Abstract

A growing number of longitudinal studies report associations between adverse psychosocial factors at work and sleep problems, but the evidence regarding the direction of these associations and the effects of changes in exposure across time is limited. Based on a randomly drawn cohort from the general working age population in Norway, we tested the plausibility of normal, reversed and reciprocal associations between 10 psychosocial factors at work and sleep problems. In addition, we analysed if reduced exposure across time had the anticipated result of reducing the risk of sleep problems. Eligible respondents had an active employee relationship in 2009 and 2013 (N = 5760). We computed several sex-stratified logistic regression models with adjustments for a various plausible confounders, i.e. sleep problems at baseline, sociodemographic characteristics, chronic health problems, health behaviours, and organisational and mechanical factors at work. In addition to the commonly hypothesised causal relation between psychosocial stressors and sleep problems, we found reverse as well as reciprocal associations. There was one stressor only in which reversed causality could be ruled out, namely a robust effect of work-family imbalance and subsequent sleep problems among women (adjusted OR = 1.90; 95% CI = 1.37–2.64). Moreover, we found that reducing the exposure to effort-reward imbalance and low social support among men, and work-family imbalance among women were followed by a robust reduced risk of sleep problems (p < 0.05). This point to the potential health benefit of preventive measures aimed at reducing exposure to these three specific psychosocial stressors at work.