Ecological association between Grindr user characteristics and HIV prevalence in Malaysia: a cross-sectional study.

Abstract

Objective: Men who have sex with men in Malaysia face a high HIV burden, but demographic, behavioral, and digital health data remain limited by social constraints. This nationwide, cross-sectional ecological study utilized a grid-based geographic sampling method to systematically collect publicly accessible Grindr profiles across all Malaysian states and federal territories. Population-normalized Grind user rates, profile characteristics, and regional people living with HIV rates were analyzed using Spearman's rank-order correlation to evaluate whether geosocial networking platform data offer insights into regional HIV epidemiology.

Results: A significant positive correlation was observed between population-normalized Grindr user rates and PLHIV rates (r_s = 0.5821, P = 0.0252). While unadjusted analyses showed nominal positive correlations between regional PLHIV rates and specific profile characteristics, none maintained significance following false discovery rate correction (Q = 5%). This lack of post-correction significance is driven by the limited statistical power of the regional sample size (n = 16) and high percentages of profile non-disclosure. Thus, while grid-based sampling systematically assesses digital user density, individual profile characteristics cannot reliably indicate regional HIV prevalence when data omissions are widespread.