| dc.description.abstract | Body mass index (BMI) is an established risk factor for many cardio-metabolic conditions including hypertension, coronary heart disease (CHD), cardiovascular disease (CVD), and type 2 diabetes (T2D), as well as being associated with certain cancers including breast cancer and prostate cancer. Obesity (BMI>30) cases have increased over the last 70 years, with higher rates of increase now observed across developing countries. Obesity is both a global health concern and an economic burden.
Genome-wide association studies (GWAS) have enabled the identification of ~1000 common genetic variants that are associated with BMI, accounting for approximately 6% of the phenotypic variance explained. A shift in the distribution of BMI over the last 70 years, is largely attributed to recent changes in environment and lifestyle. Examples of environmental risk factors include an increase in the availability of ‘fast foods’ such as fried foods and sugar-sweetened drinks, and a decrease in physical activity, for example through a smaller proportion of the population taking up labour-intensive jobs, and a large proportion of private and public sector involving office-based work. Studies have investigated the gene x environment interaction (G x E) on BMI and found there are obesogenic environments which accentuate the genetic effect, which include socioeconomic position, sugar-sweetened drinks, alcohol consumption and sedentary lifestyle. These studies have focused on genetic risk scores or individual common variants with the strongest main effects, such as variants in the FTO locus. It is critical that studies investigate the effect of rarer variants on BMI, identifying variants in known BMI-associated genetic loci and some novel loci. In chapter two I identify G x E interaction of non-modifiable risk factors sex and age, and subsequently in chapter three I find G x E interaction of modifiable risk factors, on BMI using 76 of the 97 common BMI-associated variants previously identified by BMI GWAS. I test for correlations between the variant main effects and interaction effects, and determine outliers from the correlation if present, where the interaction effect is greater than expected. In chapter four I identify rare variant burden-BMI associations in regions where a common variant is associated with BMI. I then extend analyses by testing for rare variant burden-BMI associations exome-wide, where I found novel BMI-associated loci not previously reported in BMI GWAS. In chapter five I tested for G x E interaction of a rare variant burden with modifiable and non-modifiable exposures where previous evidence of interaction was identified. Finding no widespread evidence of G x E interaction, I recommend further replication in larger studies with increased power to detect rare associations. This thesis illustrates that G x E effects are difficult to ascertain but are important to account for when assessing genetic associations for complex traits such as BMI. | en_GB |
