Condom use and its associations with HIV/sexually transmitted diseases in four urban communities of sub-Saharan Africa

Objectives: To estimate rates of condom use in four urban populations in su b-Saharan Africa and to assess their association with levels of HIV infecti on and other sexually transmitted diseases (STDs). Methods: Data were obtained from a multicentre study of factors that determ ine the differences in rate of spread of HIV in four African cities. Consen ting participants were interviewed on sexual behaviour, and also provided b lood and urine samples for testing for HIV infection and other STDs. Data o n sexual behaviour included information on condom use during all reported s pousal and non-spousal partnerships in the past 12 months. Results: A total of 2116 adults aged 15-49 years were interviewed in Cotono u (Benin), 2089 in Yaounde (Cameroon), 1889 in Kisumu (Kenya) and 1730 in N dola (Zambia). Prevalence rates of HIV infection were 3.4% in Cotonou, 5.9% in Yaounde, 25.9% in Kisumu and 28.4% in Ndola. Reported condom use was lo w, with the proportions of men and women who reported frequent condom use w ith all non-spousal partners being 21-25% for men and 11-24% for women. A h igher level of condom use by city was not associated with lower aggregate l evel of HIV infection. The proportions of men reporting genital pain or dis charge during the past 12 months were significantly lower among those repor ting frequent condom use in all sites except Yaounde: in Cotonou, adjusted odds ratio (OR) = 0.28, 95% confidence interval (CI) = 0.09-0.94; in Kisumu , adjusted OR = 0.34, 95% Cl = 0.14-0.83; and in Ndola, adjusted OR = 0.33, 95% Cl = 0.12-0.90. The same association was found for reported genital ul cers in two sites only: in Cotonou, adjusted OR = 0.14, 95% Cl = 0.02-1.02; and in Kisumu, adjusted OR = 0.18, 95% Cl = 0.04-0.75. There were few stat istically significant associations between condom use and biological indica tors of HIV infection or other STDs in any of the cities. Conclusion: Similar levels of condom use were found in all four populations , and aggregate levels of condom use by city could not discriminate between cities with high and low level of HIV infection. It seems that rates of co ndom use may not have been high enough to have a strong impact on HIV/STD l evels in the four cities. At an individual level, only a male history of re ported STD symptoms was found to be consistently associated with lower rate s of reported condom use. (C) 2001 Lippincott Williams & Wilkins.