Indirect and direct effects of factors associated with diabetes amongst the rural black population in the Dikgale Health and Demographic Surveillance System, South Africa.

backgroundDiabetes is an enormous, growing clinical and public health problem, which together with hypertension contributes significantly to the high risk of cardiovascular diseases (CVDs) globally.

aimTo examine the indirect and direct effects of risk factors simultaneously as a network of multiple pathways leading to diabetes in the rurally based adult population (aged 15+) using a household survey.

methodsThis investigation was based on a predictive model using a cross-sectional community-based study to identify the direct and indirect effects of diabetes risk factors in the Dikgale Health and Demographic Surveillance System (HDSS) consisting of 15 villages, with 7200 households and a total population of approximately 36 000. Fasting blood glucose and total cholesterol were measured using ILAB 300 with the following cut-off values: high fasting blood glucose 7 mmol/L and triglycerides 1.70 mmol/L.

resultsA total of 1407 individuals were interviewed, of whom 1281 had their blood pressure (BP) measured. The conceptual model was validated by the goodness-of-fit indexes (comparative fit index [CFI] = 1.00, Tucker Lewis index [TLI] = 1.041, root mean square error of approximation [RMSEA] = 0.001). Hypertension had the strongest direct effect of 0.0918 on diabetes, followed by age (0.0039) and high waist circumference (-0.0023). Hypertension also mediates the effects that high waist circumference (0.0005) and triglycerides (0.0060) have on diabetes status.

conclusionThe results in this study confirm the conceptual model considered in the risk factors for diabetes and suggest that hypertension, age and high waist circumference are the key variables directly affecting the diabetes status in the South African rural black population. The direct effect of triglycerides on diabetes suggests mediation by some measured factor(s).