Association of rotating shiftwork with preterm births and low birth weight among never smoking women textile workers in China.

1035 married women workers in three modern textile mills in Anhui, China were surveyed to investigate the association of rotating shiftwork with low birth weight and preterm birth in 1992. Information on reproductive health, occupational exposure history, and other covariates including age at pregnancy, time and duration of leave from job since pregnancy, and mill location was obtained by trained nurses with a standardised questionnaire. This analysis was limited to 845 women (887 live births), who were middle or high school graduates, never smokers, and non-alcohol drinkers. About 72% of the women worked an eight day cycle with shift changes every two days throughout pregnancy. Mean gestational age was 38.8 and 39.0 weeks for shift and regular schedule workers, respectively. Multiple linear regression was used to adjust for confounding factors including maternal age at pregnancy, order of live birth, mill location, job title, occupational exposure to dust/gases/fumes, stress, carrying and lifting of heavy loads, working in a squat position, time and duration of leave from the job since pregnancy, and indoor coal combustion for heating. The adjusted difference in gestational age associated with rotating shifts was statistically significant (beta = -0.44 (SE 0.20) weeks.) Mean birth weights were 3248 g and 3338 g for rotating shift workers and regular schedule workers respectively. The estimated effect of rotating shiftwork on birth weight was -79 (SE 42) g. When the analysis was restricted to first order live births or to production workers, the estimated effects of rotating shiftwork on both gestational age and birth weight were significant. The proportions of preterm birth (<37 weeks) and low birth weight (<2500 g), respectively, were 20% and 9% for shift workers and 15% and 6% for regular schedule workers. The adjusted odds ratio of shiftwork was 2.0 (95% CI) 1.1-3.4) for preterm birth and 2.1 (95% CI 1.1-4.1) for low birth weight. This association remained significant when the analysis was restricted to production workers or first order live births.