Date of Award


Document Type

Open Access Thesis


Epidemiology and Biostatistics


The Norman J. Arnold School of Public Health

First Advisor

Michael Wirth


Background: Shiftwork is associated with many chronic diseases stemming from the disruption of circadian rhythms. Police officers have a high risk for many adverse conditions including depression. Abnormal bodily functions, such as increased stress and disturbed sleeping patterns, may play an important role in the development and severity of depression. This study was designed to assess the association between depressive symptoms and shiftwork in a police cohort from Buffalo, NY

Methods: This longitudinal analysis consists of 470 police officers from the Buffalo Cardio-Metabolic Occupational Police Stress (BCOPS) study. Data were collected during visit 3 between 2004-2005 and visit 4 in 2010. Officers were classified as working the day, evening, or night shift based on the shift they spent the most of their work hours according to electronic payroll records. Other shift work variables included short-term shiftwork over the previous two weeks and number of shift changes. Two self-reported depression measures were used as the outcomes, the Center for Epidemiological Studies – Depression (CES-D) scale and the Beck Depression Inventory (BDI). Generalized linear models were used to estimate least squares means and odds ratios from repeated data. Possible effect modification was only assessed for the stress measures.

Results: Linear regression models did not produce significant results. With logistic regression, those working the evening/night shift had higher odds for depressive

symptoms according to the BDI than those working the day shift (OR = 4.60, 95% CI = 1.15-18.39). Excluding the evening shift, those working the night shift had higher odds for depressive symptoms according to the BDI than those working the day shift (OR = 4.90, 95% CI = 1.20-19.57).

There were no significant results for short-term shiftwork or number of shift changes in relation to either depression measure. Crude analyses showed highest depressive symptoms in the evening shift for the linear and logistic models. Stratifying by the stress measures did not show significance, however mean values for depressive symptoms were higher in the high stress categories for Impact of Events (IES) and perceived stress score (PSS).

Discussion: Associations were only found in the logistic regression model in regard to long-term shiftwork and depressive symptoms assessed by the BDI. After stratifying by the stress measures, mean values for depressive symptoms were higher in the highstress categories. This may indicate that depressive symptoms are more influenced by stress than by shiftwork. Further research should include usage of biomarkers for depression, a longer study design, and assessment of total shiftwork history.

Included in

Epidemiology Commons