Skip to main content

Diet, genes and Harvard conservatism

{My new book "Ever seen a fat fox ~ Human obesity explored" which has a chapter and section on diet and genes is due out mid May}
The Harvard School of Public Health has a recent web article on genes and obesity bearing the somewhat negative title: “Genes are not destiny” [1]. From my standpoint, this review is in fact defensive of a School where genomics appears to be low on their nutrition research agenda. This brief review of the Harvard position is written by a blogger who works both in public health nutrition but also in molecular nutrition.

The overview (hardly a review) begins by pointing out that obesity conditions linked to a single gene defect such as the Prader-Willi Syndrome are utterly rare. Equally, common ordinary everyday obesity is linked to so many genes that looking for one is like looking for a needle in a haystack. However, needles can be found in haystacks and we now know that high blood pressure is dramatically reduced in individuals marginally deficient in the B-vitamin, riboflavin, when given dietary riboflavin supplements but this applies only to those carrying a single variant of a gene we all have, the natural variant of the folic acid-related FTHFR. That variant isn’t rare. One in 10 of us carry it[2]. So lets not throw the baby out with the bath water as they say.

  Equally, the web article goes on to make the case that modern gene analysis technology allows us to explore how multiplesets of genes might explain common obesity. The more such complex sets of genes explain obesity, the fewer people we find with such a unique blend of genes and even at its best, this technology can explain only a few percentage points of the obesity problem. So far I am in full agreement with the Harvard team and I remain somewhat in agreement when they write:

“So if our genes have stayed largely the same, what has changed over the past 40 years of rising obesity rates? Our environment: the physical, social, political, and economic surroundings that influence how much we eat and how active we are. Environmental changes that make it easier for people to overeat, and harder for people to get enough physical activity, have played a key role in triggering the recent surge of overweight and obesity.”

Its here I start to divert from the Harvard dons’ views. According to the US Center for Disease Control[3], about 1 in 3 adults is classically obese while the same figure for children (2-19 years) is 1 in 6 (35% and 17% respectively).  The figure rises if we consider overweight but obesity is the sharp end of the problem in the Harvard analysis. Now if 1 in 3 US citizens are obese, then 2 out of 3 are not. They share as stated above, the same “environment: the physical, social, political, and economic surroundings that influence how much we eat and how active we are”.

So why isn’t everyone obese? Why some, not others. The answer is of course genes and we’ve known this for years. The Harvard review pays simple lip service to twin studies, the extensive literature on which tell us that obesity is about 70% heritable (only 20% in non-identical twins who do not share an exact genome as identical twins do).  In their simple reference to twin studies, the Harvard Dons reveal a deep prejudice when they write:
“The strong correlation for BMI between monozygotic twins and its attenuation with lesser degrees of shared genes suggest a strong genetic influence on BMI. However, this conclusion is based on the assumption that identical and fraternal twins have the same degree of shared environment-and it’s an assumption that may not hold in practice.”

The literature is peppered with outstanding scientific papers, old and new, which this overview sweepingly ignores. They completely fail to refer to the studies of twins raised apart for periods of almost 50 years in radically different environments and still showing an obesity heritability of >70%. They completely fail to refer to metabolic studies where identical twins were subject to 1000 additional calories per day[4]or a reduction in caloric balance through bouts of intense cycling[5]. Take for example the former. Some twin pairs showed a dramatic gain in weight with the extra calories while others didn’t. Different genomes predispose more or less to obesity. However, whatever one identical twin did, the other followed exactly. With weight loss experiments the same conclusions held.

In summary, we do know of some instances where a very common single gene variant can determine dietary responsiveness to a complex phenotype such as high blood pressure. We do know that identical twins show a 70% heritability of obesity (this exceeds the heritability of height, alcoholism and depression for example) and we know that identical twins raised apart for very long periods also show a very high obesity heritability[6]. Metabolic studies also explain variation in obesity heritability both between and within identical twin pairs. So, the case for optimism is huge and not so futuristic, as the Harvard dons would suggest.
One of the main flaws in the diet-obesity-gene link is the constant belief that it is to reductionist biology that we must turn to understand this problem. Metabolic pathways of fat and carbohydrate metabolism including fat cell, liver cell and muscle cell biology dominate. For me, the future is the genetics of food choice. Why are some people sensitive to the external cues to food intake – its sight, smell, ambiance, availability, time of day, etc? Why are some apparently more self restrained, less sensitive to genetic cues? Applying advanced neurobiological techniques to study neurobiological determinants and responses to food cues are the start.  I would then move to twin studies to rank the heritability of such neurobiological indices of appetite. In time I’d find more and more examples of heritable neuronal regulators and her I’m talking complex neuronal patterns. I don’t need to go to genes at this stage.  Moreover, I’d be looking not for determinants of individual behaviour but membership by individuals of clusters of neuronal networks as we currently apply to metabolomics.

[2]Mahmud N et al (1999); Gut 45:389-394 

[4] Bouchard, C et al. (1990) ‘The response to long-term overfeeding in identical twins’. The New England Journal of Medicine, 322:1477–82.
[5] Bouchard, C et al. (1994) ‘The response to exercise with constant energy intake in identical twins’. Obesity Research, 2: 400–10
[6] Stunkard, A. J et al. (1990) ‘The body mass index of twins reared apart’. The New England Journal of Medicine, 322: 1483–7.


Popular posts from this blog