More than just days off: how work characteristics shape absence rate, frequency and duration

Purpose This study examines which psychosocial work characteristics, drawn from multiple theories of work stress and work design, retain predictive relevance for objectively recorded sickness absence when considered jointly rather than in isolation. Design/methodology/approach Using the Copenhagen Psychosocial Questionnaire (COPSOQ), 19 work characteristics were assessed in 1,935 employees from 63 organizational units in the German construction materials sector. Least absolute shrinkage and selection operator regression was applied to aggregated unit-level data to predict sickness absence rate, frequency, and duration recorded 12 months after baseline. Findings Results revealed distinct predictor constellations across absence metrics. Gender composition was the sole predictor of absence rate (R2 = 0.12). Absence frequency (R2 = 0.21) was additionally predicted by work–privacy conflict and social support, while absence duration (R2 = 0.24) was associated with role clarity, social support, quality of leadership, and trust and justice. Sensitivity analyses excluding gender confirmed that psychosocial predictors were robust and not reducible to demographic composition effects. Practical implications The findings inform targeted approaches to psychosocial risk assessment and workplace health management, suggesting that organizational climate and justice factors are particularly relevant for reducing prolonged absences. Originality/value Unlike prior research relying on narrow sets of work characteristics from single theoretical frameworks and treating absence as a unitary construct, this study integrates predictors from multiple work stress theories and demonstrates that sickness absence rate, frequency, and duration are shaped by different work characteristics.