How does HIV-related stigma correlate with HIV prevalence in African countries? Distinct perspectives from individuals living with and living without HIV

BACKGROUND: Population-level research evaluating HIV-related stigma among countries with varied national HIV prevalence is scarce. To better understand HIV-related stigma and mitigate its potential negative effects, it is necessary to evaluate its relationship with HIV prevalence, as well as the mechanisms that influence it. This study aimed to analyze how HIV-related stigma correlates with subnational HIV prevalence in three African countries with varied HIV epidemics. METHODS: This paper used data from the nationally representative Population-based HIV Impact Assessment (PHIA) surveys conducted from 2015–2017 in Malawi, Zambia, and Tanzania. Each country's sub-national geographic divisions were used to categorize them as low (0–5.4%), middle (5.5–11.2%), and high (11.3–17.1%) HIV prevalence regions in the main analysis. Questions from the survey stigma module were used to measure HIV-related stigma. Logistic regression and multilevel models were performed to assess the associations between the level of sub-national HIV prevalence and HIV-related stigma measures among persons living with, and without, HIV. RESULTS: The results show that the odds of people living without HIV expressing stigmatizing behavior towards PLWH was significantly lower in regions of middle (OR = 0.80, 90%CI = (0.68–0.96)) and high (OR = 0.65, 90%CI = (0.53–0.80)) HIV prevalence when compared to low prevalence regions. The odds of reporting discriminatory attitudes were also lower for those in middle (OR = 0.87, 90%CI = (0.78–0.98)) and high (OR = 0.64, 90%CI = (0.56–0.73)) HIV prevalence regions compared to others. Living in middle and high HIV prevalence regions was associated with lower odds of expressing prejudice toward PLWH (OR = 0.84, 90%CI = (0.71–0.99) and OR = 0.60, 90%CI = (0.45–0.80), respectively) among people living without HIV. Notably, PLWH living in high prevalence regions had higher odds of reporting internalized stigma (OR = 1.48, 90%CI = (1.02–2.14)) compared to those living in low prevalence regions. CONCLUSIONS: The results indicate that among people not living with HIV, subnational HIV prevalence was negatively associated with discriminatory attitudes and prejudice towards PLWH, but HIV prevalence was positively associated with self-reported internalized stigma among PLWH. These results provide insight on how resources could be invested to reduce HIV related stigma among both PLWH and those not living with HIV. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12889-023-16545-3.