support healthy cholesterol levels are still warranted.” Which wasn’t what I was asking at all.
THE FOURTH often-repeated lie about weight and health is that dieting makes us thinner and healthier. At the very least, we consider dieting benign, something that can’t hurt us even if it doesn’t really help. But the truth is, dieting is actually harmful for many of us for all sorts of reasons. And it doesn’t make most of us thinner or healthier. On the contrary.
*Per projections from “The State of Obesity: Better Policies for a Healthier America,” Trust for America’s Health, www.healthyamericans.org/report/115/.
*Europeans got heavier, too, in the same time period; according to the World Health Organization, 30 to 70 percent of European adults are overweight (that’s a pretty big range), and 10 to 30 percent are obese. See www.euro.who.int/en/health-topics/noncommunicable-diseases/obesity/data-and-statistics for more information.