Characterisation of gene-lifestyle interactions associated with obesity- related traits in African populations (360G-Wellcome-214205_Z_18_Z)

Obesity is increasing in Africa and elsewhere, but little is known of how genetic and environmental factors interact.The proposed research has the following components: (i) discovery phase: heterogeneity of variance analysis (HEVA) to determine candidate gene-environment interacting (GEI) variants in the Human Hereditary and Health in Africa (H3Africa) cardiovascular (CVD) working group to replicate the top-ranking signals. Additional variants will be selected from the GWAS trans-ethnic meta-analysis of East Asians, Europeans and Africans based on exhibiting heterogenous effects across the ethnic groups. Gradient Boosted LD adjusted (GraBLD) machine learning algorithms will be used to compute highly predictive genetic risk scores; (ii) cross-sectional and longitudinal analysis phase: single candidate GEI variants and genetic risk scores (of candidate GEI variants and from GraBLD analysis) will be combined with lifestyle factors in generalised linear models (cross-sectional analysis) and linear mixed models (repeated measures analysis) to determine plausible variants responsible for gene-lifestyle interactions in obesity- related traits. Assessments of sensitivity and specificity will be used to determine the predictiveness of these gene-lifestyle interactions and genetic risk scores. The transferability of the identified GEIs across African and European populations will be evaluated. The proposed research may help elucidate relevant risk factors and/or identify obesity intervention targets.