Tuesday, July 16, 2013

The Genetics of Obesity, Part III

Genetics Loads the Gun, Environment Pulls the Trigger

Thanks to a WHS reader* for reminding me of the above quote by Dr. Francis Collins, director of the US National Institutes of Health**.  This is a concept that helps reconcile the following two seemingly contradictory observations:
  1. Roughly 70 percent of obesity risk is genetically inherited, leaving only 30 percent of risk to environmental factors such as diet and lifestyle.
  2. Diet and lifestyle have a large impact on obesity risk.  The prevalence of obesity has tripled in the last 30 years, and the prevalence of extreme obesity has increased by almost 10-fold.  This is presumably not enough time for genetic changes to account for it.
To explore this concept further, consider two hypothetical examples.  If everyone in Sameville lives in the exact same way and eats the exact same food, environmental variability will not contribute to body fat variability at all.  Genes will be the only major source of variability, and therefore the heritability of body fatness will approach 100%.  Obese people in Sameville almost always have obese children, and lean people almost always have lean children.

If each person in Differentville lives in a radically unique way and eats radically unique food, environmental variability will be huge, and the heritability of body fatness will be lower, perhaps 25%.  High environmental variability means genetics gets a smaller share of the total variability.  In Differentville, obese parents often have lean children and lean parents often have obese children.

As this example illustrates, estimates of heritability only apply within a specific environmental context.  Within the US and within many other affluent countries, most people live in a relatively similar manner (as opposed to comparing the environment of the US vs. Uganda).  If you don't live like the typical 21st century American, and don't eat like the typical 21st century American, then there's no reason to believe that obesity heritability estimates derived from 21st century American studies will apply to you.  By living and eating significantly better than average, many people can escape their genetic tendency toward obesity.

Genes Determine Susceptibility

If genes are so important, how do we explain the fact that numerous traditionally-living cultures around the world remain lean throughout life, with very few overweight individuals, even if they have sufficient food availability, yet these same groups readily become obese once they modernize?  How do we explain the fact that obesity prevalence has increased three-fold in the US since 1978?  My view is that genes don't determine a person's absolute level of body fatness, but they do determine a person's degree of susceptibility to a fattening environment.  The best example I've seen of this phenomenon comes from a blood pressure study conducted in traditional and urban women in New Guinea (Maddocks et al. Med J Australia. 1:1123. 1967).

The following graphs show systolic blood pressure on the horizontal axis, and the percentage of the population at a given blood pressure on the vertical axis (this type of graph is called a histogram).  The top graph represents young and older women from a traditionally-living horticulturalist group in the New Guinea highlands***.  What you can see is that nearly all young women cluster in the healthy range between 100 and 120 mm Hg.  Although the curve representing older women is flatter, it's similar in shape and shares nearly the same peak and mean value.  This is a traditionally-living population that maintains a healthy blood pressure as it ages.  

The bottom graph is from a genetically similar population living in the port city of Hanuabada, where trade has brought in Western processed foods and lifestyles.  What you can see is that the curve for young women looks quite good: it's sharply centered around a healthy blood pressure just as it is in the highlands group.  However, in older women, the curve flattens out considerably and develops a pronounced rightward "tail" representing serious hypertension.  Cardiovascular mortality is probably extremely high at the right end of that tail.  Yet despite the industrial diet and lifestyle, many older women on the left side of the graph maintain a healthy blood pressure!

In a natural environment, genetic differences did not result in very much blood pressure variability (i.e., nearly everyone clustered around a healthy value), and hypertension was uncommon.  In an unhealthy environment, genetic susceptibility presumably started to matter more, population blood pressure variability increased with age, and a large fraction of women developed life-threatening hypertension.

The example above focuses on blood pressure, but essentially the same phenomenon has been demonstrated for obesity (1).  Below is a similar graph of the US population, with body mass index (BMI; a rough measure of body fatness) instead of blood pressure on the horizontal axis.  Each line represents a different time point, beginning with NHANES II (1976-80) and ending in 1999-2004.

You can see that although the difference is less pronounced than the the blood pressure example above, the curve "flattens out" somewhat over time, producing a longer rightward "tail" of obese and extremely obese individuals, and a higher mean value.  Note that the US in 1978 was not quite a "natural environment" by historical standards.  The NHANES II curve has likely already flattened considerably compared to the ancestral population, which would probably exhibit a sharp peak centered around a BMI of ~22, with very little rightward stretch into the obese category.  Also, the curve has shifted even further to the right since 2004, as the prevalence of obesity has increased.

Below is a crude estimate of what the curve probably looked like in the ancestral population (green) prior to industrialization, based on my knowledge of currently existing populations living a subsistence lifestyle.  Leanness is most common, overweight has a modest prevalence, and real obesity is rare but not totally absent.  I've also estimated what the curve looks like in the US today (purple).  I want to be perfectly clear that the colored lines are not real data; I drew them in by hand to illustrate a concept.

