A recent article in the UK Daily Mail featured the “Weight of the World” chart made by Visual.ly. It graphically represents the average body mass index (BMI) of the inhabitants the countries of the world. In keeping with my recent series on the matter, I wanted to see what it’d look like if I turned it into a map:
This is the result. The colors refer to average BMI as shown, rounded up (based on data gathered from the World Health Organization, WHO). For the U.S., I’ve used the White average BMI, and I’ve drawn data for Britain from here, and Canada here.
Now with this map, I’ve got to say: come on man, has no one really ever considered that perhaps there is a genetic explanation for (at least part of) this pattern?
The average BMI numbers are roughly clustered by ethnic group. While we can see the effects of poverty and food scarcity across Africa and much of South and (perhaps) Southeast Asia, the overall pattern is fairly striking.
This can be further highlighted by the following charts. This is a chart of all the countries sorted by BMI, with their ethnoracial clustering shown:
We can see some clear trends. The East Asian nations cluster near the top in the low BMI category. At the other end, we can see a clear cluster made by the Anglosphere and their British progenitors. Also clustering high are many Latin American countries. The general pattern follows East Asians at the top, followed by the Eastern European countries, then the Mediterraneans, the Germanics/Scandinavians, and finally the Latin Americans and the Anglo/Celtic world. The Middle Eastern/North African nations are segregated with the more impoverished ones near the top and the oil-rich ones near the bottom.
Here the countries again, sorted by ethnoracial group:
Within each group, there aren’t many wild variations and ethnic categories are fairly consistent. France seems to form a cluster with the Low Countries (and perhaps Denmark), so I’ve grouped some of them together. Interestingly, with the Scandinavian countries, it is Iceland (with its significant Celtic ancestry) and Finland (whose people are perhaps closer to their hunter-gatherer ancestors) that score highest.
It’s impossible to escape the observation that daughter countries, such as those in the New World, cluster with their ancestral countries in Europe. Despite all the alarm over rising obesity rates, perhaps the various peoples of the world are just genetically fated to pack on the pounds in a modern, calorie-rich environment.
No country illustrates this better than perhaps multi-ethnic Canada:
These maps track the variations in the obesity rates across the various Canadian provinces. As we’ve seen before, there is a distinct pattern, with the Québécois routinely being fairly light, and the Maritime and Indigenous provinces being significantly heavier. The Anglo-Germanic provinces in the center come in somewhere in between. The ethnic pattern is striking. The Maritimes are heavily Celtic in origin (particularly Scottish), and apparently take on weight like their brethren across the Atlantic. Native American groups are known for their propensity towards obesity. British Columbia, with nearly a quarter of its population being Asian, leans towards the light end.
A similar pattern is visible in the United States:
This is the CDC’s state-level obesity rates, broken down by race. It is clear that most of the obesity statistics that are commonly given are misleading because there are clear racial distinctions. The White obesity rate is significantly higher in the Scotch-Irish states of Greater Appalachia.
As we’ve seen before with cardiovascular disease, obesity also follows a clear pattern, one which suggests that it has genetic roots. In hindsight, it is silly anyone would not at least entertain that possibility, but as Greg Cochran so eloquently ranted recently, people refuse to acknowledge biology when it comes to humans. We know heredity is heavily responsible for differences between individuals in a group (with one’s BMI being 80% heritable, as heritable as height), and it is clearly considerably responsible for the racial differences seen above. That heredity may be responsible for differences between ethnic groups within a race, say between the English and the French, should be at least considered a distinct possibility.
This may mean that for the Anglosphere countries, with whose anti-obesity campaigns I’m most familiar, the levels of obesity seen in modern populations may be endemic and ultimately intractable.
Of course, it’s not clear that reducing obesity is even a worthwhile goal. For one, I’ve yet to see any evidence that the problems that people commonly believe that obesity leads to, such as the aforementioned cardiovascular disease and “premature” death are in fact caused by the obesity itself (and not by say common genetic factors for both obesity and heart trouble). For one, in the course of gathering this data, I’ve found that the association between obesity and cardiovascular disease isn’t all that strong: obese people are at less than double the risk for cardiovascular mortality during the course of most studies. This is fairly weak for a wholly correlational link, and the link may in fact be weaker still, since many studies rely on people to self-report their BMI, and people tend to underreport such measures.
The other fear with obesity is that the obesity rate will continue to shoot up indefinitely, and that soon there will be no thin, able-bodied people left in the population. This of course is also patently silly. Not only is the increase in the obesity rate plateauing in the developed world, there is no reason to expect that it would do anything different. Unless the environment continues to change in the relevant ways, at some point, everyone genetically susceptible to becoming obese will do so, leaving the rest of the population as thin as ever.
All this runs squarely against the hopes and aims of those who wish to propagate a “fat-shaming” campaign with the goal of getting people to lose weight. While genetics are at play in generating individual and group differences, obesity has indeed changed over time. This is apparently due to unclear environmental changes over this time, probably broadly related to increased food availability and decreased physical activity (of course, in exactly how this has happened is still far from clear). Proponents of “fat shaming” think that by ridiculing or otherwise pressuring overweight people, they can get them to make behavioral changes that get them to lose weight; as if it was that simple. Needless to say, not only this idea unproven, all evidence lies squarely against the broad effectiveness of dieting strategies. Even low-carb & “paleo” diets can’t be all that helpful for most people, for if they were, the obesity “problem” would have been tackled by now.
In that vein, interestingly, as I was gathering evidence for this post, the announcement came that Samoa Airlines was to begin charging its passengers by weight. This is move long called for by many as a great move in the spirit of fat shaming. Now, the Samoans are no small people, with body types such the one shown being quite common among them. And yet, all that extra weight seems not to impact them as much as one might think, as evidenced from their cardiovascular mortality rate, which is comparable to that of the U.S.
Indeed, the link between cardiovascular mortality and obesity between groups worldwide seems quite questionable.
I think that serious attention needs to be paid to genetic factors in investigating variation in obesity around the world, perhaps for resolving mysteries like the “French Paradox“. Of course, there are caveats in the data I’ve presented here. I have no idea how the WHO gathered this data, so I can’t vouch for its accuracy. As well, BMI is an imperfect measure of fatness, since there is a great variation in the typical builds of different peoples across the world. Hence, these numbers need to be taken for what they’re worth. And clearly, heredity isn’t the whole story, as there are marked differences between say France and Quebec in terms of obesity and cardiovascular death rates (11.3% and 40/100,000 vs 21% and 191/100,000, respectively). And genetic factors include not only physiological ones, but things like taste response and behavioral factors as well. Indeed, a critic might level that the observed ethnoracial associations don’t represent heritable physiological or even behavioral factors, but shared dietary customs. Perhaps that’s so, but I think the connections are too strong to dismiss, especially given the role heredity is known the play in individual differences in obesity. In any case, more research is needed, and to quote Simon Baron Cohen, “don’t forget about biology.”