Job strain and the risk of depression: is reporting biased?

It is unknown whether the relation between job strain and depression reflects causal characteristics of the working environment or reporting bias. The authors investigated reporting bias by analyzing individual versus work-unit measures of job strain and the risk of depressive symptoms (n = 287) and a diagnosis of depression (n = 97) among 4,291 employees within 378 work units in Aarhus, Denmark, 2007. All participants reported psychological demands and decision latitude, and the authors estimated mean values for each work unit. The odds ratios predicting depressive symptoms or a diagnosis of depression for the highest versus the lowest levels of individual, self-reported high psychological demands and low decision latitude were significantly increased above 2.5. When participants were classified by the work-unit mean levels, these associations were substantially smaller. For depressive symptoms, the odds ratios were 1.49 (95% confidence interval (CI): 0.88, 2.53) and 1.08 (95% CI: 0.84, 1.39), respectively, for psychological demands and decision latitude. For a diagnosis of depression, the odds ratios were 1.33 (95% CI: 0.57, 3.09) and 1.02 (95% CI: 0.68, 1.56), respectively, for psychological demands and decision latitude. These findings indicate that reporting bias inflates associations between job strain and the occurrence of depression, if studies rely on individual self-reports.