Introduction: No epidemiological data exist concerning the influence of an earthquake on the risk of stroke. Whether the incidence of cerebrovascular stroke increased after the 1995 Hanshin-Awaji earthquake (EQ) in Japan and whether seismic intensity affected stroke risk dose-dependently was examined.
Methods: A retrospective cohort study was conducted among residents, who were living in two towns on the island of Awaji and were participants of the National Health Insurance (NHI) program. The two towns were divided into 11 districts and their respective damage and socioeconomic states were investigated. Reviewing the NHI documents issued before and after the EQ, people who had strokes (9th International Classification of Diseases, codes 430-431 or 433-434.9) were identified. Risk of stroke in relation to the seismic intensities, was assessed with the Cox proportional hazard model.
Results: Among subjects aged 40 to 99 years, 45 of 8,758 (0.514%) had a stroke the year before the EQ 72 of 8,893 (0.810%) had a stroke in the first year following the EQ, and 49 of 8,710 (0.566%) had a stroke in the second year following the EQ. In districts where the earthquake's intensity was < or = 9.5 on the modified Mercalli intensity (MMI), compared with the year prior to the EQ, the relative risk (RR) of stroke was 2.4 (95% confidence interval (CI) = 1.1, 5.0) in the first year following the EQ, after adjusting for age, gender, and income. In that year, compared with MMI of < 8.5-9.0, RRs for 9.0-9.5 and > or = 9.5 were 1.6 (CI = 0.9, 2.1) and 2.0 (CI = 1.1, 3.7), respectively (p for trend 0.02). No trend for the RR was observed in the year before the EQ or in the second year following the EQ.
Conclusion: The incidence of stroke increased in the first year following the EQ. The increase was associated with seismic intensity in a dose-response manner. Results suggest a potential threshold for RR of > or = 2.0 in areas near 9.5 on the MMI scale.