Genetic Determinants of obesity in relation to diet, weight gain and mortality
Obesity is one of the major health concerns that has reached epidemic proportions globally. It is generally believed to be a result of interactions between genetic and environmental factors. In this thesis we investigated the role of dietary factors in modifying the genetic susceptibility to obesity (papers I to III), studied the association between genetic susceptibility to obesity and weight gain at different time-points in life (paper IV) and tried to dissect the causality between cardiometabolic traits and mortality (paper V). The work in this thesis was conducted using data from the population based prospective Malmö Diet and Cancer Study (MDCS; N= ~30,000) and the Gene-Lifestyle interactions And Complex traits Involved in Elevated disease Risk (GLACIER; N= ~5000) cohorts. In paper I, we did not observe any evidence for macronutrient, fiber or total energy intake in modifying the genetic susceptibility to obesity when genetic susceptibility was represented as a Genetic Risk Score (GRS) based upon 13 BMI associated genetic variants. In individual SNP analyses, after correcting for multiple comparisons, some of the individual obesity loci such as NEGR1 rs2815752 associated with fat, carbohydrate and fiber intakes (P≤1x10-4 for all) and BDNF rs4923461 interacted with protein intake on BMI (Pinteraction=0.001). In paper II, pooled analyses of MDCS and GLACIER suggested 0.16 (SE=0.04) kg/m2 increase in BMI (P=8x10-5) in the lowest quartile of GRS (comprised of 30 BMI-associated genetic variants) for each increment in category of sugar-sweetened beverages (SSB) intake vs. 0.24 (SE=0.04) kg/m2 higher BMI in the highest GRS quartile (P=1x10-7). We also observed evidence for the role of SSB intake in modifying the genetic susceptibility to obesity (Pinteraction=0.049). In paper III, a copy number variant (CNV) in the salivary amylase gene (AMY1) did not associate with obesity traits neither in men nor in women (P>0.05 for all). However, upon stratification by dietary starch intake, BMI decreased with increasing AMY1 CNV in low starch intake group (P=0.035) and increased with increasing AMY1 CNV in the high starch intake group (P=0.04) among females. These results suggest a putative role of starch intake in modifying the association between AMY1 CNV and obesity in women (Pinteraction=0.041). In paper IV, a GRS based on 31 BMI-associated genetic variants was associated with increased annual weight change (β=0.003 kg; SE=0.01; P=7x10-8) and increased odds for substantial weight gain (OR=1.01; 95% CI= 1.00-1.02; P=0.013) per risk allele from young to middle age in MDCS. However, the GRS was associated with decreased annual weight change (β=-0.005 kg; SE=0.002; P=0.002) and decreased risk for substantial weight gain (OR=0.97; 95% CI= 0.96-0.99; P=0.001) per risk allele during and after middle-age in the pooled analyses of MDCS and GLACIER. These results suggest a paradoxical inversed relationship between genetic susceptibility to obesity and weight gain during and after middle age compared to increased weight gain in younger age. In paper V, observations from multivariable Mendelian randomization analyses suggest a direct causal association of TG (P=0.017 and P=0.028) and an inverse association of HDLC (P=0.049 and P=0.005) with total- and cardiovascular mortality, respectively. In conclusion, the results from this thesis suggest a role of specific dietary factors in modifying the genetic susceptibility to obesity and that genetic variation affect weight gain differently at different time-points in life but the underlying mechanisms need to be further understood. Additionally,our findings points towards causal associations between TG and HDLC and mortality which can help to devise better treatment strategies in clinical practice.