Pure Pedantry

i-f8e99c6696249bb00b35a68655b3ec72-cuba-flag.gifCastro’s Cuba has seen a precipitous drop in obesity rates and in the deaths associated with cardiovascular risk:

Cuba’s economic crisis of 1989-2000 resulted in reduced energy intake, increased physical activity, and sustained population-wide weight loss. The authors evaluated the possible association of these factors with mortality trends. Data on per capita daily energy intake, physical activity, weight loss, and smoking were systematically retrieved from national and local surveys. National vital statistics from 1980-2005 were used to assess trends in mortality from diabetes, coronary heart disease, stroke, cancer, and all causes. The crisis reduced per capita daily energy intake from 2,899 calories to 1,863 calories. During the crisis period, the proportion of physically active adults increased from 30% to 67%, and a 1.5-unit shift in the body mass index distribution was observed, along with a change in the distribution of body mass index categories. The prevalence of obesity declined from 14% to 7%, the prevalence of overweight increased 1%, and the prevalence of normal weight increased 4%. During 1997-2002, there were declines in deaths attributed to diabetes (51%), coronary heart disease (35%), stroke (20%), and all causes (18%). An outbreak of neuropathy and a modest increase in the all-cause death rate among the elderly were also observed. These results suggest that population-wide measures designed to reduce energy stores, without affecting nutritional sufficiency, may lead to declines in diabetes and cardiovascular disease prevalence and mortality. (Emphasis mine.)

So I guess if you really want to cut obesity in half you should just bankrupt your entire nation.

Glibness aside, I actually have a point for posting this article. I take two things away from this paper:

1) As I have argued before, obesity is not primarily genetic. It is the consequence of an environment and gene interaction, and environment including diet and exercise plays a large role.

When I use the term environment and gene interaction, I think I am still confusing people so let me clarify what I mean. The environment — how cheap food is, how much exercise people get, etc. — determines the average rate of obesity for a society. (This is overwhelmingly demonstrated by the study where the Cuban society has undergone a “shift of the mean” as a consequence of government policy.) Genetics, on the other hand, determine the variation around that mean. If you have a predisposition to obesity and you are living in a society that has a high caloric intake on average, you are very likely to be obese. Likewise, in a society that has a low overall caloric intake, genetics may push you towards the high end of the distribution, but it will still be comparatively low.

2) I don’t know if anyone has looked at this (please tell me if someone has), but I have a theory about how you can combat obesity. To demonstrate my theory, I want to show a graph.

I argue that there is an Inverted U correlation between Wealth and Obesity rates.

i-c9642ad4ddb936bdd5e92d7470526883-invertedu.jpg

At the extreme low end of the wealth axis, obesity rates are low because people are desperately impoverished. Just so people don’t jump on me for this, I am not talking poverty as in the US. The US has poverty, but not in the way that Ethiopia has poverty. We are talking really poor here.

At the extreme high end of the wealth axis, obesity rates fall again because obesity like most things health-related has a socioeconomic status (SES) gradient. This means that the richer you get the less likely you are to be obese.

Now you can debate a couple things about the top end of this graph. First, the factors that determine the SES gradient for obesity are complex. There is undoubtedly an element of education. Also, the relative contributions of race and social class as opposed to actual income are unclear. Second, the steepness of the downturn on the top end of the U is something that is fuzzy. For example, it isn’t clear to me whether if you made everyone richer you could get down to the obesity rates down to as low as you would get in the deeply impoverished.

However, keeping both of those caveats in mind, I think you could still reasonably argue that if you made people wealthy past a certain point you would lower the rates of obesity. Is that a fair assumption? I am curious to see if people agree.

Anyway, why would the existence of an inverted U matter?

Well, it matters when we look at public policy options for fighting the obesity epidemic. Look at it this way. Say we have a group of people in the middle of the U. I would say that this would comprise the majority of poor people in the US — not so poor that they are starving but clearly poorer than the other people in our society.

What policies could we create that would lower the obesity rates for people in the middle of the U? I can think of two.

