Methods: Data were drawn from the Ugandan violence against children survey collected utilizing a multi-stage cluster sample design. We aggregated the data to the district level (N=136) and used Multivariate Clustering tool in ArcGIS Pro to assess the possibility that districts in Uganda can be differentiated based on their level of CSV, as measured by four proxies: experienced any unwanted sexual touching in childhood, experienced any unwanted attempted sex in childhood, experienced any physically forced sex in childhood and experienced any pressured sex in childhood. A non-spatial regression model was conducted to examine the association between the geographical patterns of histories of CSV on HIV vulnerabilities.
Results: We found three clusters of histories of CSV: 16.9% (n=23) of districts had severe histories of CSV, while 36.03% (n=49) had moderate histories of CSV and 47.06% (n=64) had low histories of CSV. Non-spatial regression analyses showed that compared to districts with low histories of CSV districts, districts with severe and moderate patterns of CSV histories were positively associated with multiple sexual partners, transactional sex engagement, condomless sex, STI symptoms, and suboptimal HIV testing.
Conclusion: Findings point to the need for multi-component interventions that screen and concurrently address adverse childhood experiences and HIV vulnerabilities. Specifically, there is a need for targeted interventions in districts with severe and moderate CSV histories.