Do Genetic Variations Predict Physical Activity Response To Lifestyle Intervention Among Obese Adults With Diabetes?

Abstract

PURPOSE: Those who are prone to high physical activity (PA) in natural environments may respond better to a PA promotion intervention than those who are prone to low PA. A genome-wide association study (GWAS) and a candidate gene study identified 4 single nucleotide polymorphisms (SNPs) related to PA: rs978656 near DNAPTP6, rs10887741 near PAPSS2, rs7279064 near C18orf2, and rs6265 near BNDF. We hypothesized that the 4 SNPs will predict greater change in PA phenotypes in response to a lifestyle intervention. METHODS: This is a secondary analysis of Look AHEAD, a multi-center randomized controlled trial among participants who are overweight/obese and have type 2 diabetes (ages 45-76). Look AHEAD is designed to test the health benefits of an intensive lifestyle intervention (ILI), combining calorie restriction and PA promotion for weight loss, as compared to diabetes support and education alone. We examined the moderating effects of the 4 SNPs individually and in a weighted genetic score (GS). Of the 3649 participants who were successfully genotyped for all the 4 SNPs, we examined only those cases that also completed Paffenbarger PA questionnaire (PPAQ) and cardiorespiratory fitness (CRF) measures at baseline, year 1, and year 4 (n=2675). We used linear mixed effects models to regress PA phenotypes (PPAQ and CRF) on genetic variations, time, intervention, as well as interactions between the three. Models controlled for age, sex, body mass index, ancestry principal components (for population stratification), and study sites and included a Bonferroni correction for multiple testing. Additive, recessive, and dominant models were tested. RESULTS: None of the individual SNPs or the GS were associated with baseline CRF or PPAQ. The rs978656 interacted with time (year1) and intervention (p=0.04), such that the main effect of the intervention on CRF was significant (p=0.04) only among A allele carriers (less-PA-prone) at year 1; however, this finding did not persist following Bonferroni correction (alpha<0.006). GS was not predictive of change in CRF or PPAQ. Individual SNPs were not predictive of change in PPAQ. CONCLUSIONS: The ILI may have a more salient effect on CRF among A allele carriers of rs978656. Future intervention studies on the genetic basis of PA change are recommended to include more GWAS-identified SNPs.