  • 1) We make food more expensive. We, in essence, simulate poverty for the purposes of food consumption. This does not mean making our economy more like the Cuban one. Instead, we raise taxes on food or take steps to make more expensive so that people will eat less. This may sound like a ridiculous option, but lots of people advocate various brands of libertarian paternalism — a term that I still reject as a contradiction — to fight obesity. They advocate policies such as taxing specific type of high calorie food to make people eat less of them, or they are arguing that we ban certain foods, which will have the indirect effect of making food more expensive.
  • 2) We could try and make the people in the middle of the U richer. We exploit the fact that as people get richer they are more likely to exercise and eat right, that they are more likely to have access to the weight control measures that at present only the very rich have.

Thus far, I haven’t heard many people in the obesity debate argue for the second set of policies which is really a shame because I think that it would work. Further it involves significantly less government intrusion into people’s lives than the alternatives.

The argument I am making for the second set of policies is analogous to the effects of wealth on birth rates. All societies go through a demographic transition as they become more wealthy. This is because people actually have less kids — after a time — as they get wealthier. The most effective means of controlling population is not to restrict the number of births by government fiat; it is actually to help people get rich.

There is a prevailing assumption in some quarters that obesity is a direct and uncontrollable consequence of our society being wealthy, and that as a consequence the only way to fight obesity is to limit people’s access to food by government policy. However, I argue that this has the situation bass ackwards. The problem isn’t that American society is wealthy. The problem is that the wealth has not filtered down sufficiently to the people in the middle of the U. The poor in this and other rich countries have the money to buy high calorie foods but cannot afford access to good food and exercise. You solve that problem by making people richer not by restricting the high calorie foods.

What policies would be most effective then? Pro-growth policies, I guess. You might even argue that American society is going through a sort-of health transition where obesity rates will temporarily soar, but as people get richer will begin to fall again.

Alright, so let’s hear it. I am really curious to hear whether people think this theory has merit. I know that I am speculating, but I have been thinking about this subject a lot lately.

Hat-tip: Hit & Run

Comments

  1. #1 Jorge Gajardo Rojas
    September 26, 2007

    Interesting theory,but population are peoples not a unity.Some eats for hedonistic pleasure without reference of
    wealth.In other countries eating action have cultural roots,you can be obese with a very good food quality or with only carbohidrat or fatty diets.
    I think clima in Cuba is also relevant to have low obesity rates.

  2. #2 Sharon
    September 26, 2007

    “As I have argued before, obesity is not primarily genetic. “

    To be pedantic, that is not what you argued before. You argued for obesity resulting from a mixture of genetics and environment (no dispute there), and in that post you said explicitly that you did not dispute the 70% figure concerning inheritability of obesity.

    Thus “primarily” is an extremely apt word to use, given that 70% is a substantial amount of correlation between obesity and genetics. Whilst technically one can’t assert from this evidence “obesity is primarily genetic”, as in theory there could be some unknown factor correlating even higher than 70%, which would better deserve the description “primary”, the evidence is very strongly pointing to the assertion “obesity is not primarily genetic” being false.

  3. #3 Scott S.
    September 26, 2007

    Neither of your ideas will work. The solution are drugs that curb appetite. Just make it so that people are not interested in eating. A fortune awaits he who finds such a solution. Riches beyond measure, as they say.

    I lost thirty pounds while in the hospital fighting pneumonia, but I don’t recommend that as a widespread solution.

  4. #4 hoary puccoon
    September 26, 2007

    If wealth is largely a stand-in for social status and/or education, generally higher incomes won’t help. I’m wondering, though, if the widening gap between rich and poor in America is one of the driving forces behind the obesity epidemic, as the middle and working classes have less time to cook and to exercise with all adults in a family in the labor force, people feeling behind the 8 ball economically and not putting the time into self care, and so on.

  5. #5 Jake Young
    September 26, 2007

    Sharon,

    Let me clarify something that I didn’t understand when I wrote the earlier post. Basically Razib from Gene Expression explained to me what heritability estimates are actually measuring.

    Heritability estimates — including that 70% figure — are NOT the percentage of genetic contribution to a trait. The represent the VARIANCE of that trait around a population mean that is attributable to genes.

    In this sense, you could have a trait that is 100% heritable, but also have a large environmental component that shifts the MEAN for that population.

    It would work like this. The average BMI for a population is say 25 and genes determine whether you go up or down by 5 or so. If you added a lot of high calorie foods that average might go up to 27. Now the genes still affect the variance, but if you remove those high calorie foods you are going to lower the average back again.

