Article Text
PDF
Abstract
Background It has been suggested that the decline in HIV prevalence in Zimbabwe from the late 1990s may be partly due to reductions in sexual concurrency but little is known about levels and trends in the different forms of concurrency or their association with HIV risk.
Methods We use data from four rounds (1998–2000, 2001–03, 2003–05, 2006–08) of a large longitudinal population-based HIV survey to investigate patterns of non-spousal (two or more current sexual partners, at least one of whom is not a spouse) and spousal (multiple spouses but currently no extra-marital sexual partners) concurrency, and associations with prevalent HIV infection in rural east Zimbabwe.
Results 15.4% (95% CI 14.4% to 16.5%) of men (17–54 yrs, N=4327) and 1.9% (1.5% to 2.3%) of women (15–44 yrs, N=5148) reported concurrent sexual partners at baseline. 11.7% (10.8% to 12.7%) and 3.7% (3.2% to 4.3%) of men reported non-spousal and spousal concurrency, respectively. Non-spousal concurrency was most common in single (16%) and divorced/widowed (11%) men but was also reported by married men with single (6.5%) and multiple (4%) spouses (Abstract O1-S06.06 figure 1). HIV prevalence is similar in men with (19.2%) and without (19.5%) concurrent partners but is higher in those with concurrent partners after controlling for age (age-adj. OR (aOR), 1.37; p=0.02). However, after excluding non-sexually experienced men, this difference was not statistically significant (aOR, 1.27, p=0.08). Among all men, HIV prevalence was higher for those with non-spousal concurrency (aOR, 1.38; p=0.015) but not for those with spousal concurrency (aOR, 1.26; p=0.2). For women, concurrency (all non-spousal) was 1.9% (95% CI 1.2% to 2.7%) in single women, 0.4% (0.2% to 0.7%) in married women, and 5.4% (4.1% to 7.0%) in divorced/widowed women. HIV prevalence was higher in women with concurrent partners than in those without (55.7% vs 25.4%; aOR, 3.26, 2.08 to 5.11) even after excluding women who had not started sex (aOR, 2.83; p<0.001). For males, non-spousal concurrency fell from 11.7% (95% CI 10.8% to 12.7%) in 1998–2000 to 6.1% (5.3% to 7.0%) in 2001–2003 and 4.3% (3.7% to 5.0%) in 2006–2008; prevalence of spousal concurrency fell from 3.7% (3.2% to 4.3%) to 2.6% (2.0% to 3.2%) to 1.3% (1.0% to 1.7%) over the same period. For females, concurrency declined from 1.7% (1.4% to 2.1%) in 1998–2000 to 1.0% (0.7% to 1.3%) in 2001–2003 and 0.5% (0.3% to 0.7%) in 2006–2008.
Prevalence of non-spousal concurrency by marital status, Men, 17–54 years.
Conclusion A 2/3rds reduction in (mainly non-spousal) concurrency may have contributed to HIV decline in east Zimbabwe.