Genetic Variations in Type 2 Diabetes and Cardiovascular Disease: A Focus on Gene-Lifestyle Interactions and Mendelian Randomization

Sammanfattning: Popular Abstract in English The prevalence of type 2 diabetes (T2D) and cardiovascular disease (CVD) is increasing in epidemic proportions around the globe. This epidemic accounts for huge health and economic burdens as a leading cause for morbidity and mortality. The development of these diseases is very complex and involves both genetic and lifestyle factors. The shift to sedentary lifestyles and increased caloric intake leading to obesity along with cigarette smoking are the main culprits behind this epidemic. The genetic component contributing to the etiology of these conditions is also evident due to familial aggregation and differences among ethnic groups. Gene-lifestyle interactions are also believed to be an important factor in the etiology of these diseases. Understanding gene-lifestyle interactions is believed to be important and may contribute to the understanding of complex diseases as T2D and CVD. In simple terms interactions exist when the magnitude of the association between a certain lifestyle factor and the disease changes among people based on their genetic background. Understanding interactions may help us in identifying biologic mechanisms through which lifestyle factors and genetic factors affect the risk of disease. This can open the door into identifying better drugs for treating and preventing these diseases. Studying interactions may also help us to personalize prevention and treatment strategies in people according to their genetic background. This discipline is still relatively new and much work is still needed to understand interactions. This thesis aims to investigate gene-lifestyle interactions in T2D and CVD using the strongest genes previously identified to associate with these diseases. We have studied the Malmö Diet and Cancer Study that includes more than 30,000 individuals. We have observed interactions between the strongest T2D genetic variant (TCF7L2) and dietary fiber intake influencing the risk of this disease. Previous studies have consistently reported that higher fiber intake protects against T2D. We have observed this protective association only among carriers of the non-risk CC genotype who constitute around 55% of the population, while among carriers of the risk genotypes (CT and TT) fiber intake did not protect against diabetes. In addition, TCF7L2 genotype similarly modified the association of fiber intake with the metabolic syndrome. Individuals with the metabolic syndrome usually have a clustering of different T2D and CVD risk factors, as obesity, high blood pressure, high blood levels of cholesterol and glucose. More than 50 locations on the human genome have been associated with increased risk for T2D. TCF7L2 is a transcription factor in a cellular pathway called the WNT signaling pathway. To understand if fiber intake affects T2D through the WNT signaling pathway, we investigated more than 51 T2D genes for connections to this pathway and identified 7 such connections. In addition to TCF7L2, we observed interactions between 2 genetic variants in the NOTCH2 and ZBED3 genes, indicating that fiber intake could exert its protective actions through this pathway. We have also observed interactions between the strongest CVD genetic variant (chromosome 9p21) with both vegetable and wine intakes influencing the risk of CVD. The protective association between high vegetable intakes and CVD was restricted to carriers of the non-risk AA genotype. The protective association between wine consumption and CVD was restricted to carriers of the risk genotypes (AG and GG). The ultimate goal in epidemiological studies is to obtain causality. However, observational studies are often biased due to several reasons. One of the main challenges for these studies is called confounding that happens due to the correlation of different factors that makes it difficult to pinpoint the causal one. This bias can be eliminated in randomized controlled trials through balancing the confounders between comparison groups by randomization. Since we know that genetic variants are randomly allocated at conception, they can also be used as means for balancing confounders and obtaining causality. In our study we have used genetic variants in genes that affect obesity, blood pressure, blood glucose and cholesterol to study the causal relationship between these traits and T2D and CVD. Our main observation was the causal association between lower LDL cholesterol (the bad cholesterol) and increased risk of T2D. However, lower LDL cholesterol is causally associated with lower risk of CVD. Our study is first in the world to connect lower LDL-cholesterol to increased risk of T2D. Our results are important because they raise the question if we should develop new drugs to further reduce LDL cholesterol in blood due to the possible increased risk of developing T2D. To summarize, this thesis has provided novel evidence for gene-lifestyle interactions in T2D and CVD which need to be followed-up in future studies. Our results raise concerns for increased risk of T2D associated with lower levels of LDL cholesterol. Future studies need to be performed to understand the mechanisms that connect LDL-cholesterol to T2D.