Trends and factors associated with illicit drug use and HIV outcomes in South Africa: findings from multiple national population-based household surveys

Background: Illicit drug use has the potential to impact on health and wellbeing, but there is little data on its trends and predictors in sub-Saharan Africa. We used data from South Africa to describe trends in recent drug use, assess factors associated with it, and evaluate whether recent drug use predict HIV outcomes.

Methods: We analysed data among adults aged 15+ years from five national household surveys conducted over 2002 to 2017. Multivariable logistic regression was used to assess predictors of recent drug use (defined as, any past three-months use of illicit drugs, excluding alcohol/tobacco), and to evaluate whether recent drug use predicted HIV outcomes while controlling for age, sex, race, and province.

Results: There were 89,113 respondents aged 15+ years from the five survey rounds. The prevalence of recent drug use increased from 1.5% in 2002 to 10.0% in 2017. Recent drug use was negatively associated with being female (adjusted odds-ratio [aOR] 0.21, 95% Confidence Interval [CI] 0.19-0.24), aged 35-49 vs 15-24 years (aOR 0.56, 95%CI 0.49-0.65), rural informal vs urban (aOR 0.70, 95%CI 0.58-0.83), grade 8-11 vs 0-7 (aOR 0.84, 95%CI 0.73-0.96) and grade 12+ vs 0-7 (aOR 0.65, 95%CI 0.56-0.76). Conversely, individuals of mixed-ancestry race (compared to black-African race, aOR 1·87, 95% CI 1·58-2·21) and white race (compared to black-African race, aOR 1·52, 95% CI 1·25-1·85) and the unemployed (compared to employed, aOR 1·34, 95% CI 1·19-1·51]), were more likely to have recently used drugs. Compared to not using drugs, recent drug use was associated with having multiple sexual partners (last-year), earlier sexual debut, hazardous or harmful alcohol use or alcohol dependence. Recent drug use was not associated with HIV positivity, different aspects of the HIV continuum-of-care or condom use.

Conclusion: Illicit drug use has increased substantially in South Africa and is associated with poorer socio-demographic characteristics and numerous risk behaviours.