The role of stress in the onset of type 2 diabetes is a widespread lay belief, yet observational studies have produced inconsistent results. This study aimed to test the hypothesis that the association between perceived stress and incident diabetes might depend on occupational status (OS).

The four-item Perceived Stress Scale (PSS-4) was completed at baseline by 22,567 participants in the labour force (16,193 men, 6374 women; mean age: 44.5±9.8 years) who had undergone two health checkups subsidized by the French national healthcare system. All subjects were free from diabetes at baseline, defined as a fasting blood glycaemia≥7mmol/L or the use of antidiabetic drugs.

After a mean follow-up of 5.3±2.1 years, 527 participants (2.3%) had incident diabetes. After adjusting for sociodemographic, behavioural and biomedical risk factors as well as self-rated health, the association between baseline perceived stress and diabetes at follow-up was non-significant for the total study population. However, perceived stress was significantly associated with incident diabetes in participants of low OS [odds ratio (OR) for a five-point increment: 1.39; 95% confidence interval (CI): 1.02–1.90]. In contrast, there was a negative association between perceived stress and diabetes among those of high OS (OR: 0.60; 95% CI: 0.41–0.88) and no association within other occupational categories. The interaction between perceived stress and OS was significant (P<0.01).

This study suggests that the association between perceived stress and diabetes onset is dependent on OS. Furthermore, this association does not appear to be explained by the classical risk factors for diabetes.