Background: This study tests associations between psychosocial stress at work measured by the effort-reward imbalance model in a dynamic perspective, and multiple indicators of poor mental health, in a prospective design.
Methods: 1986 male and female employees from four Belgian enterprises were followed-up over one year within the framework of the Somstress study. Based on two consecutive measurements, an index of cumulative job stress was constructed and its associations with five indicators of mental health were studied, excluding caseness at entry (for depression, anxiety, somatisation, chronic fatigue and psychotropic drug consumption respectively). Taking into account the longitudinal design, four categories of job stress are defined: 1) employees free from stress at both measures, 2) job stress present at first measure but not at the second one, 3) recent onset of job stress as evidenced by second measure 4) workers exposed to stress at both measures. Multivariate logistic regression with appropriate adjustments was applied.
Results: In bivariate analysis, a clear graded association of cumulative job stress with all five mental health indicators is observed, both in men and women. In multivariate logistic regression analysis, recent onset of stress is strongly associated with poor mental health among men (odds ratios ranging from 1.8 to 4.6), while cumulative stress shows strongest effects on mental health in women (odds ratios ranging from 1.4 to 7.1).
Conclusion: Cumulative experience and recent onset of job stress in terms of high effort spent and low reward received is associated with elevated risk of all five indicators of poor mental health at follow-up in a large cohort of employees.