Reliability of retrospective self-report of shift work exposure in a large occupational cohort

Retrospective self-reports of shift work exposure in police employees showed 86% overall agreement with baseline (kappa = 0.72), with no evidence of asymmetric misclassification.

• Why it matters: Many studies rely on retrospective exposure recall, but reliability can vary by health status, potentially biasing associations with cancer or cardiometabolic outcomes.
• What they did: In a subgroup of the Airwave Health Monitoring study, participants reported shift work status at entry and retrospectively at follow-up; agreement was quantified overall and by later disease status.
• The result: Agreement was similar for later cancer (kappa = 0.76) but lower for later cardiometabolic disease (kappa = 0.56), and discordant reporting odds did not differ significantly by cancer (OR = 0.96) or cardiometabolic disease (OR = 1.85).