Differential Impacts of HIV status on short-term fertility desires among couples in Rakai, Uganda

BACKGROUND: Fertility desires of female and male partners in current relationships are often correlated. We examined the influence of HIV seropositive status of female and male partners on short-term fertility desires in Rakai, Uganda, a setting with high fertility and HIV infection rates. METHODS: Participants were couples (15-49 years old) enrolled in the Rakai Community Cohort Study, from 2011 to 2013 (n = 2,291). Cohen's kappa coefficient was used to measure the correlation of female and male partners' short-term fertility desires (measured as 'wanting a child in the next 12 months'), in both total sample and stratified serostatus groups. HIV serostatus and additional characteristics of female and male partners were included in Poisson regression models to estimate the rate ratios (RR) for each partner's short-term fertility desires. Individual and partner characteristics included HIV status, partner HIV status, age in years, partner age in years, educational attainment, number of living children, community of residence, and socioeconomic status (SES). RESULTS: Short-term fertility desires among female and male partners were moderately associated (Kappa = 0.37, p-value<0.001). The association was weakest among female sero-positive and male sero-negative couples (Kappa = 0.29, p-value<0.001). When adjusting for parity and other covariates in the model, women's short-term fertility desires were significantly associated with their positive sero-status regardless of male partners' sero-status (adjRR = 1.58, p<0.001 for F+M-; adjRR = 1.33, p = 0.001 for F+M+; in comparison with F-M-). Men's short-term fertility desires were significantly associated with their positive sero-status, in addition to their female partners' positive sero-status (adjRR = 1.23 with p-value = 0.022 for F-M+; adjRR = 1.42 with p-value<0.001 for F+M-; adjRR = 1.26 with p-value<0.001 for F+M+; in comparison with F-M-). When the differential effect of parity was included in the model, similar associations remained for both female and male partners when the number of living children was small, but largely reduced when the number of living children was large (3 or more). CONCLUSION: Female and male partners in couple dyads demonstrated moderate agreements about short-term fertility desires. The HIV seropositive status of female partners was most strongly associated with short-term fertility desires of both genders, and this association was even stronger for women who had few or no living children.