Objectives This study investigated the association between combined exposures and hyperglycemia incidence, as well as the dose–response relationship between the duration of night work and hyperglycemia among long-term night workers.

Methods In this prospective cohort study, 12 716 night workers from the nationwide population were recruited. Hyperglycemia incidence was based on the one-year change in fasting blood glucose levels. Occupational noise exposure was defined as exposure to 8-hour time-weighted average sound levels of ≥85 decibels. Personal factors, including body mass index, and work-related factors, like monthly night work duration, were assessed. Multivariable logistic and linear regression models were used to explore the association.

Results In the multivariate logistic analyses, each additional day of night work was associated with an increased risk of hyperglycemia [adjusted odds ratio 1.05, 95% confidence interval (CI) 1.02–1.07]. In the normal fasting glucose group, each additional day of night work was associated with a linear increase of +0.07% (95% CI +0.03% – +0.12%) in the change in fasting glucose levels, and noise exposure was associated with a linear increase of +1.34% (95% CI +0.55% – +2.12%) increase in fasting glucose levels. Furthermore, the population exposed to noise and working ≥10 days of night work had a significantly higher increase of fasting glucose levels (β +5.71%, 95% CI +4.48% – +6.95%), with significant interaction effects (P for interaction <0.01).

Conclusions The possible dose–response relationship between duration of night work and changes in fasting glucose levels was found. The combined exposure to night work and noise posed a higher risk for hyperglycemia than exposure to night work alone.

Key terms cohort study; exposure; hyperglycemia; interaction effect; night shift; night work; night worker; noise; population-based

• Supplementary material