Objectives This study aimed to examine the association between job strain and incident coronary heart disease (CHD) in Denmark, while accounting for changes of job strain.

Methods We included all employees residing in Denmark in 2000, aged 30–59 years with no prevalent CHD (N=1 660 150). We determined exposure to job strain from 1996–2009 using a job exposure matrix (JEM) with annual updates. Follow-up for incident CHD was from 2001–2010 via linkage to health records. We used Cox regression to calculate hazard ratios (HR) and 95% confidence intervals (CI) for the association between job strain and incident CHD.

Results During 16.1 million person-years, we identified 24 159 incident CHD cases (15.0 per 10 000 person-years). After adjustment for covariates, job strain in 2000 predicted onset of CHD during a mean follow-up of 9.71 years (HR 1.10, 95% CI 1.07–1.13). When analyzing changes in job strain from one year to the next and CHD in the subsequent year, persistent job strain (HR 1.07, 95% CI 1.03–1.10), onset of job strain (HR 1.20, 95% CI 1.12–1.29) and removal of strain (HR 1.20, 95% CI 1.12–1.28) were associated with higher CHD incidence compared to persistent no job strain. Associations were similar among men and women.

Conclusions Job strain is associated with a higher risk of incident CHD in Denmark. As we used a JEM, we can rule out reporting bias. However, under- or overestimation of associations is possible due to non-differential misclassification of job strain and residual confounding by socioeconomic position.