    In this case environment still affects what is a highly heritable trait.

    I’m sorry this wasn’t clear in my earlier post, but I don’t think that I understood it very well then either.

  6. #6 kevin
    September 26, 2007

    I am not talking poverty as in the US. The US has poverty, but not in the way that Ethiopia has poverty. We are talking really poor here.

    I’d like to jump on you for this… :)

    Having lived in a “really poor” country for 2 years, and also seen some “really poor” in the U.S., I think the difference is not what you are implying.

    If you are poor in the U.S., I think mostly you end up eating crap — your only options tend to be preprocessed junk food from walmart. Your chance for exercise goes down, since in the U.S. exercise is a leisure activity for the most part. You do not start subsistence farming in your backyard using hand tools.

    In Ethiopia (assuming it is anything like the West Africa I am familiar with), when you are poor, you start do do manual labor. Either for someone else, or as a subsistence farmer. Cheap labor — I’ve seen desperate people crush rocks for gravel & run large mining operations (all with primitive hand tools), transport goods on their backs or heads for miles day in and day out, and worse. And then you go buy the cheapest food available, which tends to be home-grown, preservative-free, natural foods. Local rice, local fish maybe, perhaps some locally-produced oils, peanuts, or vegetables. But what you don’t do is buy a twinkie for 2 day’s wages.

    -Kevin

  7. #7 Coin
    September 27, 2007

    So I guess if you really want to cut obesity in half you should just bankrupt your entire nation.

    Only at this late date do I understand the true genius of George W. Bush’s economic policies.

  8. #8 Barbara
    September 27, 2007

    Checking my last supermarket bill, I found I paid 2 GBP for a readymeal and 3GBP for some peppers that will form part of a healthier meal. The problem is not that food is too cheap, but that the wrong kind of food is cheap.
    Of course, I like the idea of just making everybody richer, but this is at best a very long-term solution. So in the meantime it’s more useful to make the healthy food cheaper, instead of or in addition to making the high-calorie food more expensive.

  9. #9 Sharon
    September 27, 2007

    “Heritability estimates — including that 70% figure — are NOT the percentage of genetic contribution to a trait. The represent the VARIANCE of that trait around a population mean that is attributable to genes.”

    I am well aware that it represents variance and not the percentage of contribution, as you might have been able to detect by my speculating about the existence of another unknown factor with correlation more than 80%.

    To repeat, 70% is a huge correlation factor. Try trawling through the medical literature and see where you can find a bigger correlation. The large correlation does not provide complete evidence for the assertion “obesity is primarily genetic” but it most certainly provides evidence to suggest that “obesity is not primarily genetic” is untrue.

    To put it another way, let’s try a variation on your example, whereby some trait is 100% inherited (genetics accounts for all the variance) and yet there is a environmental factor that affects to what extent this trait is expressed. If you start describing this trait as not primarily due to genetics, that’s just a crazy description. Kinda like saying “whether an adult has a beard is NOT due to their gender”. Obviously beard-growing is heavily influenced by cultural factors but it is even more strongly influenced by the gene-dictated trait of gender.

    A better description of the beard-growing situation amongst adults is to describe gender as having a strong influence and also culture. Likewise, the best way to describe the factors influencing obesity is by saying that genetics and environmental factors both play a part.

    Though I have to say, whilst I think gender and culture do play a strong role in beards, I’m not very convinced by how much of a role in weight the environment plays, from what you’ve said. It’s like (to take a crude simplistic example to try and make my point) genetics determines the body weight range a person is comfortable at, and the environment determines where within that range they are, EXCEPT that the genetics has much more of an influence on how many pounds someone weighs, and the environment doesn’t seem to be able to have much influence. For example, from the evidence about exercise, I understand that the amount of exercise someone does affects weight by about 4-5lbs, on average.

    It just doesn’t seem to be that big an effect.

  10. #10 Sharon
    September 27, 2007

    “1) We make food more expensive. We, in essence, simulate poverty for the purposes of food consumption.”

    I meant to comment on this one too. I think that would have the opposite effect to that which you intended. If you make food more expensive then you encourage the poorer people who don’t have much money to spend on food towards food that are cheap and more energy dense and more filling. It seems to be a popular opinion that people think that cheap energy dense foods are responsible for obesity.

