We investigated the prevalence of stroke in Taiwan in an epidemiologic study of stroke, diabetes, and cardiovascular disease that used a two-phase survey design. The study population was drawn by cluster sampling and consisted of both urban and rural communities from four regions of Taiwan. There were 8,705 people 36 years of age or older interviewed during the period of October 1 to December 31, 1986, and 143 cases of completed stroke were later identified by a neurologist. The point prevalence rate for people aged 36 or older in our study was 1,642/100,000 population (95% confidence interval 1,389-1,942/100,000). Prevalence rates differed significantly among the four study regions and between urban and rural communities; prevalence was greater in northern Taiwan and in urban communities. Percentages of the major types of stroke in 143 stroke survivors were as follows: cerebral infarction 67.1% (96 cases), cerebral hemorrhage 14.0% (20 cases), subarachnoid hemorrhage 4.2% (six cases), and unclassified 14.7% (21 cases). Of the stroke survivors, 67.1% were independent in activities of daily living, and 75.5% were independent in ambulation. Hypertension, heart disease, diabetes mellitus, and a family history of stroke were significantly more common in stroke survivors than in strokefree individuals. {Stroke 1989;20:858-863) C erebrovascular disease ranked as the most common cause of death in Taiwan in the 20 years from 1963 through 1982. It is now the second cause of death, behind cancer.1 Heart disease, diabetes mellitus, and hypertension were fourth, fifth, and eighth, respectively, on the list of leading causes of death in 1984. To date, few epidemiologic studies concerning incidence, prevalence, and risk factors of these diseases have been conducted in Taiwan. 2 -3 In an attempt to conduct an epidemiologic investigation of these diseases, we began a study entitled "Epidemiologic Study of Stroke, Diabetes, and Cardiovascular Disease," and we present our results concerning prevalence, clinical features, and disability of stroke patients.
Subjects and MethodsSince stroke, heart disease, and diabetes affect chiefly the elderly, we selected our study population from people 36 years of age or older in both rural and urban communities. The family was used Received October 13, 1988; accepted January 8, 1989. as the study unit and consisted of people born before 1950. On the basis of geographic location, we randomly chose one rural and one urban community from the northern, central, southern, and eastern regions of Taiwan (Figure 1). Thus, a stratified cluster random-sampling method was used and yielded a total sample population of 13,930. Age, sex, and educational level representative of each community were tested before we began the survey. For the stroke survey, we used a two-phase design that included a door-to-door household survey and physical examination. The 88 interviewers consisted of public health nurses and student nurses who were specifically trained to ensure a high level of sensitivity and specifi...