PREDICTING THE RISK OF DEVELOPING TYPE-2 DIABETES MELLITUS IN URBAN AND RURAL YOUNG ADULT POPULATIONS IN DELTA STATE

ABSTRACT

Diabetes mellitus is gradually attaining an increasingly worrisome proportion. As a result, there is need to stem the tide and such efforts are targeted at preventing or delaying the onset of the disease. This study was aimed at predicting the 10-year risk of developing type-2 diabetes mellitus in urban and rural young adult populations in Delta state. The study involved 188 male and 212 female human participants (aged 15 – 35 years) without a prior diagnosis of diabetes. Study locations were respectively Asaba (urban) and Igbodo (rural). Diabetes risk score assessment was carried out using the Finnish Diabetes Risk Scoring (FINDRISC) tool and all measurements were done following standard procedures. High risk participants were further subjected to assessment of selected biochemical parameters following standard protocols. Of the 400 participants, 50% rural (48% male, 52% female) and 50% urban (46% male, 54% female) was determined. A total of 4.7% (6.6% female vs 2.7% male; 9.0% urban location vs 0.5% rural location) attained FINDRISC ≥ 12 while a high risk prevalence of 0.7% (1.1% male vs 0.5% female; 1.5% urban location vs 0.0% rural location) was found in the entire population. Hyperglycemia was found in 19.2% in the entire population (22.1% female vs 17.0% male; 30.0% urban location vs 8.5% rural location). The prevalence of metabolic syndrome in the subpopulation studied was 23.8% and was higher in males (27.58%) relative to females (20.58%). There was a high prevalence of the different BMI-metabolic-risk sub-phenotypes with metabolically obese normal weight (MONW) and metabolically healthy obese (MHO) subphenotypes occurring at rates of 23.68% and 60.0% respectively. The diagnostic accuracy analysis revealed that FINDRISC had Area Under Curve (AUC) = 0.69 with optimal cut point of 5.50 in the overall population; AUC = 0.69 with optimal cut point of 5.50 in males; AUC = 0.66 with optimal cut point of 6.50 in females; AUC = 0.68 with optimal cut point of 6.50 in the urban population; AUC = 0.49 with optimal cut point 5.50 in the rural population. Generally, FINDRISC score had no significant positive correlation with the studied biochemical parameters but correlated well with adiposity indices viz; body mass index = +0.610 (P < 0.001), body adiposity index = +0.561 (P < 0.001), waist-to-height ratio = +0.557 (P < 0.001), waist circumference = +0.495 (P < 0.001). From this study, a novel risk score prediction equation was developed. These findings suggest that diabetes risk susceptibility preponderate in the urban location relative to the rural xxii location and in female participants relative to male participants, thus creating a double jeopardy for urban female population. The FINDRISC tool demonstrated a fairly good discriminative ability in the study population. The use of the novel diabetes risk prediction equation in other populations is recommended