Genetic variants for liver fat accumulation and circulating triglyceride levels

Abstract: Non-alcoholic fatty liver disease (NAFLD), obesity and type 2 diabetes(T2D) are strongly associated and today their prevalence reaches epidemic proportions. NAFLD has a multi-factorial aetiology, both genetic and environmental factors are equally important. The aim of this thesis was to investigate genetic factors associating with NAFLD and circulating triglyceride (TG) levels in overweight or obesity and their interaction with diet, as well as their association with lipid and glucose traits, and incidence of T2D and CVD. In total 417 obese Finnish subjects and >5000 subjects from the population-based prospective Malmö Diet and Cancer Study (MDCS), with data on genotypes, lipid traits and diet, were included. Two apolipoprotein c3 (APOC3) variants did not associate with liver fat content or apoc3 concentrations among Finns, which was in contrast to results in a previous study. Overweight modified the association between the PNPLA3 rs738409 and TG and ALAT levels (Pinteraction=0.003 and 0.03, respectively) and we additionally observed that carbohydrate (CHO) and fat intakes, in particular sucrose among normal-weight individuals, and possibly the ω-6: ω-3 PUFA ratio in overweight individuals, may modify association with TG levels (Pinteraction=0.03 and 0.08, respectively). Importantly, our results suggest that the mechanism by which PNPLA3 variant increases liver fat accumulation may depend on BMI status and dietary intake. The variants in MLXIPL and GCKR genes, were associated with lower TG and higher glucose traits in MDCS-CC study, and a combination of TG lowering alleles in MLXIPL and GCKR was strongly associated with lower TG levels (P=2.4-12) and higher fasting glucose and HbA1c levels (P=0.009, for both). The same alleles associated with lower risk of hypertriglyceridemia at the baseline (P=0.015 for MLXIPL, P=0.00025 for GCKR and P=0.000012 for combined alleles), and the MLXIPL T-allele was associated with lower incidence of hypertriglyceridemia in the prospective analysis (OR: 0.37; 95%CI: 0.20-0.70; P=0.002).These loci did not associate with the risk of T2D, although a nominally significant interactions between total CHO intake and MLXIPL and between fiber intake and GCKR on T2D risk was observed (Pinteraction=0.043, Pinteraction=0.007, respectively). Analysis of combined genotypes reflected the results observed in the interaction analyses for the individual genotypes. Additionally, five NAFLD susceptibility loci did not show any association with T2D, CVD or total mortality, except for that nominally significant associations between the GRS and increased incidence of T2D (P=0.03) and death (P=0.048) was observed among severely obese (BMI≥35kg/m2) individuals. In conclusion, this thesis suggests putative role of obesity and dietary CHOs in modifying the genetic susceptibility to NAFLD, which may facilitate further studies to elucidate the mechanisms and factors behind increased liver fat accumulation and potential new therapeutic approaches.