Abstract/Summary

Low vitamin D concentration has been shown to be a risk factor for metabolic traits in several observational studies; however, the findings have been inconsistent. Certain genotypes could potentially contribute jointly to obesity and vitamin D deficiency, and these may be modulated by lifestyle factors (dietary factors and physical activity levels) across different ethnic groups. The implementation of a genetic approach to establish the relationship between vitamin D and metabolic diseases could be a more desirable option over observational studies, as results are less prone to confounding factors. Hence, the main aims of this thesis were to examine the association of common vitamin D-related single nucleotide polymorphisms (SNPs) and metabolic SNPs, as genetic risk scores (GRSs), with vitamin D concentrations and metabolic outcomes in multiple ethnic groups. In addition, the interaction between these GRSs and dietary factors (protein, fat, carbohydrate, and fibre) on vitamin D concentrations and metabolic traits was investigated. A total of five different studies were used. These studies included two case-control studies [the Chennai Urban Rural Study (CURES; Asian Indian, n = 545) and a study in Turkish adults (n = 396)] and three cross-sectional cohort studies [the Minangkabau Indonesia Study on Nutrition and Genetics (MINANG study; Indonesian women; n = 110), the Obesity, Lifestyle and Diabetes in Brazil (BOLD study; Brazilian young adults; n = 187), and the Genetics of Obesity and Nutrition in Ghana (GONG study; Ghanaian adults; n = 302)]. Gene-diet interactions were observed in the Indonesian and Ghanaian populations between vitamin D related-GRS and dietary intake on obesity markers, carbohydrate intake (g) on body fat percentage (pinteraction = 0.049) and fibre intake (g) on body mass index (pinteraction = 0.020), respectively. Additionally, in the Ghanaian population, an interaction was found between vitamin D-related GRS and dietary fat intake (g) on glycated haemoglobin levels (pinteraction = 0.029), a marker of long-term glycaemic status. In the Brazilian population, the vitamin D related-GRS showed a significant interaction with protein intake (g) on 25(OH)D levels (pinteraction = 0.006). Furthermore, we observed novel interactions in metabolic disease related-GRSs on vitamin D concentrations with carbohydrate energy intake (%) in the Indian population (pinteraction = 0.047), and with fat energy intake (%) in the Turkish populations (pinteraction= 0.040). An additional interaction was observed in the Indian population with the metabolic disease related-GRS and fat energy intake (%) on low-density-lipoprotein cholesterol (pinteraction = 0.032), a marker of cardiovascular disease. In summary, these studies in multiple ethnic groups show that the relationship between vitamin D status and metabolic outcomes may be influenced by dietary factors such as protein, fat, carbohydrate, and fibre intake. Replication of these findings in larger cohorts and in dietary intervention studies is warranted before these findings could be considered for personalised dietary recommendations.