  11. #11 Jamesaust
    September 27, 2007

    “The problem isn’t that American society is wealthy. The problem is that the wealth has not filtered down sufficiently to the people in the middle of the U.”

    I’m not so sure that this is true. Despite the common tongue-in-cheek saying, the wealthy are not different from the rest of us because they have more money. Bigger bank accounts are in fact the consequence – not the catalyst -of what makes the wealthy different: the host of environmental and behavioral characteristics that are highly correlated with wealth – such as, good parents and grandparents, high quality and quantity of education, hard work, drive for success, socioeconomic-cultural factors among the ‘better classes’ that push them to adopt habits of the well-off including those involving diet and exercise.

    Saying that wealth “hasn’t filtered down” makes it sound like some external power forces money into peoples’ pockets. Give middle class people more money? You’ll get middle class health for people with lots of money. (Millionaire lottery winners may increase their consumption of Courvoisier but it doesn’t follow that they become indistinguishable from others in their elite tax bracket.) If you want people to have more wealth (and per Jake more health), you need to find ways to give them opportunity and incentive to transform themselves into wealthy people. I’d start with the two factors that have been turning the rich into the super-rich over the last few decades: education and technology. Find ways to democratize those tools for the masses and you’ll end the obesity epidemic at the same time.

  12. #12 wj
    September 27, 2007

    While the low end of your inverted U is driven by physical reality (i.e lack of available calories), I’m not so sure about the upper end. In particular, the impact of status-driven (SES) decrease in obesity is very much a matter of relative, rather than absolute, wealth. (An income which would put you up among the very rich in Zimbabwe at the moment, would have you getting food stamps in California.) Which means that the chances of reducing obesity by making the people in the middle of the U richer isn’t going to work. Not that you can’t make them richer, but because their relative wealth won’t change. And so their SES won’t change . . . and their motivation to exercise and avoid obesity won’t change either.

  13. #13 Jake Young
    September 27, 2007

    Sharon,

    So as I understand it (and correct me if I am wrong) you have two issues with my argument.

    One is a nomenclature problem. You object to the statement, “obesity is not primarily genetic.” Would the statement, “obesity is the consequence of a gene X environment interaction,” work? Because I think that is more accurate. I guess that I am always arguing the environment end because I feel like people tend to exaggerate the genetics, but in the future I will try and be more precise with my statements.

    Two, there is the issue of the effect size of changing the environment. My argument that the effect size is large because if you look at people coming from non-Western cultures into Western cultures, there is a BMI change that happens in a couple generations. As the people become more Western in their behavior and have a Western diet, their BMIs increase. There is no way that their genes changed that much in two generations.

    Likewise for the historical argument. BMIs were significantly less 30 years ago.

    I make of these argument to say that environment has a LARGE effect in the gene x environment interaction.

    With respect to the exercise issue, I can see why most people who exercise would only change their weight 4-5 lbs. This is because A) most people’s impression of exercise is pretty limited and B) most people aren’t changing their diets. But I assert that exercise and diet when changed right are sufficient to cause larger changes in weight. Just look at the members of the National Weight Control Registry.

    Jamesaust,

    I think that you make a good point about the fact that the health benefits of being wealthy are not necessarily related to having money. They are indeed related to a lot of behavioral factors that covary with having money but are not the same. So I don’t think that a wholesale redistribution of wealth is likely to work for these reason.

    However, pro-growth policies in the area of trying to get people to make themselves wealthy I assert will still work. These would be more related to encouraging entrepeneurship via tax breaks and subsidizing education. Wouldn’t you say that these would be more likely to confer the behavioral correlates with wealth that also confer good health?

    wj,

    I don’t know if I think that the SES gradient is based on relative wealth vs. absolute wealth, but I think that is a fair criticism. If the SES gradient in wealth is related to income distribution than pro-growth policies would no doubt be ineffective because they will just increase income disparity.

    I frankly don’t know what the effect size would be at the high end of the U. This theory is all speculation there. It is clearly not a big a decline as the lower end. What I would like to see is someone look at how big the effect would be, so that we can make a wise choice about obesity abatement strategies that would cause a minimum of human misery. If deprivation is the only effective means for people to lose weight, so be it. We need to know that. I think that their might be another way.

