Regular readers of Obesity Panacea will know that one of my favourite topics (and the focus of my PhD) is the health impact of sedentary behaviour. Last fall computational biologist Larry Parnell left a comment linking to a recent paper of his exploring how genes modify the health impact of sedentary behaviour. The paper was fascinating, but since genes aren’t my area of expertise I wasn’t sure that I could do it justice on my own. Luckily, Larry put me in touch with the study’s lead author Caren Smith, who did a fantastic job of explaining the study in terms that non-geneticists like me can understand! Below are my questions and Caren’s answers.
1. Could you briefly summarize the findings of your study in lay-terms?
This study, and other “gene-environment interaction” studies are based on growing awareness that our health status is based not only on our genes or only on our environment (diet, activity level, smoking, alcohol), but on the ways in which these two components “interact” with each other. For example, depending on genetic makeup, some people may be more sensitive than others to the effects of physical inactivity. The current study examines the effects of physical inactivity (quantified as computer time and television time or ‘screen time’) on HDL (the ‘good cholesterol’). Depending on the version of a particular gene that a person carries, inactivity may be more or less likely to adversely affect (eg, lower) the individual’s HDL.
2. What is a genetic polymorphism?
A genetic polymorphism is a variation of a specific gene at a specific location in the DNA. The word polymorphism means ‘many shapes’, reflecting that at a given position, there are different varieties of a particular segment of DNA . Although overall we (as humans, and even as higher primates) are much more genetically similar than we are different, at least 10 million polymorphisms have been identified in people. Some of these are associated with health-related traits such as diabetes, heart disease or lipid levels, although our understanding of the consequences of polymorphisms or even how many exist across all of the world’s people is still quite limited.
3. Your results showed a significant interaction between screen time, genotype and HDL-cholesterol levels in women, but not men. Why do you think the interaction was only seen in women? Does this mean that men don’t need to worry about screen time, at least when it comes to HDL?
The reasons for our detection of the interaction only in women are unclear. We know that HDL differs, overall between men and women, with women having higher HDL on average. Several earlier (non-genetic) studies reported similar relationships between television viewing and risk of disease which were limited to women. These differences could be related to gender-specific biology, but could also be related to broader issues, such as differences in the ways men and women spend their leisure time, or what kind of work they perform inside or outside the home.
With respect to the issue of exempting men from any concern about screen time and HDL, we need to emphasize that we examined a single polymorphism in a single (relatively small) group of white individuals living in the US Midwest. Our observations are therefore incomplete, and may not be generalizable to all individuals. Further, HDL is a ‘complex trait’ which is determined not only by a single gene and a single exposure (eg, physical activity) but in fact by many genes and many other exposures including smoking, drinking alcohol and body weight. Other genes (we only studied one) may be more important in men.
4. What is the clinical significance of the relationship between screen time, genotype, and HDL cholesterol that you observe in your findings? Should people be worried about their genotype in addition to the amount of time that they spend being sedentary?
When we think about clinical significance of a genetic finding, we often look at the amount of difference between the two genotype groups, and/or between the two environment groups (low and high inactivity, in this case). In the current study, women with high screen time and the ‘higher risk’ genotype had an average HDL of 48 whereas women with the same genotype but low screen time had an average HDL of 57. A difference of 9 units for HDL would be considered clinically relevant, since even an increase of 1 unit will substantially reduce risk of cardiovascular disease. With respect to worrying about genotype vs. worrying about sedentary behavior, at the moment we know more about the overwhelmingly beneficial effects of physical activity than we do about the effects of interactions with certain genotype. While it is likely that some people derive greater benefits than others from exercise (or conversely, at put at higher risk through inactivity), the vast majority of people would benefit from increasing their overall physical activity.
5. This paper examined endothelial lipase, but other work in animals suggests that sedentary behaviour also has a dramatic influence on lipoprotein lipase activity. For example, this study suggests that 6 hours of muscle inactivity reduces LPL activity by roughly 50% in animal models. Could the polymorphisms examined in this study be related to LPL activity as well as endothelial lipase?
