Date of Award


Document Type

Campus Access Thesis


Epidemiology and Biostatistics



First Advisor

James B. Burch


Police officers suffer from various adverse physical and mental health problems that may be mediated by stress, including effects imposed by shift work. Using data from the Buffalo Cardio-Metabolic Occupational Police Stress (BCOPS) cohort, this study examined effects of long term or short term shift work on measures of diurnal salivary melatonin or cortisol, and depressive or stress-related symptoms as measured by the Center for Epidemiologic Studies Depression instrument (CES-D) and Brief Symptom Inventory (BSI), respectively. Sleep symptoms were also ascertained. Long term shift work variables included cumulative afternoon, night, or combined hours, shift changes, or shift work status from 1994 until participation (1999 -2001). Short term shift work included similar measures summarized over the preceding 3, 5, 7, or 14 work days. Relationships between the melatonin or waking cortisol rhythm and the CES-D or BSI scores were also examined. Subjects collected serial saliva samples: on first awakening, and 15, 30, and 45 minutes after waking over one work day. The area under the curve with respect to ground (AUCg) summarized total cortisol output after waking, and the area under the curve with respect to increase (AUCi) characterized the waking cortisol response. The melatonin ratio was calculated as then ratio of first morning to afternoon plus baseline concentration. Generalized linear models were used to compare adjusted (least squares) means of dependant variables among participants grouped by quartile of quantitative shift work variables. Similar analyses compared shift workers (afternoons, nights, combined) with day workers. The melatonin ratio was not associated with long or short term shift work variables. Adjusted mean waking cortisol rhythms (AUCg) were reduced by up to 70% among night workers compared to those working day shifts over the preceding 3, 5, 7, or 14 days. Long term shift changes were associated with reduced CES-D scores, longer weekday sleep, and lower waking cortisol (AUCi) levels compared to those with fewer shift changes. Shift work over a 5-day period elicited maximal disruption of the waking cortisol rhythm. Shift changes over a period of years appear to influence duration sleep, depressive symptoms, and waking cortisol rhythms.