Background: The association between perceived workplace psychological stress, over the entire work career, and cancer among men has never been assessed. This was explored in the context of a population-based case-control study conducted in Montreal, Canada.

Methods: 3103 incident cancer cases (11 types) diagnosed in 1979-1985 and 512 population controls were interviewed. Subjects described in detail each job held during their lifetime, including the occurrence of stress, and its reason. Logistic regression estimated odds ratios (OR) and 95% confidence intervals (CI) for the association between perceived workplace stress and its duration, and each cancer site, adjusting for lifestyle and occupational factors.

Results: Employment in at least one stressful job was associated with increased odds of cancers of the lung (OR=1.33, 95% CI: 1.01-1.75), colon (OR=1.51, 95% CI: 1.15-1.98), bladder (OR=1.37, 95% CI: 1.03-1.81), rectal (OR=1.52, 95% CI: 1.10-2.10), and stomach (OR=1.53, 95% CI: 1.08-2.15). A duration-response trend was found for cancers of the lung, colon, rectum, stomach, and for NHL. Subjects reported changes in stress level over their career. Perceived stress was ascribed to several sources, including high demand and time pressure, financial issues, job insecurity, and hazardous conditions.

Conclusion: Prolonged exposure to perceived stress at work was associated with greater odds of cancer at 5 out of 11 sites. While over reporting of stress by cases cannot be fully ruled out, these associations, if substantiated, would bear important public health significance. Prospective studies building on detailed stress assessment protocols considering all sources and changes over the career are necessary.