There has been an overall decrease in stroke mortality in Japan. However, Dr. Omae said that the incidence rates of lacunar infarction and cerebral hemorrhage are still higher than in Western countries. He also discussed specific lifestyle and medical risk factors for cerebral infarction.

In 1965, age-adjusted stroke death rates in Japan were more than 200 per 100,000 persons, according to statistics from the World Health Organization. At that same time, stroke rates were much lower in Western countries such as Sweden and the United States.

Vital Statistics in Japan show that there was a sharp decline in stroke mortality between 1965 and 1990 and has been leveled off thereafter. This is mainly due to a decrease in mortality due to cerebral hemorrhage; Today, the death rate from stroke in Japan is not particularly high compared with those in western countries.

Several factors could be contributing to the precipitous decline in stroke mortality in Japan. The reduction in fatal stroke could be due to better control of hypertension or to improvement in nutritional conditions. It could also be due to improved care in the acute stage of stroke.

Certain stroke subtypes such as lacunar infarction are more common in Japan than in Western countries. One large, long-term Japanese population study provided data on incidence of stroke by type. The study investigators reported a 45 incidence of lacunar infarction in cases that occurred between 1961 and 1993. Another study looking at hospital admissions in Japan between 1978 and 1991 confirmed a high incidence of lacunar infarction. In comparison, reported lacunar infarction rates in the United States and Switzerland have ranged from 12% to 20%.Cerebral hemorrhage rates were 20% and 22%, respectively, in the Japanese population and hospital admission studies. Reported rates of cerebral hemorrhage in the United States and Switzerland are 11% to 12%.

The incidence of each stroke subtype in Japan may be changing over time. Some data suggest an increase in atherothrombotic infarction and cardioembolic infarction, with a reciprocal decrease in lacunar infarction.

The Japanese population study described above is a long-term, prospective evaluation in a rural town, Hisayama, that is adjacent to Fukuoka. The evaluation has taken place from 1961, when the population was approximately 6,500, to the present (population 7,500). Researchers evaluated only individuals at least 40 years of age in 3 separate time periods. 80% who died during the study period came to autopsy to identify the cause of death.

Investigators noted a reduction in stroke incidence between the earliest cohort and the second and third cohorts. There was no change in incidence of ischemic heart disease. Of stroke subtypes, both cerebral infarct and cerebral hemorrhage declined in incidence from the first to second cohort. There was no change in subarachnoid hemorrhage.

The investigators also looked at incidence of cerebral infarction by blood pressure group: normotensive, borderline hypertensive, and hypertensive. Stroke incidence increased with higher blood pressure for both men and women.

Hypertension occurred in 28% of men and 24% of women in the first cohort. The frequency of it decreased significantly for men in the second and third cohorts (24% and 23% respectively) but not significantly for women (24% and 21%).

Frequency of metabolic abnormalities increased over time despite the decrease in stroke incidence. The rate of obesity was relatively low in the earliest cohort, and significantly higher in both the second and third cohorts. Similarly, there were significant increases in hypercholesterolemia and abnormal glucose tolerance tests over time.

The incidence of lacunar infarction varied according to level of blood pressure. In individuals not receiving antihypertensive drugs, there was a stepwise increase in incidence of lacunar infarction starting at blood pressures of 140/90 mmHg and higher.

Multivariate analysis revealed that risk factors for cerebral infarction varied by subtype. Smoking was significant only for lacunar infarction. Only age and systolic blood pressure were significant for atherothrombotic infarction. Findings were fundamentally the same for men and women. Results for men are as follows:

The findings of the Hisayama study provide valuable insight into the character of stroke in Japan over time. Although much of Japanese lifestyle has become westernized, the risk factors and the clinical profile of stroke have not been as westernized.