Workplace Levels of Psychosocial Factors as Prospective Predictors of Registered Sickness Absence Karl Bang Christensen, PhD Martin L. Nielsen, MD Reiner Rugulies, PhD Lars Smith-Hansen, BaTechnSoc Tage S. Kristensen, DrMedSci Objective: We sought to investigate whether workplace levels of psychosocial work environment factors predict individual sickness absence. Methods: Data were collected in a prospective study in 52 Danish workplaces in three organizations: municipal care, technical services, and a pharmaceutical company. Psychosocial factors were aggregated as workplace means. We used multilevel Poisson regression models with psychosocial factors as predictors and individual level sickness absence from absence registries as outcome. Results: High workplace levels of decision authority predicted low sickness absence in the technical services (rate ratio = 0.66, 95% confidence interval = 0.51– 0.86) and high workplace levels of skill discretion predicted low sickness absence in the pharmaceutical company (rate ratio = 0.74, 95% confidence interval = 0.62– 0.88) after control for relevant confounders. Workplaces in municipal care did not differ with respect to the psychosocial factors. Conclusions: Psychosocial factors at the work- place level may be important predictors of sickness absence. (J Occup Environ Med. 2005;47:933–940) A bsence from work because of sick- ness has considerable negative ef- fects for employees, employers, and the society. Sickness absence is a strong predictor of disability pen- sioning, 1,2 as well as morbidity and mortality. 3,4 Psychosocial work environment factors have been found to predict sickness absence after adjustment for relevant confounders in six prospec- tive studies. 5–10 These studies are predominantly based on the de- mand– control–support model 11 and have consistently found that a high level of decision authority is related to a low level of absence. Skill dis- cretion, the other variable in the job control/decision latitude concept, was tested in three articles and two found significant associations. Four of the six articles showed a signifi- cant association between a high score on some measure of social support and low rates of absence, although in two of these studies, this was only found in men. No consis- tent associations with psychological demands at work were found; in some studies high demands predicted high absence, in other studies it pre- dicted low absence, particularly in men. Beyond the demand– control– support model, other psychosocial factors, such as low job satisfaction, 5 high job insecurity and downsizing, 7 and low organizational justice, 8 have been found to be significant predic- tors of absence. Recently, our research group ana- lyzed prospective data on individual psychosocial work environment de- From the National Institute of Occupational Health, Copenhagen, Denmark. Address correspondence to: Karl Bang Christensen, PhD, National Institute of Occupational Health, Denmark, Lerso Parkalle ´ 105, DK-2100 Copenhagen, Denmark; E-mail: kbc@ami.dk. Copyright © by American College of Occupational and Environmental Medicine DOI: 10.1097/01.jom.0000172864.16001.85 JOEM • Volume 47, Number 9, September 2005 933