Objective: To examine the construct validity of different formulations of job strain based on the demand-control(-support) model in the context of a hospital-based survey.
Methods: We defined the best measure of job strain as the one that would be better predicted by professional status and job characteristics and better predict health outcomes, stress and back-pain leave. We received 1,298 responses from collaborators at the University Hospitals of Geneva who responded to a questionnaire survey including all the constructs cited above and the Job Content Questionnaire.
Results: The difference between decision latitude and psychological demands (strain by subtraction), corresponded to the definition of the best measure of job strain and significantly, albeit weakly, predicted all outcomes. The logarithmic approach was the second best measure of job strain and was in fact a better predictor of stress.
Conclusions: We would encourage researchers to explore different formulations of job strain, in particular the subtraction approach, to compute strain from demands and decision latitude.