You can see that the separation between curves is more pronounced when we consider the entire transition period between the ancestral subsistence lifestyle (green) and the contemporary industrial lifestyle (purple).  Leanness went from typical to atypical, yet there are plenty of lean people even today.  This suggests that when the environment is not fattening, obesity susceptibility genes rarely cause obesity.  When the environment is fattening, susceptibility genes start to matter more, and those who carry them become overweight and obese.  Population BMI variability increases, and the BMI curve flattens out and shifts rightward.

This also suggests that people with a genetic tendency toward obesity would most likely remain lean(ish) in the ancestral environment, and I believe that replicating key aspects of that environment can help us sustain leanness today.


Here are the main points to walk away with:
  1. In a healthy environment, genes alone do not usually cause obesity and lifestyle-related diseases, and population variability in these characteristics is low (i.e., most people are lean and free of these diseases).
  2. In an unhealthy environment, genetically susceptible people become obese and/or develop disease, while others remain lean and healthy because they are not genetically susceptible.  
  3. This results in an increase in population variability (e.g., more variability in body fatness between individuals).
  4. Diet and lifestyle environment have a major impact on obesity risk.  For a genetically susceptible person, maintaining a leaner state is usually possible but it requires stepping out of the typical fattening environment.  Genes are not destiny.

* I lost your e-mail.  Send me a message if you'd like attribution.

** Reader Unknown remarked that this quote was used as early as 1920 by Dr. Elliott Joslin.

*** The diet was primarily sweet potatoes.


Unknown said...

Another interesting post, and especially love the quote re: gun and trigger. Apparently the first person to use it was Dr. Elliott Joslin in the 1920s. He used it in reference to diabetes, but I've also seen it in reference to heart disease and eating disorders. Not trying to be confrontational! As a PhD student, I'm a little obsessed with giving credit where it's truly due. Not sure how to find the original source.

polyhex said...

Excellent post, as always. I love seeing posts from you in my reader!

Margaret Anne said...

Great Post.

Robert said...

Thanks for the very informative and succinct post, Dr. G. I know a few people who have not made environmental changes to prevent/reverse obesity even after losing a loved one to an obesity-related condition like heart attack or cancer. These people are very smart and aware of the risks yet apathetic. This seems to suggest that top-down environmental changes are likely needed to prevent obesity from ever occurring. I'm wary that these changes will occur in our lifetimes, however, and I fear for the future.

Willow said...

Interesting post.

Anonymous said...

Great post, very good discussion of the matter. I would only hope this serves to clear up the apparent contradiction in the minds of those who first hear about the very high heritability of obesity.

That said:

"This also suggests that people with a genetic tendency toward obesity would most likely remain lean(ish) in the ancestral environment, and I believe that replicating key aspects of that environment can help us sustain leanness today."

Good luck with that part.

Adam Noel said...


Although I do agree there is a lot of difficulty in the task of changing our environment I do think reproducing certain elements of our ancestral conditions could be easier then others.

Even improving sleep quality for most of the population may help with appetite regulation and result in weight reduction. Of course this will not bring people down to leanness but for some people it may do the trick.

Simple decisions like eating at home twice a day as opposed to once a day could reduce food reward and may result in an increase in leanness.

At the same time I think the act of dieting (as commonly done) is counter-productive as we live in an incredibly fattening environment and dieting will likely just result in weight regain.

Dawn said...

Stephan: I am a longtime follower of your blog, your voice remaining one of the very few I still listen to after being dazzled then disillusioned by low-carb, paleo/primal, "wheat belly," etc. Your mention of replicating our "ancestral environment" prompts me to ask you if you have read "An Epidemic of Absence" by Moises Velasquez-Manoff. Premise: agriculture didn't ruin us; it was the Industrial Revolution, urbanization, and modern medical approaches that completely changed our natural microbiota for the worse. He describes research in which people are being intentionally infested with hookworms and whipworms and seeing their allergies disappear and autoimmune diseases such as MS go into complete remission. The point is that we traditionally had symbiotic relationships with microbes and parasites that beneficially affected our bodies in many ways, one of which was weight regulation. I would really appreciate a comment from you regarding this idea, or even better an entire blog post reviewing the book.

Stephan Guyenet said...

Hi Dawn,

I haven't read that book, sounds like it might be interesting. I agree that the Industrial Revolution was the deleterious dietary shift that's most relevant to our health concerns today.

I use the term "ancestral" though I recognize that it can be a squirrely term. What I mean when I use that term is a subsistence lifestyle where most people in a community are primary producers of food. That covers hunter-gatherers as well as non-industrial agriculturalists.

Regarding parasites, symbiotic bacteria and weight regulation. Yes it's probably true that certain bacteria and helminths, in the right context, have health benefits due to our long-co-evolution with them. That being said, helminths and certain bacteria can clearly be devastating to health in the wrong context.

