The Role of Leisure-Time Physical Activity in the Change of Work-Related Stress (ERI) over Time

Background: Every second employee in Europe complains about work-related stress. Occupational stress due to an imbalance between efforts spent and rewards gained (effort-reward imbalance = ERI) is well investigated and it is associated with mental and physical health. A common guess is that leisure-time physical activity (LTPA) has beneficial effects on work-related stress. Yet, evidence in support of this assumption is weak, especially regarding ERI-stress. Longitudinal studies investigating the role of LTPA on ERI are missing. Therefore, this study aims to investigate the effect of LTPA on work-related stress by ERI over time. Methods: 3961 socially insured employees that were born in 1959 or 1965 and working in the first (t(1): 2011) and second wave (t(2): 2014) of the lidA-study were included. Work-related stress was measured by ERI, LTPA by the self-rated weekly frequency of physical activities. Besides the direct effect, a moderating effect of LTPA on ERI over time was tested in the multiple linear regression analysis. Results: The ERI at t(1) was strongly associated with ERI at t(2). While LTPA had no direct effect on ERI(t(2)), it was a significant moderator of ERI from t(1) to t(2): The higher the frequency of LTPA, the lower ERI was over time. This interaction of LTPA with ERI remained after adjustment for socio-demographic factors. Conclusions: The long-term moderating effect of LTPA on ERI is in agreement with former investigations on the role of LTPA on work-related stress, generally, and on its cross-sectional effect on ERI-stress, specifically. Some of Hill’s criteria of a causal association in epidemiology (biological gradient, temporality, consistency) support our findings. As LTPA has also been shown to exert a protective effect on health outcomes that are associated with ERI, the moderation of ERI by LTPA could partly explain this protective effect. Future observational and interventional studies are required to support our results over more than two age groups and study times.