Abstract
BACKGROUND: Black/African American men experience a disproportionate burden of type 2 diabetes (T2D) and high exposure to discrimination, yet it is unclear whether social networks buffer discrimination’s effects on mental health. This study tested moderation by network characteristics among Black/African American men with T2D. METHODS: In a cross-sectional online survey (February-June 2024), 1,225 Black/African American men with T2D completed validated measures of everyday discrimination and healthcare discrimination, mental health, and egocentric network inventories elicited via multiple name generators. Network indicators included mean support, percentage of very supportive members, general communication frequency, and diabetes-specific communication frequency. Multiple regressions with interaction terms estimated direct and moderating effects, controlling for demographics. RESULTS: Both perceived everyday (β = 0.72, 95% CI [0.67, 0.78], p < .001) and healthcare (β = 2.74, 95% CI [2.50, 2.99], p < .001) discrimination were associated with elevated levels of depressive symptoms, anxiety, and stress. Regarding network characteristics, a higher percentage of very supportive network members was associated with better mental health (everyday models: β=-4.97, 95% CI [-7.67, -2.28], p < .001; healthcare models: β=-5.00, 95% CI [-7.79, -2.20], p < .001). More frequent communication was associated with worse mental health: diabetes-specific communication (everyday models: β = 2.53, 95% CI [1.94, 3.13], p < .001; healthcare models: β = 2.28, 95% CI [1.65, 2.92], p < .001) and general communication (everyday models: β = 2.04, 95% CI [1.34, 2.74], p < .001; healthcare models: β = 1.94, 95% CI [1.20, 2.68], p < .001). Modest interactions emerged for everyday discrimination with diabetes-specific communication (β = 0.07, 95% CI [0.03, 0.11], p = .002) and with general communication (β = 0.05, 95% CI [0.00, 0.11], p = .04). CONCLUSIONS: Perceived discrimination was consistently associated with poor mental health. Networks characterized by a higher proportion of very supportive alters were associated with better mental health, whereas greater communication frequency, including diabetes focused exchanges, was associated with poor mental health. These findings indicate that interventions should prioritize improving the quality and cultural responsiveness of support rather than increasing contact volume. Longitudinal studies are warranted to elucidate causal mechanisms and to clarify the content and context of network communication.