DIABETES-MELLITUS AS A CAUSE OF DEATH IN SUB-SAHARAN AFRICA - RESULTSOF A COMMUNITY-BASED STUDY IN TANZANIA

The aim of this study was to determine the contribution of diabetes me llitus to all-cause mortality and diabetes mortality rates in adults 1 5 years and above living in one urban and two rural areas of Tanzania (Dar es Salaam, Hai and Morogoro Rural Districts). The three surveilla nce populations comprised 307 912 persons. Prospective monitoring of a ll deaths between 1 lune 1992 and 31 May 1995 was carried out. Cause o f death was determined by verbal 'autopsy' conducted with relatives of the deceased. In total, 4299 deaths were recorded in children (aged < 15 years) and 8054 in adults. In children there were no reported death s associated with diabetes (due to or in children with diabetes). The adult male mortality rates associated with diabetes were 34, 30, and 1 5 per 100 000 per year in Dar es Salaam, Hai and Morogoro Rural Distri cts respectively. The figures in women were 21, 18, and 4 per 100 000 per year, respectively. The percentages of all adult male deaths assoc iated with diabetes were 2.6 %, 2.1 % and 0.7 % respectively. In women the percentages were 1.7 %, 1.8 %, and 0.2 % respectively. Acute meta bolic complications, infection, and stroke each accounted for approxim ately 30 % of all diabetic deaths. Thus diabetes mortality rates varie d between the three surveillance areas, being lowest in the poorest ru ral area. Rates were higher in men in all areas. While care is require d in the comparison of mortality rates between countries, it was notew orthy that Tanzania, a country with a low diabetes prevalence, had dia betes mortality rates which were higher than or comparable to rates in Mauritius and the United States. Most patients died from preventable causes, indicating a need for improved case-management of diabetic eme rgencies as well as better detection and treatment of hypertension.