The specific polymorphism that we examined was in the gene that ‘codes’ for the enzyme called ‘endothelial lipase’. That is: this gene contains instructions which are used to construct this enzyme, specifically, and not other enzymes. However, you bring up an important point which is that many other enzymes (such as LPL ( lipoprotein lipase)), or proteins within metabolic pathways are affecting HDL level. It is also possible that an individual with the endothelial polymorphism might also have polymorphisms in the genes encoding any of these other components, and that these polymorphisms could affect HDL or any other health-related trait, either directly or through interaction with an environmental exposure. The full range of polymorphisms which affect even the single trait of HDL is unknown.
6. What is the next step as far as research is concerned?
The best way to continue research in this area would be to carry out an ‘intervention trial’. In this kind of study, people volunteer to enroll in a controlled study for a specified period of time and to agree to the terms of the study. For example, the design of the study might specify periods of activity and periods of inactivity (or record these very precisely using a device that quantifies movement), and also control for other factors which influence HDL (see the next answer below). Before enrolling in the study, people would be tested (genotyped) to see whether they have the polymorphism that we studied, to make sure that there are similar numbers of people with and without the polymorphism to compare. The intervention trial is a more reliable design for establishing ‘causality’ – that is, for being more certain that the effects we see are due to gene-screen time interactions, and not to some other (unknown) factor.
7. What is the take-home message for people who want to improve their HDL cholesterol levels? [feel free to speculate wildly here 🙂 ]
For most people reading this interview, the single most important factor affecting HDL is body weight or waist size. The majority of people in the US are overweight or obese. HDL levels are very sensitive to extra weight which tends to decrease HDL levels, especially if the weight is concentrated in the abdomen. The same is true for another lipid in the blood called triglyceride, which increases in response to extra weight. The combination of high triglycerides and low HDL is associated with a pre-diabetes condition called ‘metabolic syndrome’ which affects more and more people in the US. Physical activity helps in the maintenance of a healthy body weight, leading not only to healthier HDL but also to an overall prevention of metabolic and other diseases.
Other lifestyle factors which affect HDL are smoking and drinking alcohol. Smoking decreases HDL (perhaps for some people more than others, which may depend on their genes). Cigarette smoking also increases triglycerides, so that improving blood lipids represents yet one more reason to stop smoking. Moderate use of alcohol can increase HDL, although given the risks associated with overuse and abuse of alcohol, recommendations which encourage consumption of alcohol should be considered with advice from a physician.
8. Is there anything else you’d like to add?
This kind of study illustrates concepts which are intuitively familiar to all of us – that there are some people who live long lives despite unhealthy behaviors and others who die prematurely while engaging in similar or even less risky behaviors. We conclude that the one who died early (or suffered disability from stroke or diabetes) carried some sort of ‘genetic risk’. The goal of studies like this one is to identify and quantify the components of this risk, so that we might predict, for example, who is most likely to develop diabetes or experience heart disease as a result of weight gain or inactivity. In the case of physical activity, we know it’s beneficial to virtually everyone, but people experience these benefits in different ways, and to different degrees. Some of us will be “ okay” with the extra 30 minutes of screen time, and others should really put that time(and more) into walking.
Thanks again to Larry Parnell for linking to the study and passing along my questions to Caren, and to Caren for breaking down the study so eloquently!
And for more computational biology (and who doesn’t need a bit more computational biology?!?), be sure to check out Larry’s blog Variable Genome.
Smith, C., Arnett, D., Tsai, M., Lai, C., Parnell, L., Shen, J., Laclaustra, M., Junyent, M., & Ordovás, J. (2009). Physical inactivity interacts with an endothelial lipase polymorphism to modulate high density lipoprotein cholesterol in the GOLDN study Atherosclerosis, 206 (2), 500-504 DOI: 10.1016/j.atherosclerosis.2009.03.012