Factors affecting HIV concordancy in married couples in four African cities

Objective: To determine risk factors for HIV transmission within married couples in four urban populations in sub-Saharan Africa. Methods: Data from a cross-sectional population-based study were used. Representative random samples approximating 1000 men and 1000 women in each of four cities of Kisumu (Kenya), Ndola (Zambia), Cotonou (Benin), and Yaounde (Cameroon), were interviewed and tested for sexually transmitted infections (STI). Married couples were identified as concordant negative, discordant, or concordant positive for each STI. After excluding concordant HIV negative couples, analysis of behavioural and STI risk factors for HIV positive concordancy was undertaken across the four cities and in each city separately where sample size allowed. Results: Among 221 couples in which at least one member was HIV positive, we found that the only significant risk factor for positive HIV concordancy was herpes simplex type 2 (HSV-2) status. After adjusting for age and city of residence the odds ratio for HIV concordancy compared to couples with neither spouse HSV-2 positive was 3.4 (95% confidence interval, 0.62-18.4) for couples with one partner HSV-2 positive and 8.6 (95% confidence interval, 1.6-45.0) for couples with both partners HSV-2 positive. The same trends were seen in Kisumu and Ndola when they were analysed separately (numbers were small in the other cities). Conclusions: Although cross-sectional studies are not ideal for delineating the sequence of transmission events, this study adds to the evidence that HSV-2 is a key risk factor in promoting HIV transmission. (C) 2004 Lippincott Williams Wilkins.