Multilevel Analysis of the Determinants of HIV infection: Evidence from Nigeria
Nigeria ranks second globally in the number of people living with HIV/AIDS. HIV prevention is of strategic public health importance needing timely evidence from available data. Understanding determinants of HIV infection is crucial in designing appropriate strategies that will be deterrence in effective prevention programs. This study evaluated determinants of HIV and its zonal effects towards reducing the burden of the virus.
METHODS:
This was secondary data analysis of 2012 National HIV/AIDS Reproductive Health Survey (NARHS). This is nationally representative survey that contained information on socio-demographic, sexual and reproductive health, and HIV test. Random-effects logistic regression model was used.
RESULTS:
Sample size was 31,235 with 49.9% female. Male median age was 32years (15-64years); female median age was 28years (15-49years) and median age at first sex was 20years. About 51.3% had at least secondary education; 63.9% were currently married; 68.7% were rural dwellers; 5.1% had sex in exchange for gifts; and 67.2% had sexual intercourse in the last 12 months. National HIV prevalence was 3.4% with heterogeneity across zones from 1.8% to 4.4%. The HIV determinants were once a week of alcohol OR: 1.4 95% CI (1.1 -1.8); had casual sex OR: 1.6 95 % CI (1.1 - 2.5); living in rural area OR: 1.2 95% CI (1.1 -1.5); while being a male was protective with OR: 0.7 95% CI (0.6 - 0.8). Estimated variance between zones was 0.36 with a standard error of 0.13.
CONCLUSIONS:
This study has shown that being a female, living in rural areas, once a week of alcohol and having casual sex were risk factors for HIV. There is a need to scale-up targeted behavioral interventions among rural dwellers and women. More analyses are needed to inform HIV programming at zonal/state levels, and to provide parameter values for mathematical modeling of HIV incidence in the zones.