Background: Whether workplace social capital affects employees' mental health is debated. We examined the association between work-unit aggregated social capital and incident purchase of psychotropic medications among employees.

Methods: We used data from the WHALE (Well-being in Hospital Employees) cohort study. The study population comprised 21,711 employees without recent psychotropic purchase-history nested within 2283 work units in the Capital Region of Denmark. Employees were invited to participate in a survey in March 2014 (86% response). We assessed workplace social capital by eight items (covering trust/justice and collaboration) and aggregated the mean of responses up to each work unit and categorized the scores into quartiles. Data on psychotropic purchases (antidepressants and anxiolytics/hypnotics/sedatives) were extracted via linkage to national registers. Using two-level mixed-effects survival models, we analyzed the association between work-unit social capital and psychotropic purchases during a one-year follow-up period adjusting for individual-level workplace social capital.

Results: Low work-unit social capital was associated with higher purchases of overall psychotropic medications in a dose-response manner (low-versus-high: HR=1.32, 95% CI=1.05-1.65), but this effect attenuated after adjusting for individual-level workplace social capital (HR=1.14, 95% CI=0.88-1.46). Low work-unit social capital was associated with higher purchases of antidepressants (HR=1.78, 95% CI=1.16-2.73) even after adjusting for individual-level workplace social capital (HR=1.69, 95% CI=1.05-2.73).

Limitations: Medical doctors/dentists were underrepresented in the data on workplace social capital.

Conclusions: Low work-unit social capital may be associated with higher use of antidepressants among health-care employees. Interventions to improve social capital could potentially promote mental health at work in the healthcare setting.