Background
Obesity has reached epidemic proportions in South Africa with significantly higher levels in women compared with men. Linked to the obesity epidemic is the high and rising prevalence of cardio-metabolic diseases of diabetes, hypertension and dyslipidaemia in the country. However, the patterns of these associations have not been examined locally by gender and/or population groups.

Aims
To determine whether adiposity, calculated by body mass index (BMI), was differentially associated with diabetes, hypertension and hypercholesterolaemia by gender and population groups in the Heart and Stroke Foundation South Africa’s screening programme.

Methods
This cross-sectional study was conducted nationally in five of the nine South African provinces in 2013 among ≥18-year-old adults self-selected for screening. Participants were classified according to the previously defined official South African population groups of black African, coloured, white and Indian. Clinical measurements comprised anthropometry, blood pressure (BP), and point-of-care random blood glucose (RBG) and cholesterol assessments. Diabetes was defined as RBG ≥11.1 mmol/l or known diabetes, hypercholesterolaemia as total cholesterol ≥5 mmol/l, and hypertension as systolic BP ≥140 mmHg and/or diastolic BP ≥90 mmHg or the use of antihypertensive agents. Categories of BMI were defined as underweight (<18.5 kg/m²), normal weight (18.5-24.9 kg/m2), overweight (25-29.9 kg/m2) and obesity (≥30 kg/m2). A univariate logistic regression was performed for each variable (age, gender, population group, province, family and personal medical history, diet relating to fruit and vegetable, salt and fat intake, physical activity, smoking, alcohol misuse and adiposity) to determine the effect on the odds of diabetes, hypertension and hypercholesterolaemia by gender and population group. Variables with a p-value of ≤0.05 from the univariate analysis were then used to construct the multivariate models. These results are presented as odds ratios (ORs) with the corresponding 95% confidence intervals (CIs). 

Results 
Among the 7711 participants, 2488 men and 5223 women, mean ages were 47.6 years and 48.6 years, respectively. In men and women, prevalence were: obesity 25.6% vs. 43.7%, p<0.001, diabetes: 13.8% vs. 12.8%, p=0.232, hypertension: 51.8% vs. 48.9%, p=0.017, and hypercholesterolaemia: 31.6% vs. 37.3%, p<0.001. In obese compared with their normal weight counterparts, the ORs for diabetes were 2.36 (95% CI: 1.66-3.36) for men and 1.76 (95% CI: 1.36-2.28) for women. Similarly, although not significantly different, obese men versus obese women had a greater likelihood for hypercholesterolaemia (OR: 1.53, 95% CI: 1.21-1.94 vs. OR: 1.19, 95% CI: 1.02-1.39) and hypertension (OR: 3.11, 95% CI: 2.44-3.97 vs. OR: 2.44, 95% CI: 2.04-2.92) compared with their normal weight counterparts. By population group, compared with normal weight, overweight was significantly associated with diabetes in Indians only (1.79, 95% CI: 1.28-2.52), while obesity was associated with diabetes in black Africans (OR: 2.47, 95% CI: 1.54-3.97), coloureds (OR: 2.04, 95% CI: 1.28-3.23) and whites (OR: 2.9, 95% CI: 1.72-4.88). Although not statistically significant, the odds for diabetes was greater in underweight compared with normal weight black Africans (OR: 2.54, 95% CI: 0.52-12.40). A generally linear pattern was observed for hypercholesterolaemia and BMI category in black Africans and whites while there was no discernible pattern in Indians or coloureds. There was a greater risk for hypercholesterolaemia in obese compared with normal weight black Africans (OR: 1.63, 95% CI: 1.25-2.12) and whites (OR: 1.55, 95% CI: 1.19-2.03). Hypertension and BMI category were generally linearly associated across all population groups. Women compared with men were less likely to have diabetes but this was significant only for black Africans (OR: 0.64, 95% CI: 0.46-0.89). Black African, coloured and white, but not Indian, women were significantly less likely to have hypertension compared with their male counterparts. 

Discussion
The similar to higher prevalence of diabetes and hypertension with significantly lower obesity levels in men, compared with women, suggests their greater susceptibility to these cardio-metabolic abnormalities at a particular adiposity level or perhaps the influence of other risk factors. Similarly, Indians, compared with other population groups, had a propensity for diabetes at lower BMI levels, which highlights their need for stricter weight control. The association of underweight with diabetes in black Africans may be linked to the Barker hypothesis and/or the high prevalence of HIV infection in South Africa, and warrants further investigation. The absence of an association between hypercholesterolaemia and BMI in Indian and coloured populations indicates the importance of other contributory factors.

Associations between cardio-metabolic abnormalities and adiposity by gender and population groups in South Africa