Umeå universitets logga

BACKGROUND: This study examined whether sick leave due to severe stress (stress leave) and duration of leave are associated with future cancer risk.

METHODS: We conducted a matched case-control study using complete-population data from Swedish national registers (2005 to 2018), including 516,678 primary cancer cases and 2,357,433 matched controls. Odds ratios (OR) were calculated by conditional logistic regression and adjusted for pre-specified confounders.

RESULTS: Stress leave of any duration, reported to the Swedish Social Insurance Register, was associated with a slightly increased cancer risk, with the highest risk estimate for 1-30 versus 0 days (adjusted OR 1.05, 95% CI 1.02-1.09). In men, a clear exposure-response trend was present. We observed increased risks of prostate cancer (adjusted OR for > 90 days: 1.10, 95% CI 1.01-1.20) and cervical cancer (adjusted OR for > 90 days: 1.11, 95% CI 1.05-1.17, including cancer in situ). In etiology-based analyses, a positive association was found for smoking-related cancers, and the risk relationship for non-cervical HPV-related cancers was similar to that for cervical cancer. Risk estimates were above one for several types of stress in relation to overall cancer risk, including an exposure-response trend for acute stress reactions (p-trend 4.0 × 10-4) but a null association for post-traumatic stress disorder.

CONCLUSIONS: Stress leave was associated with a modestly higher risk of cancer overall and prostate and cervical cancers specifically. Regardless of whether the link is biological or reflective of lifestyle mediators or for cervical cancer, lower participation in screening, these findings suggest a potential relevance of severe stress for cancer prevention.