OBJECTIVE: The aim of this study was to assess mortality and causes of death in subjects with visual impairment caused by diabetic retinopathy (DR). Only limited data are available concerning the mortality of subjects with DR, and to our knowledge, no data so far have been published on the mortality of subjects with visual impairment caused by DR. RESEARCH DESIGN AND METHODS: We identified 34 men and 73 women living in northern Finland with visual impairment caused by DR on 31 December 1993. The median age of the subjects was 71 years (range 27-88). The mortality of these 107 diabetic subjects was followed up for 4 years, until 31 December 1997, and compared with the mortality rates of 3 age- and sex-matched control groups. The first control group consisted of subjects treated for DR by laser coagulation from 1990 to 1993. The second control group consisted of diabetic subjects who had had fundus photographs taken from 1991 to 1992. The third control group comprised nondiabetic subjects selected from the population register. Information on deaths was obtained from official death certificates. RESULTS: A total of 91 diabetic and 10 nondiabetic subjects died during the follow-up. Of the deaths, 51 occurred in the subjects with visual impairment caused by DR, with a 4-year mortality rate of 477/1,000 (95% CI 382-571/1,000). Mortality rates were 224/1,000 (145-303/1,000) for the diabetic subjects with retinopathy treated by laser coagulation; 150/1,000 (82-217/1,000) for the diabetic subjects who had undergone fundus photography; and 94/1,000 (46-165/1,000) for the nondiabetic subjects. Cardiovascular diseases were the underlying cause of death in 55% of the subjects with visual impairment. Nephropathy was mentioned as the immediate cause of death for only 10% of the subjects. Compared with the nondiabetic control subjects, the odds ratios (ORs) for all-cause mortality were 5.1 (2.6-11) in the diabetic subjects with visual impairment caused by DR, and 5.6 (2.1-19) for mortality caused by diseases of the circulatory system. The ORs for all-cause mortality were 2.4 (1.1-5.6) in the diabetic subjects with retinopathy treated by laser coagulation and 1.6 (0.68-4.0) in the diabetic subjects with fundus photographs taken. CONCLUSIONS: The survival of diabetic subjects with visual impairment caused by DR was poor. The high mortality rate was attributed mainly to cardiovascular diseases. Therefore, severe retinopathy proves to be a risk marker of cardiovascular death in diabetic patients.

No abstract