Shift work, job strain and changes in the body mass index among women: a prospective study

Objectives The effects of job strain and shift work on weight gain have not been studied jointly. Cross-sectional and longitudinal studies on shift work and weight gain have reported different results. This study examines potential effect modification by job strain on the link between shift work and weight gain, and concurrent and delayed effects of shift work on weight gain. Methods Data came from 52 622 women who participated in the Nurses' Health Study II, a prospective cohort study. Using linear regression, we modelled change in body mass index (BMI) over 4 years as a function of change in job strain, cumulative exposure to rotating night shift previously and during the 4 years (ie, previous and concurrent exposures) and the interaction between job strain and concurrent shift work exposure. Age, race/ethnicity, pregnancy history, baseline BMI, job types and health behaviours at baseline were controlled for. Results Job strain and rotating shift work, concurrent and previous, all had independent associations with BMI change during the 4-year period. There was no evidence for effect modification by job strain. Concurrent and previous exposures to rotating night shift had different associations with BMI change: an inverted U-shape for concurrent exposure (ranging from 0.01 to 0.14 kg/m(2) increase), a dose-response for previous exposure (-0.02 to 0.09 kg/m(2)). Conclusions Job strain and rotating night shift work have independent contributions to weight gain. Reducing job strain and supporting night shift workers are both important intervention goals.