  14. #14 Jake Young
    September 27, 2007

    Oh and forgot about this…

    I think that would have the opposite effect to that which you intended. If you make food more expensive then you encourage the poorer people who don’t have much money to spend on food towards food that are cheap and more energy dense and more filling. It seems to be a popular opinion that people think that cheap energy dense foods are responsible for obesity.

    I don’t think that making food more expensive will be effective for precisely the reasons you say. That is exactly my point. Taxing food might really backfire.

    But the fact that it probably won’t be effective hasn’t prevented some people from suggesting it.

  15. #15 Mike
    September 27, 2007

    Eliminating subsidies for the sort of cheap, crappy food that most promotes obesity, in particular corn and sugar, really is libertarian.

  16. #16 brian gulino
    September 28, 2007

    An alternate theory for the right side of the graph, where people can afford to overeat.

    The ability to postpone gratification is a trait which correlates with wealth in capitalist societies. This same ability may cause people to adopt a healthier lifestyle. To be specific, lower middle class people are more likely to order a pizza and video from blockbuster on a Friday night while upper middle class people will make a salad at home and read a book.

    The trait that made them upper middle class, to postpone gratification, to stay in school, to save money, to work overtime, makes them more inclined to forgo the pizza.

    If this is the case, then giving poor people more money is unlikely to change their eating habits.

    By the way, I don’t buy the “poor people can’t afford healthy food” argument. Potatoes are cheaper than potato chips, ground beef is cheaper than beef jerky, vegetables are cheaper than cookies, and nobody needs to drink soda.

  17. #17 Michel Phillips
    September 28, 2007

    On relative v. absolute wealth: When the Food Stamps program began in the mid-1960s, malnutrition was a real problem among the American poor. Today a bigger problem among the American poor is obesity. The SES (relative wealth) of the American poor, however, has not changed. This at least shows that more absolute wealth will move people from the bottom of the “U” to the middle.

    On the efficacy of government programs: We have a large-scale experiment proceeding with this right now, but there’s no good monitoring of data, and no control group. Corn sugar is heavily subsidized. School menus are subsidized by the U.S. Department of Agriculture — with priorities dictated by the farm lobby — NOT by the Department of Health and Human Services, which might prioritize nutritional needs instead. Compared to the 1960s and 1970s, a much smaller percentage of young people take physical education in school. Many elementary schools used to have recess twice per day, but now don’t have it AT ALL — and fear of lawsuits has resulted in removal of swings, monkey bars, and teeter-totters from most playgrounds. Rising asthma rates and crime means more poor children play indoors. Factors affecting children and adults: video games are bigger business than movies; TVs now have several hundred channels; and then there’s the internet. The shift of the economy from a manufacturing & agricultural base to a services base means a smaller percentage of jobs require physical exertion. If an experimenter or social engineer wanted to impose environmental conditions to CAUSE obesity, isn’t this pretty much what he or she would do? And wouldn’t our real-world obesity data lead this experimenter to conclude his environmental modifications were a raging success? Of course, some of these environmental factors are government-controlled and some are not. But shouldn’t we want the government programs pushing AGAINST the other obesity-causing environmental factors, and not WITH them? At any rate, the co-occurrence of all these known environmental factors with the skyrocketing of obesity rates leads me to view with great skepticism the theory that we shouldn’t worry about policy changes because obesity is due to personal choices.

  18. #18 Sharon
    September 28, 2007

    Jake:
    One is a nomenclature problem. You object to the statement, “obesity is not primarily genetic.” Would the statement, “obesity is the consequence of a gene X environment interaction,” work? Because I think that is more accurate.

    Yes that would work better (assuming that “environment” could be made to cover everything that wasn’t genetic, which I think it can). You could also say that “obesity is not solely genetic”, that would work.

    I guess that I am always arguing the environment end because I feel like people tend to exaggerate the genetics, but in the future I will try and be more precise with my statements.

    I think precision is good, so that the impression given isn’t too far off in either direction.

