Work characteristics as predictors of diabetes incidence among apparently healthy employees.

The objective of this article was to investigate the associations of the Job Demand Control-Support (JDC-S) model's components, job demands, job control, and work social support, as well as their interactive terms, with the risk of Type 2 diabetes. Participants were apparently healthy 5,843 men and women who underwent routine health checks at two points of time, about 41 months apart from one another. New cases of diabetes (N = 182) during follow-up period were defined based on fasting glucose value ≥ 126 mg/dl or glycosylated hemoglobin value ≥ 6.5% or self-reported physician diagnosis of diabetes and taking medications to treat it. The measures for assessing workload (representing job demands), job control, and work social support were all based on validated scales constructed to test the JDC-S model. In testing the hypotheses, we used logistic regressions and controlled for well-established risk factors for diabetes, including sociodemographic, physiological, and behavioral risk factors. We also controlled for depressive symptoms. The hypothesis that the higher the baseline levels of work social support, the lower the risk of diabetes, was supported (Odd Ratio = .78, significant at the p < .05 level). In an exploratory analysis, workload was found to have a U-shaped relationship with diabetes risk. We did not find direct effects of job control nor of any interactive term including the JDC-S model components on diabetes risk. Work social support is a protective factor, reducing the risk of diabetes. Both underload and overload may increase the risk of diabetes.