Impact of shift work on the risk of depression

• Published in: Chronobiology international 2021-12, Vol.38 (12), p.1761-1775
• Main Authors: Behrens, Thomas, Burek, Katarzyna, Rabstein, Sylvia, Wichert, Katharina, Erbel, Raimund, Eisele, Lewin, Arendt, Marina, Dragano, Nico, Brüning, Thomas, Jöckel, Karl-Heinz
• Format: Article
• Language: English
• Subjects: Chronotype; depressive symptoms; mental health; night work; over-commitment; work stress
• ISSN: 0742-0528; 1525-6073
• DOI: 10.1080/07420528.2021.1962903

We studied the association between shift work and depressive symptoms in the prospective Heinz Nixdorf Recall Study, considering various demographic, lifestyle and work-related factors. Depressive symptoms were assessed using the Center for Epidemiologic Studies-Depression (CES-D)-Scale (≥17 points defined as high symptoms) and the Patient Health Questionnaire (PHQ) with a cutoff ≥9, or prescription of an anti-depressant. The definition of shift work included work hours outside 7:00 to 18:00, whereas night work was defined as a shift including work between 0:00 and 5:00. Poisson regression with robust error variances was calculated to estimate relative risks (RR) and 95% confidence intervals (CI), adjusted for age at follow-up, diurnal preference, monthly household income and education. Analyses were stratified by sex. We performed various sensitivity and stratified analyses to test the robustness of our results. At baseline, 1,500 gainfully employed subjects, 45-73 years of age and without a history of depression, were included. Until the 5-year follow-up, 896 participants were observed, and 486 participants survived through the 10-year follow-up. Although most analyses did not reach the level of formal statistical significance, women working night shifts tended to show increased relative risks for depressive symptoms according to the PHQ (RR = 1.78; 95% CI 0.71-4.45), in particular when working night shifts for ≥20 years (RR = 2.70; 95% CI 0.48-15.4). Stratification by age group revealed no increased risks among women above 60 years of age. Stratified analyses indicated that over-commitment was associated with higher risks for depressive symptoms among women (RR = 4.59; 95% CI 0.95-22.2 in the CES-D and RR = 12.7; 95% CI 2.89-56.1 in the PHQ). Exclusion of subgroups for the purpose of sensitivity analyses generally strengthened associations in women, whereas little evidence for an increased risk of depression remained among male shift workers. In summary, negative effects on depression were suggested among female shift workers, although results were based on small numbers. Among men, we did not identify consistently increased risks for depressive symptoms in relation to shift work.