Background: The prevalence of sickness absence is particularly high among employees in health and social care, where psychosocial work stressors are pertinent. Managerial leadership is known to affect sickness absence rates, but the role leadership plays in relation to sickness absence is not fully understood; that is, whether poor leadership (i) is associated with sickness absence directly, (ii) is associated with sickness absence indirectly through the establishment of poor psychosocial working conditions, or (iii) whether good leadership rather has a buffering role in the association between work stressors and sickness absence.

Methods: Four biennial waves from the Swedish Longitudinal Occupational Survey of Health (SLOSH, 2010–2016, N=2333) were used. Autoregressive cross-lagged analyses within a multilevel structural equation modelling (MSEM) framework were conducted to test hypotheses i)–iii), targeting managerial leadership, register-based sickness absence and psychosocial work stressors (high psychological demands, poor decision authority and exposure to workplace violence).

Results: A direct association was found between poor leadership and sickness absence two years later, but no associations were found between leadership and the psychosocial work stressors. Finally, only in cases of poor leadership was there a statistically significant association between workplace violence and sickness absence.

Conclusions: Poor managerial leadership may increase the risk of sickness absence among health and social care workers in two ways: first, directly and, second, by increasing the link between workplace violence and sickness absence.