Correlation of cerebro-cardiovascular findings to mortality of middle-aged and elderly in a population from a fishing village in Shizuoka Prefecture

Six hundred and eighteen men and 993 women ranging in age from 30-69, living in Kamo Village, Shizuoka Prefecture, underwent baseline health examinations in 1964-1966. During the follow-up period of 20-years, 175 men and 170 women died. The most frequent cause of death was malignant neoplasms (57 men and 45 women), followed by stroke (47 men and 44 women) and heart disease (29 men and 37 women). The relationship of 22 cerebro-cardiovascular disease variables investigated in the baseline examination to stroke and heart disease mortalities, and, in addition, to cancer and all-cause mortalities were analyzed using Cox's proportional hazard model. In univariate analysis controlled for age and sex, systolic and diastolic blood pressures, albuminuria, hypertensive and sclerotic changes in fundus oculi, body mass index, atrial fibrillation, and use of antihypertensive drugs had significant positive relationships to stroke mortality. As for heart disease mortality, albuminuria, glucosuria, high R wave, and ST and/or T changes in ECG were positively and significantly related. Only Q.QS in ECG significantly correlated with cancer mortality. Variables showing significant relationship to all-cause mortality were systolic and diastolic blood pressures, albuminuria, glucosuria, hypertensive changes in fundus oculi, Q.QS, high R wave, ST and/or T changes, atrial fibrillation, use of antihypertensive drugs, and history of stroke. In multivariate analysis (step-wise) of all examination variables including age and sex, stroke mortality was significantly related to age, atrial fibrillation, use of antihypertensive drugs, and hypertensive changes in fundus oculi. Age, albuminuira, and ST changes in ECG were significantly related to heart disease mortality. Age, sex, and Q.QS in ECG were significantly related to cancer mortality. The relationship of age, sex, albuminuria, Q.QS, and ST changes in ECG to all-cause mortality was significant.