International Journal of TROPICAL DISEASE & Health

Introduction: HIV/AIDS and Noncommunicable diseases are the major public health threats of developing countries. Analysis of joint epidemiological patterns of these diseases will help in designing and implementing appropriate interventions to mitigate their impacts.
Objectives: The overall aim of this study was to analyze Epidemiological patterns of HIV/AIDS and Diabetes in developing countries.
Methods: Country level HIV/AIDS and Diabetes prevalence data at four time points, between 2000 and 2010, for 68 countries in Sub-Saharan Africa, Southern and South Eastern Asia were transformed and analyzed. Joint geographic and temporal trends were described using numerical and graphic summaries. The level of Covariation between HIV and Diabetes prevalence was measured by Pearson correlation. K-means cluster analysis was conducted after the appropriate number of clusters was determined using scree plot technique. Analysis of variance was used to identify factors that differentiate the clusters.
Results: Diabetes had higher mean prevalence with increasing trend while HIV/AIDS had higher disability-weight adjusted mean prevalence with a decreasing trend during the study period. The findings suggest that HIV/AIDS and Diabetes were negatively correlated throughout the study period (r > 0.3, P <.05 in all four time points). Hence, countries with higher prevalence of Diabetes tend to have lower prevalence of HIV/AIDS and vice versa. Four clusters of countries with size 29, 12, 12 and 14 countries were identified. These clusters were found to have significant variation with respect to their mean HIV and Diabetes prevalences as well as time trends in their mean prevalences.
Conclusions: Diabetes and HIV are heading in reverse directions during the study period in the study regions. The identified clusters were found to describe these patterns of variation across geography and time. The clusters may be useful in considering a set of coordinated country level interventions.