by Kenneth W. Krause.
Kenneth W. Krause is a contributing editor and “Science Watch” columnist for the Skeptical Inquirer. Formerly a contributing editor and books columnist for the Humanist, Kenneth contributes frequently to Skeptic as well. He can be contracted at email@example.com.
Why, sometimes I’ve believed as many as six impossible things before breakfast.–The Queen, to Alice in Through the Looking Glass.
Wouldn’t it be splendid to have our cakes and eat them too? Arguably, both ideology and popular culture allow their followers to do just that. Until they don’t, of course. At that point, when facts and logic can no longer be denied, the rudely awakened find themselves confronted with difficult choices.
The concept of healthy obesity, for example, has gained much traction during the last fifteen years. At one end of the continuum, members of the popular but clearly flawed “Healthy at Every Size (HAES)” movement profess the nonexistence of excess adiposity and suggest that even the most obese people can lead perfectly healthy lives (“Every size”—really?). On the other end, and somewhat more credibly, others allege the existence of an “obesity paradox” and a “metabolically healthy obesity.” Such are the tantalizing subjects of this column.
Cardiologist and obesity researcher, Carl J. Lavie, has described the paradox as follows: “Overweight and moderately obese patients with certain chronic diseases … often live longer and fare better than normal-weight patients with the same ailments” (Lavie 2014). In addition to his own research, Lavie’s conclusions are based on a revolutionary and, in some circles, much-celebrated JAMA study led by Katherine Flegal at the US Centers for Disease Control and Prevention, who reviewed 97 studies of more than 2.88 million individuals to calculate all-cause mortality hazard ratios for standard body mass index (BMI) classifications (Flegal et al. 2013).
Flegal’s team reported as follows: Relative to normal weight, all combined grades of obesity were associated with an 18 percent higher incidence of all-cause mortality. In cases of more extreme of obesity, the association rose to 29 percent. By itself, however, the mildest grade of obesity was not correlated with a significantly elevated risk, and the overweight but not obese category was actually associated with a 6 percent lower incidence of all-cause mortality. Predictably, the popular media quickly seized on the overweight population’s presumed appetite for these tempting results.
Metabolically healthy, or “benign,” obesity, on the other hand—which Lavie dubs the “ultimate paradox”—appears to have no standard definition or list of qualifying criteria, but is often characterized generally as “obesity without the presence of metabolic diseases such as type 2 diabetes, dyslipidemia or hypertension” (Munoz-Garach et al. 2016). Retained insulin sensitivity, however, is the hallmark trait of this subpopulation. Researchers have assigned up to 32 percent of the obese population to this phenotype. It applies more prevalently to women than men, but is thought to decrease with age among both sexes. Researchers have yet to determine whether these obese are genetically predisposed to decreased risks of disease or mortality. But their existence, along with that of the metabolically unhealthy normal-weight population, suggests that factors other than excess adiposity are at play.
All of which might sound at least somewhat comforting to the now 600 million obese worldwide (and still growing) who have been told for decades that obesity per se will significantly increase one’s susceptibility to heart disease, stroke, cancer, diabetes, and arthritis, for example. Preferences and popular reports aside, however, it appears we may yet be forced to choose between possessing our cakes and consuming them, because an impressive body of new and well-conceived research has called both the paradox and healthy obesity into serious question.
Consider, for example, a truly enormous international meta-analysis published last July in The Lancet by the Global BMI Mortality Collaboration (GBMC 2016). Led by Harvard professor of nutrition and epidemiology, Frank Hu, this study poured over data from more than 10.6 million participants who were followed for up to 14 years. 239 large studies conducted in 32 countries were included. Importantly, the Collaboration attempted to control for a “reverse causation bias,” in which low BMI was the result, rather than the cause, of an underlying or preclinical illness by excluding current or former smokers, those who suffered from chronic disease at the study’s inception, and those who died during the initial five years of follow-up. In other words, Hu’s team addressed the potential for potent confounders that Flegal’s team, for lack of data, was forced to ignore.
The Collaboration’s results were startling. Interestingly, Hu “was able to reproduce [Flegal’s results] when conducting crude analyses with inadequate control of reverse causality, but not when [he] conducted appropriately strict analyses.” In the end, then, the Collaboration found that, worldwide, participants with a normal BMI in the 22.5 to 25 range enjoyed the lowest risk of mortality and that such risk increased significantly throughout the overweight and obese ranges. In fact, every five units of BMI in excess of 25 was associated generally with a 31 percent greater risk of premature death—specifically, 49 percent for cardiovascular-related, 38 percent for respiratory-related, and 19 percent for cancer-related mortality. According to Hu, his team had succeeded in “challeng[ing] previous suggestions that overweight and grade 1 obesity are not associated with higher mortality, bypassing speculations about hypothetical protective metabolic effects of increased body fat in apparently healthy individuals.”
Consider too, a large prospective cohort study published last October in the BMJ in which about 115,000 participants—free of cardiovascular disease and cancer at baseline—were followed for up to 32 years (Veronese et al. 2016). Evaluating the combined associations of diet, exercise, alcohol consumption, and smoking with BMI on the risk of all-cause and cause-specific mortality, this study was also designed to address Flegal’s peculiar 2013 results. A lead author here as well, Frank Hu first noted, once again, that previous examinations suggesting an obesity paradox, including Flegal’s, had allowed for potentially confounding bias by failing to distinguish between healthy normal-weight individuals and a “substantial proportion of the US population” in which “leanness is driven by other factors that can increase risk of mortality,” including existing or preclinical chronic diseases and smoking.
