Socio-behavioural characteristics and HIV: findings from a graphical modelling analysis of 29 sub-Saharan African countries

Abstract

Introduction Socio‐behavioural factors may contribute to the wide variance in HIV prevalence between and within sub‐Saharan African (SSA) countries. We studied the associations between socio‐behavioural variables potentially related to the risk of acquiring HIV. Methods We used Bayesian network models to study associations between socio‐behavioural variables that may be related to HIV. A Bayesian network consists of nodes representing variables, and edges representing the conditional dependencies between variables. We analysed data from Demographic and Health Surveys conducted in 29 SSA countries between 2010 and 2016. We predefined and dichotomized 12 variables, including factors related to age, literacy, HIV knowledge, HIV testing, domestic violence, sexual activity and women's empowerment. We analysed data on men and women for each country separately and then summarized the results across the countries. We conducted a second analysis including also the individual HIV status in a subset of 23 countries where this information was available. We presented summary graphs showing associations that were present in at least six countries (five in the analysis with HIV status). Results We analysed data from 190,273 men (range across countries 2295 to 17,359) and 420,198 women (6621 to 38,948). The two variables with the highest total number of edges in the summary graphs were literacy and rural/urban location. Literacy was negatively associated with false beliefs about AIDS and, for women, early sexual initiation, in most countries. Literacy was also positively associated with ever being tested for HIV and the belief that women have the right to ask their husband to use condoms if he has a sexually transmitted infection. Rural location was positively associated with false beliefs about HIV and the belief that beating one's wife is justified, and negatively associated with having been tested for HIV. In the analysis including HIV status, being HIV positive was associated with female‐headed household, older age and rural location among women, and with no variables among men. Conclusions Literacy and urbanity were strongly associated with several factors that are important for HIV acquisition. Since literacy is one of the few variables that can be improved by interventions, this makes it a promising intervention target.

Clusters of sub-Saharan African countries based on sociobehavioural characteristics and associated HIV incidence

“…Most used standard statistical methods like descriptive statistics (Sangowawa & Owoaje, 2012;Smith Fawzi et al, 2016), linear or logistic regression (Mondal & Shitan, 2013;Delavande, Sampaio & Sood, 2014;Kidman & Anglewicz, 2016;Ashaba et al, 2018), or concentration indices (Hajizadeh et al, 2014;Pons-Duran et al, 2016;Kim et al, 2016), to assess health inequity and the impact of a small number of variables (5 to 13) on the HIV epidemic. In a previous work, we also investigated the associations and possible causal relationships between sociobehavioural factors at the individual-level that are potentially related to the risk of acquiring HIV in 29 SSA countries using Bayesian Network models (Baranczuk et al, 2019). But, these methods do not inform us on how HIV risk factors vary across SSA and which characteristic sociobehavioural patterns at the country-level are actually associated with different rates of new HIV infections in the region.…”