Objectives: To investigate the association between the three human retroviruses, HIV-1, HIV-2 and HTLV-I.
Design: Community-based follow-up studies of retrovirus infections in two cohorts.
Methods: A total of 2057 individuals aged 35 years and over were eligible for inclusion. Participants were interviewed and had a blood sample drawn. Samples were analysed for HIV-1, HIV-2 and HTLV infections. Uni- and multivariate analyses that included behavioural and socio-economic factors were performed using logistic regression and Poisson regression models.
Results: A total of 1686 individuals participated with a blood sample in the HIV prevalence analyses and 1581 individuals participated in the HTLV-I prevalence analyses. The overall prevalence was 2.1% for HIV-1, 13.5% for HIV-2 and 7.1% for HTLV-I. Comparing the < 45 year age group with 45+ year age groups, the female : male (F : M) prevalence ratio increased with age for all three retroviruses. Dual infections were more common in women than in men. Assuming independent distribution of the viruses, the observed prevalence of dual infections in women was significantly higher than expected, while the prevalence was not increased in men. The prevalence of dual infections increased with age in women, the odds ratio (OR) being 3.4 [95% confidence interval (CI), 1.0-11.3] for any combination of dual infection, while the prevalence decreased with age in men (OR, 0.3; 95% CI, 0.1-2.0) (test of interaction, P = 0.033). Control for behavioural factors did not modify these patterns.
Conclusions: The pattern of increased prevalence among older women could have public health implications; women of older age groups should be regarded as a potential vulnerable group and included in HIV/AIDS prevention programmes.