Contrary to the alleged paradox, Hu discovered that when lifestyle factors were taken into serious consideration, the lowest risk of all-cause and cardiovascular mortality was enjoyed by participants in the slightly low-to-normal, 18.5 to 22.4 BMI range—that is, when those subjects also displayed at least three out of four healthy lifestyle factors, including healthy eating, adequate exercise, moderate alcohol intake, and no smoking. In the end, according to Hu’s team, “the U-shaped relation between BMI and mortality observed in many epidemiological studies is driven by an over-representation in our societies of individuals who are lean because of chronic metabolic and pathological conditions caused by exposure to smoking, a sedentary lifestyle, and/or unhealthy diets.” The optimal human condition, in other words, is not overweight of any kind or to any degree, but rather “leanness induced by healthy lifestyles.”
So much for the obesity paradox, at least for now. But what of its somewhat less voracious cousin, the notion of metabolically healthy obesity?
Recognizing prior support for so-called “benign obesity,” a trio of Canadian diabetes researchers led by Caroline Kramer conducted a systematic review and meta-analysis of eight studies evaluating over 61,000 subjects—many of whom were classified as metabolically healthy obese—for all-cause mortality and cardiovascular events (Kramer et al. 2013). When all studies were considered, regardless of follow-up duration, the healthy obese subjects displayed risks similar to those of healthy normal-weight participants. However, when considering only those studies that followed-up for at least ten years, Kramer and colleagues discovered that the purportedly healthy obese were significantly more likely than their normal counterparts to perish or suffer serious cardiovascular trouble.
Should we infer, then, that the healthy obese are, in fact, healthy until circumstances render them otherwise a decade later? Not according to Kramer. Regardless of metabolic status, she warned, even in the short term, obesity is associated with subclinical vascular disease, left-ventricular abnormalities, chronic inflammation, and increased carotid artery intima-media thickness and coronary calcification. In the end, the Canadians found no support for the “benign obesity” phenotype and declared with no uncertainty that “there is no ‘healthy’ pattern of obesity.”
Most recently, however, a diverse and impressively creative group of Swedish scientists used transcriptomic profiling in white adipose tissue to contrast responses to insulin stimulation between never-obese, unhealthy obese, and, again, supposedly healthy obese subjects. (Ryden et al. 2016). Led by Mikael Ryden at the Karolinska Institutet, this group revealed, first, clear distinctions between the never-obese and both groups of obese participants, and, second, nearly identical and abnormal patterns of gene expression among both insulin-resistant and insulin-sensitive obese subjects, independent of other cardiovascular or metabolic risk factors.
Said Ryden during a post-publication interview: “Insulin-sensitive obese individuals may not be as metabolically healthy as previously believed.” (ScienceDaily 2016). His team’s findings, he continued, “suggest that vigorous interventions may be necessary for all obese individuals, even those previously considered … healthy.”
To Lavie’s credit, he generally acknowledges obesity’s proven hazards. He also recognizes serious and consistent exercise as the most reliable strategy for attaining and maintaining good health. Far less defensible, however, is Lavie’s insistence that exercise can render obesity a benign condition. First, as much of the research presented here demonstrates, the chronic diseases strongly associated with obesity are, by definition, progressive and apt to cause damage down the road. Second, in the real world, excess adiposity always leaves meaningful exercise a far more difficult and, thus, far less likely prospect.
Obese or not, our health continues to be undermined by the popular, ever-emotion-manipulating media, the misguided and oppressive forces of political correctness, and, most crucially, our own subjective prejudices and appetites. But as their numbers continue to swell, the overweight and obese grow increasingly vulnerable to seductive messages inviting self-deception and failure. As in all other contexts, their liberation from these influences derives only from an unflinching appreciation for the methods of science—that is, empiricism, rationality, candor, and the assumption of responsibility for individual experimentation. In a word, skepticism.
Flegal, K.M., B.K. Kit, H. Orpana, et al. 2013. Association of all-cause mortality with overweight and obesity using standard body mass index categories. Journal of the American Medical Association 309(1): 71-82.
Global BMI Mortality Collaboration. 2016. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. The Lancet 388: 776-786.
Kramer, C.K., B. Zinman, and R. Retnakaran. 2013. Are metabolically healthy overweight and obesity benign conditions? Annals of Internal Medicine 159(11): 758-769.
Lavie, Carl J. 2014. The Obesity Paradox: When Thinner Means Sicker and Heavier Means Healthier. NY: Plume.
Munoz-Garach, A., I. Cornejo-Pareja, and F.J. Tinahones. 2016. Does metabolically healthy obesity exist? Nutrients 8: 320.
Ryden, M., O. Hrydziuszko, E. Mileti, et al. 2016. The adipose transcriptional response to insulin is determined by obesity, not insulin sensitivity. Cell Reports 16: 2317-2326.
ScienceDaily. 2016. More evidence that “healthy obesity” may be a myth.” 18 August 2016. https://www.sciencedaily.com/releases/2016/08/160818131127.htm>.
Veronese, N., L. Yanping, J.E. Manson, et al. 2016. Combined associations of body weight and lifestyle factors with all cause and cause specific mortality in men and women: prospective cohort study. BMJ. DOI:10.1136/bmj.i5855.