November 30, 2010

How much would exercising matter to me for weight-loss, given my genomic markers?

By: Douglas S. McNair, MD, PhD

li_fig_1 Well, quite a lot, actually. The article just published by Shengzu Li and Ruth Loos and colleagues at the University of Cambridge (link and pdf below) ought to provide strong encouragement to start or maintain your exercise plan. The study was conducted over a number of years and involved more than 20,000 people. It is absolutely solid and actionable, in terms of readiness to put into practice.

Changes in lifestyle, including excessive caloric intake and insufficient exercise, have caused a dramatic increase in the prevalence of overweight and obesity.

If you decrease intake and increase your energy expenditure through exercise, that logically *should* reduce body mass index (BMI) and risk of metabolic syndrome (precursor to Type 2 diabetes). But epidemiological studies have established that genetic factors also contribute to how big your risk of obesity is, and how likely it is that exercise and dieting will work in keeping weight in-check. Genetically-predisposed individuals are more susceptible to obesity in an "obesogenic" environment, which is what Li's paper is about.

They genotyped 12 SNPs (single nucleotide polymorphisms) in obesity-susceptibility related genes in a population-based sample of 20,430 individuals aged 39 to 79 years. Physical activity level was assessed using a self-administered questionnaire that categorized activity into 4 levels. They found that each additional BMI-increasing SNP allele was associated with 0.154 kg/m2 increase in BMI (equivalent to 445 grams in body weight for a person 170 centimeters tall). This association was far more pronounced in inactive people (0.205 kg/m2; 592 g in weight) than in active people (0.131 kg/m2; 379 g in weight).

• rs3101336 in NEGR1 gene on Chr. 1
• rs10913469 in SEC16B (LZTR2) on Chr. 1
• rs6548238 in TMEM18 on Chr. 2
• rs7647305 in ETV5 on Chr. 3
• rs10938397 in GNPDA2 on Chr. 4
• rs925946 in BDNFOS on Chr. 11
• rs10838738 in MTCH2 on Chr. 11
• rs7132908 in FAIM2 on Chr. 12
• rs7498665 in SH2B1 on Chr. 16
• rs1121980 in FTO on Chr. 16
• rs17782313 (or proxy rs10871777) in MC4R on Chr. 18
• rs368794 (or proxy rs29941) in KCTD15 on Chr. 19

The direct-to-consumer (DTC) genomics testing companies don't tell you anything about these, but the genotype results for them are right there in your raw results file download.

For your interest, I've prepared a little Excel tool where you can enter your genotype values for each of the SNPs above. The spreadsheet calculates your risk of obesity using the information in the Li article. It also calculates how much you are likely to decrease your BMI by, if you change your activity level from whatever it is now to 'active' (according to the definitions used in the Li paper). Click on the link below and download the spreadsheet and play with it, to see how big the changes are for different genotype values. If you've had your SNP testing done and have your results in your raw file download, go ahead and enter them and see what the model predicts for you.

[NOTE: The 'expected' decrease in BMI with increasing activity corresponds to the 20,000 people studied. Your own 'mileage' may vary, as they say, depending on how much you increase your energy expenditure through exercising, how much your food intake is, etc. That said, the expected value does represent a realistic and reliable estimate, based on what actually happened for those 20,000 people studied.]

If you have questions or suggestions, please add a comment below or send me email at dmcnair@cerner.com. Thanks!