    It’s certainly a very interesting question to consider what factors affect obesity. From the literature and elsewhere I’ve seen strong evidence for both genetics and environment being factors. Some of the twins studies are extremely convincing for genetics being a strong factor, and it is also known that some people do have the ability to have a significant effect on their weight, even if only temporarily. And of course the studies about immigration are very suggestive of it being in some way a mixture: that is, how the genetics respond to a different environment.

  19. #19 Sharon
    September 28, 2007

    Jake wrote:
    As the people become more Western in their behavior and have a Western diet, their BMIs increase. There is no way that their genes changed that much in two generations.

    Indeed. It’s clearly either all environment, or interaction between how the genes are expressing themselves differently due to a different environment.

    It is a very interesting question: what in the Western environment is causing this increase in weight? (Funnily enough, no-one seems to get worked up about height increases.) The theory I have is that this is influenced by the dieting industry and the general mindset of trying to lose weight. If trying to lose weight has a collective effect like forcefully throwing a tennis ball to the ground – as opposed to, say, dropping it – so that the rebound effect is higher than the original starting point, then this theory is certainly plausible. The timescale matches. The mentality of Western society matches. It even matches the life expectancy increasing, as people have been attempting healthy lifestyle improvements even if ultimately they didn’t manage to lose weight and keeping it off.

    There are case studies to show this kind of thing happening on an individual level, but the question is whether it happens to such an extent that this is the answer to the question for society in general.

    By the way I would like to say thanks for a good place for discussion. I always enjoy a nice polite thoughtful discussion on an interesting topic amongst people who care about truth and details, irrespective of whether they hold differing or similar views. On the topic of obesity in particular, one rarely gets that kind of thing!

  20. #20 Sharon
    September 28, 2007

    Jake:
    With respect to the exercise issue, I can see why most people who exercise would only change their weight 4-5 lbs. This is because A) most people’s impression of exercise is pretty limited and B) most people aren’t changing their diets.

    The 4-5lbs figure has come from studies where the amount of exercise was measured and significant. Such studies show a pretty huge variation, interestingly enough. Some people lose a whole bunch of weight, some barely change, and some gain. It seems to be very individual. But overall the average amount isn’t that much.

    But I assert that exercise and diet when changed right are sufficient to cause larger changes in weight.

    This is one of these statements which can’t be proved untrue even if it was wrong. If someone does manage to make a large change in their weight through diet and exercise then it supports the statement, and if they make the changes and still don’t manage to make the weight change then it can be argued that they didn’t do the change “right”.

    Just look at the members of the National Weight Control Registry.

    As for the National Weight Control Registry, the existence of that body rather argues for the opposite point. People who have managed to lose weight and continued to keep it off for a significant amount of time are so difficult to find for the purposes of studies that they had to go to special trouble to set up a weight control registry for the purpose. And even then, I hear that they dropped some of the requirements so not such a big weight loss was needed to be listed in the registry.

  21. #21 Caledonian
    September 28, 2007

    When I use the term environment and gene interaction, I think I am still confusing people so let me clarify what I mean. The environment — how cheap food is, how much exercise people get, etc. — determines the average rate of obesity for a society. (This is overwhelmingly demonstrated by the study where the Cuban society has undergone a “shift of the mean” as a consequence of government policy.) Genetics, on the other hand, determine the variation around that mean.

    Wrong.

    You don’t quite seem to be grasping that it’s not the average caloric intake of the society that’s relevant, but the average caloric intake of the individuals.

    “How much exercise people get” also isn’t part of the environment.

  22. #22 Lance
    September 28, 2007

    People at the far end of the wealth curve have the time to exercise and they eat the most diverse menu, comparatively speaking. People in the great middle eat more diversely than their parents, but they work less physically, though still to the point where they are too tired to adequately exercise. Eating is a form of entertainment in our culture. Like going to the movies, wrestling, tv, all sedentary occupations. Eating has to do with ego and with our place in the world. We may not be rich, but we can eat fatty foods and drink good booze. Surely in a large enough pool there will be those predisposed to gaining weight, but americans gain weight across the board because we sit in front of screens and we go out to dinner at restaurants that want us to come back so they make their product fatty and salty. Will we solve this? Never. People will just get fat and die. Something has to kill them. It used to be cigarettes. Cars kill teenagers. Fats kill a large percentage of us. Better dead than ever more people taking up the space. I say this as someone pushing sixty, so I know I am part of the baggage myself.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.