Nonstandard work hours have been associated with hormonal differences, possibly disturbing physiological functions that are circadian in nature. We investigated the relationship between rotating shift work and usual menstrual cycle patterns in the Nurses' Health Study II. Data from 74,436 participants who reported being premenopausal, having menstrual periods, and not using oral contraceptives were analyzed. Of these 8,519 (11.4%) and 5,239 (7.0%) reported working rotating shifts for 1-19 months and 20þ months, respectively, between 1991 and 1993. In 1993, 871 (1.2%) reported short menstrual cycle length (<21 days), 3,477 (4.7%) reported long cycles (40þ days) and 7,469 (10.0%) reported irregular cycle patterns (>7 days variability). Logistic regression was used to estimate the odds ratios (ORs) for these three outcomes, adjusting for age, parity, smoking, alcohol use, and physical activity. Rotating shift work was associated with an increased odds of having an irregular cycle pattern: 1-19 months, OR ¼ 1.2, 95% confidence interval (95% CI) ¼ 1.1-1.3; 20þ months, OR ¼ 1.4 (1.3-1.5). Rotating shift work was also associated with short cycles [ORs compared to normal cycles ¼ 1.3 (1.1-1.6) for 1-19 months and 1.4 (1.1-1.8) for 20þ months]. Long cycles were also more common among rotating shift workers [ORs for long compared to normal cycles ¼ 1.2 (1.1-1.4) for 1-19 months and 1.4 (1.2-1.5) for 20+ months]. When further adjusting for BMI, all ORs were slightly attenuated but still statistically significant. These data suggest that menstrual function is adversely affected by rotating shift work, possibly through circadian disruption.
American Journal of Epidemiology; 41st Annual Meeting Society for Epidemiologic Research Chicago, Illinois, June 24-27, 2008