Network structure of gambling-related cognitions and risk behaviors among sports bettors in the Democratic Republic of the Congo: a cross-sectional EBICglasso network analysis (N = 2261)

Background and aims: Sports betting is expanding rapidly in sub-Saharan Africa, yet the network structure of gambling-related cognitions and risk indicators remains understudied. This study analyzed these cognitions and their financial risk, psychological, and social correlates among DRC sports bettors. Methods: A cross-sectional sample of 2261 sports bettors (85.9% male; mean age = 25.59 years, SD = 5.79) was recruited at Kinshasa betting venues (March–May 2026). The Gambling-Related Cognitions Scale (GRCS; α = 0.924) was administered alongside financial, psychological, and social measures. Network structure was estimated using thresholded EBICglasso (γ = 0.5). Bridge centrality and sex-based invariance were assessed via NetworkComparisonTest (100,000 permutations). Results: The thresholded network retained 150 edges (density = 0.302). Guilt and G1 were the most central nodes. Bridge analysis identified guilt and social pressure as bridge nodes. Seven items showed zero bridge strength. Item-level network structure showed no global sex difference (M = 0.195, p = .121), whereas a 14-node sensitivity analysis detected an omnibus structural difference (M = 0.217, p = .022) not localized to any edge. After Bonferroni and FDR corrections, applied separately to 496 edge-level and 90 centrality-level tests, one edge (G1–G15), but no centrality index, differed by sex. Conclusions: This is among the first network analyses of gambling-related cognitions in sub-Saharan Africa, with Gambling Expectancies as the most connected cognitive domain, G1 as a central hub, and guilt and social pressure as bridge nodes. These findings extend the Interaction of Person-Affect-Cognition-Execution (I-PACE) model and inform intervention research.