Abstract #89  -  Hierarchical modeling of the effect of orphanhood on sexual behaviour & HIV risk in sub-Saharan Africa
Authors:   Presenting Author:   Ms Laura Robertson - Imperial College London
Additional Authors:  Prof Simon Gregson, Prof Geoff Garnett,
Aim: In countries with high HIV prevalence there will be higher incidence (and prevalence) of orphans. Increased pressure on households caring for orphans may reduce the standards of care and guidance provided to orphans. Additionally, orphans may suffer from stigma, discrimination and psychosocial distress. Evidence from studies in Southern Africa indicate that orphans, particularly maternal orphans, are more likely to engage in risky sexual behaviours and to be infected with HIV than non-orphans. There is also some evidence that education and/or marriage can help to mitigate these effects, although these results are less consistent across studies. We use data from several sub-Saharan African countries to investigate the effects of orphan status on sexual behaviour and HIV risk.
Method / Issue: The Demographic and Health Studies are nationally representative, cross-sectional surveys conducted in sub-Saharan Africa. Data was extracted from all surveys with HIV testing and data on orphan status for children aged 15-17 years. Age-adjusted, logistic, hierarchical regression models have been constructed with a random intercept at the country level. Initially, we considered the effects of orphan status, amongst women aged 15-17 years, on the outcomes “ever had sex?” and HIV status. We also investigated interactions between orphan status and education level, marital status, household socio-economic status, country level HIV prevalence and GDP per capita, with respect to sexual activity and HIV status.
Results / Comments: Data on 10,641 children was extracted from 10 DHS surveys conducted between 2003 and 2007. The hierarchical regression models suggest that maternal (OR 1.28; 95% CI 1.04-1.58), paternal (OR 1.20’ 95% CI 1.06-1.37) and double orphans (OR 1.41; 95% CI 1.18-1.70) are at increased risk of having started sex compared to non-orphans. Orphan status remains significant when education level and marital status are added to the models. There is a significant interaction between orphan status and education level (p=0.037). No other significant interactions were found with any of the proposed variables. The models also suggest that maternal (OR 2.70; 95% CI 1.58-4.60) and double (OR 3.34; 95% CI 2.18-5.12) orphans are more likely to be HIV positive. Paternal orphans did not experience an increased risk of being HIV positive (OR 1.21; 95% CI 0.78-1.88). A similar pattern of association is found when educational level and marital status are added to the model. No significant interactions were found with any of the proposed variables.
Discussion: It would appear that orphans are more likely to have started sex, independent of their marital status. They also appear to be at increased risk of being HIV positive, although the lack of an association with paternal orphanhood suggests that this pattern may be influenced, to some extent, by vertical HIV transmission. Further work will include creating larger, multivariate models; using different outcome variables; investigating the addition of new levels to the hierarchical models; looking at the effects fostering; and repeating the analysis using similar data from male adolescents.
Go Back