Vitamin D status and cardiometabolic risk factors in black African and Indian populations of South Africa

Background: South Africa is in the midst of a health transition that is characterized by a high burden of both infectious diseases and non-communicable diseases. One of the drivers of non-communicable diseases in South Africa is the current epidemic of obesity. Vitamin D deficiency, which is defined by 25(OH)D levels in blood, has been reported to be a risk factor for cardiovascular disease and shares a number of risk factors with those traditionally linked to non-communicable diseases. Osteoporosis is another non-communicable disease that is reportedly increasing in prevalence worldwide and may be linked to vitamin D levels and to body fat. There is limited data on 25(OH)D levels in South Africa and its association with cardiovascular risk factors. There is also limited data on body composition including bone mineral density. Aims: The aims of this thesis were to describe 25(OH)D levels in healthy Black African and Indian subjects recruited from the greater Johannesburg metropolis and to determine if differences in 25(OH)D levels contributed to differences in cardiovascular risk. A further aim was to describe body composition in both ethnic groups and to see if differences in body composition contribute to differences in 25(OH)D levels or to differences in bone mineral density and to determine if differences in bone mineral density are mediated by differences in 25(OH)D. Methods: This was a cross sectional study carried out from July 2011 to March 2012 on 714 male and female subjects (male: female=340:374) of whom 371 were Black African and 343 were Indian. Subjects were recruited via the caregivers of the Birth to Twenty cohort. The first step was a descriptive analysis of 25(OH)D as well as its predictors including whole body fat, visceral and subcutaneous adiposity. This was followed by examining the associations of 25(OH)D and parathyroid hormone with cardiovascular risk factors that comprise the metabolic syndrome. Final analysis was description of bone mineral density according to ethnicity and gender and the contribution of lean mass, sub-total fat mass, visceral and subcutaneous adiposity to bone mineral density in each ethnic group. Results: Vitamin D deficiency was very prevalent in Indians, 28.6% in comparison to 5.1% in the Black African group (p<0.0001). In both groups season of collection was a positive predictor and PTH was negatively associated with 25(OH)D. Neither whole body fat nor visceral or subcutaneous adiposity was predictive of 25(OH)D in either group. Using the harmonized definition of the metabolic syndrome (Met S), was diagnosed in 29% of the Black African and 46% of the Indian subjects (p<0.0001). Subjects with Met S had higher PTH than those without (p<0.0001), whilst 25(OH)D levels were not significantly different (p=0.50). Logistic regression analysis showed that Indian ethnicity (OR 2.24; 95% CIs 1.57, 3.18; p<0.0001) and raised PTH (OR 2.48; 95% CIs 1.01, 6.08; p=0.04) adjusted for 25(OH)D) produced an increased risk of Met S but 25(OH)D did not (OR 1.25; 95% CIs 0.67, 2.24; p=0.48). Whole body, hip, femoral neck and lumbar spine bone mineral density were significantly higher in Black African than Indian subjects (p<0.001 for all). Whole body lean mass positively associated with bone mineral density at all sites in both ethnic groups (p<0.001 for all), and partially explained the higher bone mineral density in Black African females compared to Indian females. Whole body fat mass correlated positively with lumbar bone mineral density in Black African (p=0.001) and inversely with sub-total bone mineral density in Indian subjects (p<0.0001). Visceral adiposity correlated inversely with sub-total bone mineral density in the Black African subjects (p=0.037) and with lumbar bone mineral density in the Indian group (p=0.005). No association was found between serum 25(OH)D and bone mineral density. PTH was inversely associated with hip bone mineral density in the Black African group (p=0.01) and with sub-total (p=0.002), hip (p=0.001) and femoral bone mineral density (p