Background: Although vital statistics have indicated large Japanese-American differences in mortality rates for coronary heart disease (CHD), the magnitude of difference has not been documented well using comparable validation of cause of death.
Methods: Population-based fatal CHD data were compared between the Oita Cardiac Death Survey, Japan and the Atherosclerosis Risk in Communities (ARIC) Study, USA. Both studies (population: Oita City 198 093; the ARIC comunities 286 820) identified possible fatal CHD events (International Classification of Diseases, Ninth Revision [ICD-9]: 410-414, 250, 401-402, 427-429, 440, and 798-799) among residents aged 35-74 years during 1992-1993. Comparable criteria for classifying cause of death were applied. Sex-specific, age-adjusted mortality rates of CHD were calculated by place of death.
Results: In all, 330 deaths in Oita and 1398 in the ARIC communities had eligible ICD-9 death certificate codes; CHD codes (ICD-9 410-414) comprised 30.6% of investigated deaths in Oita and 58.6% in ARIC. For men, the non-validated rate ratio for CHD deaths (ARIC:Oita City) was 5.9 (95% CI : 4.2-8.5), which fell to 4.7 (95% CI : 3.5-6.4) with validation and inclusion of sudden deaths within one hour of onset as fatal CHD. For women, the overall non-validated rate ratio was 4.6 (95% CI : 2.8-7.6), which fell to 3.9 (95% CI : 2.4-6.3) with validation and but there was little further change when the sudden deaths were added.
Conclusions: Our results suggest that differences in fatal CHD rates between Japanese and Americans were not as large as suggested by vital statistics when events were validated and sudden deaths were included.