The association between diabetes and disability: what is the contribution of diabetes risk factors and complications?

Background 
Identifying risk factors that explain the association between diabetes and disability is a perquisite to prevent disability related to diabetes. The majority of studies on this have been conducted among Europid populations, but studies on other ethnicities such as South Asians and Africans are rare. Mauritius is a multiethnic country comprising mainly South Asians and Creoles, mainly of African origin “Africans”.
Aims
We aimed to evaluate the association between diabetes and disability among South Asians and Africans and to assess the extent to which the effect of diabetes is explained by risk factors.
Method
Data from a national survey in Mauritius in 2015 were analysed. Disability was measured using the Katz activities of daily living (ADL) questionnaire in participants aged >50 years and was defined as having any difficulty in any of the six Katz ADL questions. This instrument ranks adequacy of performance in six major activities of daily living including: walking bathing, transferring, toileting, dressing, eating. A series of logistic regression models were conducted and adjusted for the following variables: model 1, adjusted for age, sex, education and smoking; model 2, adjustments in model 1 plus BMI and waist; model 3, adjustments in model 2 plus blood pressure and lipids; model 4 was adjusted model 3 plus chronic disease (history of cardiovascular disease, asthma, impaired glomerular filtration rate (GFR) and depression). To quantify the contribution of risk factors that explains the association of disability with diabetes, we calculated the percentage of excess odds of disability related to diabetes for each risk factor and for the total risk factors combined.
Results
 Among 3,692 participants, 487 (13.2%) reported having any disability. In the fully adjusted model, diabetes was significantly associated with disability [OR=1.29 (95% CI 1.01, 1.67)].  When stratified by ethnicity, diabetes was strongly related to disability among Africans [OR=2.06 (95% CI 1.08-3.91)] but not in South Asians [OR=1.21 (95% CI 0.91-1.62)]. Body mass index and waist circumference together explained ~33% and 20% of the association between diabetes and disability in South Asians and Africans, respectively. Chronic diseases or conditions such as cardiovascular, impaired GFR, asthma and depression together explained 26.6% and 14.2% of the excess odds ratio in South Asians and in Africans, respectively. Collectively the risk factors explained about 54% and 25% of the association of diabetes with disability in South Asians and Africans, respectively.
Conclusion
The association between diabetes and disability was stronger among Africans as compared to South Asians. In both ethnicities, BMI and waist together explained the largest proportion of the association between diabetes and disability indicating weight management should be one of the main strategies in preventing disability related to diabetes. In this study, adjustment for risk factors partially attenuate the strength of the association between diabetes and disability, suggesting that the rest of the association between diabetes and disability is explained by hyperglycaemia and other factors such as socio-economic status, dietary habits, etc.