Summary. A systematic 20-year follow-up study of 1,221 diabetic patients was carried out in Osaka, Japan. The mean annual mortality rates were 2.55% for men and 1.64% for women. The ratios of observed to expected numbers of deaths were 1.50 for men and 1.39 for women, indicating an excess mortality for diabetic patients of both sexes, and higher mortality in men than in women. Factors that predisposed diabetic patients to premature death were early age of onset, albuminutia, diabetic retinopathy and fasting glucose level > 11.1 mmol/1 at the initial examination. Insulin dependence was also associated with poor prognosis. Cerebro-cardiovascular and renal diseases were the major causes of death in the diabetic patients; heart disease was the cause of death in 16.9%, cerebrovascular disease in 16.4% and renal disease in 11.9%. The relatively high incidence of renal disease as cause of death in diabetic patients was striking. Malignant neoplasms of liver and of pancreas and cirrhosis were also associated with increased ratio of observed to expected number of deaths in the patients.Key words: Prognosis of diabetic patients, long-term followup, excess mortality, causes of death, vascular complications, Japan.Studies on diabetic patients in Western countries have consistently shown a poor prognosis compared with the life expectancy of the general population [1][2][3][4][5][6][7][8][9]. This is due largely to late manifestation of cardiovascular complications of the disease [5][6][7][8]. Some previous studies suggest that there may be differences between diabetic patients in Japan and those in the West with regard to clinical manifestation and complications of the disease and causes of death [10][11][12]. However, detailed information is lacking on long-term prognosis in Japan. A long-term follow-up study of diabetic patients was started at our hospital in Osaka in 1962. The present report concerns the mortality rates and causes of death in our patients during the past 20 years.
Subjects and Methods
PatientsA total of 1,221 diabetic patients (705 men and 516 women) first seen at our hospital during the period 1962-1979 were studied. They were all Type 2 (non-insulin-dependent) diabetic patients, newly diagnosed (70.9%) or with duration of disease < 5 years at entry to the study (29.1%). They had fasting glucose level )6.7mmol/1 (120mg/dl) and/or 2 h glucose level i> 10.0 mmol/1 (180 mg/dl) in their first 50-g oral glucose tolerance test, in which venous blood samples were taken and plasma glucose determined by the ferricyanide method on a Technicon autoanalyzer. The mean + SEM age of the subjects at entry to the study was 50.5 + 0.4 years for men and 54.0 + 0.5 years for women (Table 1). All subjects were followed up to the end of 1981; the mean follow-up period was 7.5 years for both sexes.
MeNo~The patients were followed up by hospital outpatient computer, questionnaire, community record, or by a direct telephone call when all else failed, and their vital states (alive or dead) were confirmed. Since 1978, these proced...