Aims: To investigate the major aspects of mortality in patients with noninsulin-treated type 2 diabetes mellitus (T2DM), from 1942 till 2000.
Subjects and methods: We performed a retrospective study in 9698 noninsulin-treated T2DM patients, 5001 (51.6%) males and 4695 (48.4%) females, registered in Bucharest Diabetes Center and deceased between 1943 and 2000. For each patient the age at diabetes onset, disease duration, age at death, cause of death, sex, height and weight were recorded.
Results: The mean age at diabetes onset was 58.3 +/- 9.1 years in 1943-1960 period (no significant differences by sex) and 60.6 +/- 10.3 years in 1981-2000 (59.3 +/- 10.3 years in males and 61.8 +/- 10.1 years in females, p < 0.01 vs. males). The mean disease duration at death was 7.7 +/- 5.2 years in 1943-1960 period (no significant differences by sex) and 11.3 +/- 8.1 years in 1981-2000 (11.9 +/- 8.4 years in males and 10.7 +/- 7.6 years in females, p < 0.01 vs. males). The mean age at death was 66 +/- 9.8 years in 1943-1960 period (no significant differences by sex) and 71.9 +/- 9.7 years in 1981-2000 (71.2 +/- 9.9 years in males and 72.5 +/- 9.5 years in females, p < 0.01 vs. males). In the Cox regression analysis, an increase in mortality was associated with the masculine sex--9.6% (CI 95% 1-19%, p = 0.028) compared with feminine sex; 1 year increase in age at onset--4.8% (CI 95% 4.3-5.3%, p < 0.01); 1 kg/m2 increase in body mass index--2.9% (CI 95% 1.9-3.8%, p < 0.01); 1 mg/dl increase in mean fasting blood glucose--0.1% (CI 95% 0-0.2%, p = 0.025). The major causes of death in noninsulin-treated T2DM patients in the 1981-2000 period were: ischemic heart disease (53.8%), stroke (14.4%), cancer (9%), digestive diseases (6.3%), diabetes (5.3%), end stage renal disease (4.6%), infections (2.7%), diabetes coma (2.2%) and others (1.7%).
Conclusions: There is a statistically significant increase in the proportion of death caused by ischemic heart disease, while infections significantly decreased in importance during the study period. The masculine sex, age at onset, mean fasting blood glucose and body mass index were all significant predictors of mortality in the Cox regression analysis, adjusted for the year of death.