Objectives Since previous studies have only used past or current medical history of disease, there is no information on newly diagnosed disease in relation to job loss. Our objective was therefore to investigate whether newly diagnosed chronic disease increased job loss among middle-aged Japanese.

Methods We analysed data on 31 403 Japanese workers aged 50–59 years from a nationally representative longitudinal study. We defined two types of job loss; later job loss (within 1 year of disease diagnosis) and concurrent/later job loss (at around the time of diagnosis and within 1 year of diagnosis). Generalised estimating equation models were used to calculate ORs for job loss among current workers after a new-diagnosis of chronic disease (diabetes, hypertension, hyperlipidaemia, heart disease, stroke and cancer), using a discrete-time design and adjusting for demographic, socioeconomic and health behavioural factors. We used inverse probability weighting to account for non-response at follow-up.

Results ORs for concurrent/later job loss were 1.17 (95% CI 1.03–1.31) for diabetes, 1.01 (95% CI 0.93–1.09) for hypertension, 1.01 (95% CI 0.94–1.09) for hyperlipidaemia, 1.21 (95% CI 1.06–1.40) for heart disease, 1.48 (95% CI 1.21–1.81) for stroke and 1.38 (95% CI 1.17–1.62) for cancer diagnosed patients. The corresponding ORs for later job loss were 1.14 (95% CI 0.96–1.35), 1.00 (95% CI 0.90–1.12), 0.98 (95% CI 0.88–1.08), 1.24 (95% CI 1.02–1.49), 1.16 (95% CI 0.86–1.55) and 1.39 (95% CI 1.11–1.73).

Conclusions New diagnosis of diabetes, heart disease, cancer and stroke increased the risk of job loss; loss was concurrent for diabetes mellitus and stroke, and concurrent/later for cancer and heart disease. Our results provide fundamental information for employment support for patients with chronic disease.