Objectives: To examine relationships between shift work and markers (metabolic abnormalities) of insulin resistance (IR).
Methods: A cross-sectional study of 2,824 day and 826 shift workers. All the subjects were male blue-collar workers aged 30-59 years. Four IR markers [(1) hypertension (systolic blood pressure >or=140 mmHg or diastolic blood pressure >or=90 mmHg or under treatment for hypertension); (2) hyperglycemia (fasting serum glucose >or=7.00 mmol/l or under treatment for diabetes); (3) hypertriglyceridemia (fasting serum triglyceride >or=1.70 mmol/l or under treatment for hyperlipidemia); (4) hypo-HDL-cholesterolemia (fasting serum HDL-cholesterol <1.04 mmol/l)] were checked. Subsequently, IR syndrome, a cluster of IR markers, was expediently diagnosed by the number (N) of IR markers found in each worker. N>or=1, N>or=2 or N>or=3 was used as a cutpoint for the diagnosis. The prevalence of each IR marker and IR syndrome was compared between the two worker groups. Age, body mass index (kg/m(2)), job, drinking, smoking, and exercise were used as confounding factors. Job, work schedule and lifestyles were based on self-administered questionnaires.
Results: Hypertriglyceridemia (28.7% in day workers, 31.2% in shift workers) was most prevalent (4.6%), and hyperglycemia (5.4%) was least, in the four IR markers in both worker groups. Approximately half of the subjects had at least one IR marker in both groups (N>or=1; 46.0% and 46.6%, respectively). However, a full cluster of IR markers was rare in both groups (N=4; 0.2% and 0.1%, respectively). Prevalence of IR syndrome by any cutpoints increased with age in day workers, but shift workers aged 50 years or older had lower prevalence than shift workers younger than 50 years. In subjects younger than 50 years, all IR markers (except for hypo-HDL-cholesterolemia) and IR syndrome by any cutpoints were more prevalent in shift workers than in day workers. Contrariwise, in subjects aged 50 years or older, lower prevalence of all IR markers and IR syndrome in shift workers than in day workers was found. These results were not influenced by statistical adjustments for the confounding factors by multiple logistic regression analysis.
Conclusions: Shift work may be associated with IR syndrome in workers younger than 50 years. These relations may be underestimated mainly by broad definition of shift work and healthy-worker effects.