I'm looking forward to more research on the potential health benefits of low-level helminth infection. For the time being, there is a bit of evidence that they may reduce GI inflammation and lower circulating cholesterol, but the question needs more work.

I gave a talk on gut health (including microbiota) and the metabolic syndrome at AHS12. I plan to narrate the slides and post them on YouTube at some point.

Robert said...

Speaking of modern environmental triggers of obesity, what are your thoughts on BPA, or other plastic-based compounds?

Observational data in kids and adults suggests an association b/w BPA in urine and obesity though this association could simply be confounded by processed food consumption as the authors realize. Also, some data showing nearly universal weight gain among lab animals over the past few decades, despite being under strict dietary and env. controls, has led some to speculate that plastics or other env. triggers might have a role (I'll have to dig up this source).

Thank you for your thoughts, Dr. Guyenet.


Kids - JAMA

Kids- China, though only in girls???


Anonymous said...

The question remains, what is it about the environment that is so fattening. I am sure this was extensively covered in previous blog posts. Is it the proliferation of high food reward, constant snacking on high carbohydrate, highly industrialized "foods", sedentary patterns of activity, environmental pollutants, sleep deprivation and/or light exposure at night, or the simple imbalance between energy input into the system and energy output from the system that is implied by many of the above? How and why has food become so addictive ... it's really amazing to think about!

Jane said...

'The question remains, what is it about the environment that is so fattening.'

The only thing that seems to correlate with the obesity epidemic is Thatcherism. Could that be a clue? It made people feel helpless. We know that if rats have no alternative to sitting around eating fattening food they get fat, and if they are given an exercise wheel they no longer find the fattening food more palatable than their normal chow. They stop getting fat even if they don't use the wheel very much. Apparently they just need to know there is an alternative, one that makes them feel as good as eating does.

'The act of voluntary wheel running reverses dietary hyperphagia and increases leptin signaling in ventral tegmental area of aged obese rats'

glib said...

The subtitle at the beginning of this post (what loads and what fires the gun) is particularly appropriate. IMHO, the reward thing is what the medical class pushes to make money, but surely sedentary patterns, air conditioning, and low nutrition (highly processed) foods are major factors.

In my particular case, due to my research, I spend summers at different places. At one place they keep the temperature around 80F, everyone is in shorts, and I drop ten pounds in about 6 weeks. At my University, where I keep a sweater in the office, it is only 4 or 5 lbs drop (in 10 weeks) below my winter weight. I do walk and bike offsite, but here I do the same, and I have a garden and orchard to tend. Air conditioning makes you hungry.


"My view is that genes don't determine a person's absolute level of body fatness, but they do determine a person's degree of susceptibility to a fattening environment." Absolutely agree. I find that one sentence to be about as simple and precise as can be. The fact that many people may find adapting ancestral habits "difficult" does not mean the solution lies elsewhere. Like trying to emulate success, you won't be successful if you don't put forth the effort. Wishing for success, without a concerted and appropriate effort will work as good as wishing for leanness, or looking for the next "magic pill."

Stephan Guyenet said...

Hi Glib,

You said "the reward thing is what the medical class pushes to make money". Huh? Who is the "medical class", what is the "reward thing" they are "pushing", and how exactly is it making them money?

If you mean that researchers are writing grants to study reward mechanisms, and getting them funded because they're judged to be scientifically compelling by expert peer review, that's a funny way to put it.

Unknown said...

Thanks for another great post. It seems that genetics can replace god for some people in terms of filling in blanks, if that makes any sense. I'm somewhat tired of hearing people pass over things that they can't explain by simply passing it off as genetic destiny. It's certainly important, but so is environment.

I'd be extremely interested to see your talk on health and microbiota. It's a subject I've become increasingly interested in recently, but information on the subject around the internet can be... shall we say, somewhat suspect. I'm fairly convinced that the microbes in our body have an important role in our health, but I'm still trying to find good information on the subject though.

Anonymous said...

Colin Macdonald:
There is a good review article by Dove Press authored by Ian Spreadbury on this subject in a nutrition journal. Google Spreadbury, microbiota, queens university, kingston, ontario.

W.R.T. air conditioning making you fat, 70% of TBW is H20. One could easily lose 5 lbs by being chronically dehydrated, or even just acclimated to a different TBW. This often happens when people are put on diuretics, though they tend to regain their TBW (=total body weight). I suppose innate thermogenesis could also play a role here?

Michael Cronin, Washington, DC said...

Excellent summary. Those are line graphs, btw. Histograms use bars/rectangles to display data. And many thanks for last year's post on your work with Dr Attia's

Caroline said...

Hi Stephan, what do you know about how genes change across generations? If a mother is overweight or eats badly when she is pregnant, does she pass on worse genes to her children? If a person has one lean and one obese parent is one gene more dominant and therefore more likely to be passed on? Thanks for your blog!