Diabetes Care, Vol 23, Issue 3 290-294, Copyright © 2000 by American Diabetes Association
Mortality in patients with childhood-onset type 1 diabetes in Finland, Estonia, and Lithuania: follow-up of nationwide cohorts
T Podar, A Solntsev, A Reunanen, B Urbonaite, R Zalinkevicius, M Karvonen, RE LaPorte and J Tuomilehto
Hospital of Endocrinology, Clinicum of the University of Tartu, Estonia. toomasp@cut.ee
OBJECTIVE: To assess mortality of population-based cohorts of
childhood-onset type 1 diabetic patients from the Eastern European
countries of Estonia and Lithuania and compare this information with recent
data from Finland. RESEARCH DESIGN AND METHODS: Estonian (n = 518) and
Finnish (n = 5,156) type 1 diabetic cohorts were diagnosed between 1980 and
1994, and the Lithuanian (n = 698) cohort was diagnosed between 1983 and
1994. The mortality of these cohorts was determined in 1995. Life-table
analysis, Cox survival analysis with covariates, and standardized mortality
ratios (SMRs) were used. Causes of death were analyzed. RESULTS: Survival
after 10 years duration of type 1 diabetes was similar in Estonia (94.3%)
and Lithuania (94.0%), but much higher in Finland (99.1%). In the Cox
survival analysis with covariates, the country of origin and age at
diagnosis were found to be significant predictors of mortality. The SMR for
the Estonian cohort was 4.35 (95% CI 2.25-7.61), the highest for the
Lithuanian cohort was 7.55 (4.89-11.15), and the lowest for the Finnish
cohort was 1.62 (1.10-2.28). The most common cause of death in Estonia and
Lithuania was diabetic ketoacidosis (DKA), and in Finland, it was violent
causes. No deaths from late complications of diabetes have been documented
so far in any of the three countries. CONCLUSIONS: Our results demonstrate
a high rate of short-term deaths due to DKA and inferior survival of
childhood-onset type 1 diabetic patients in Estonia and Lithuania compared
with Finland. In Finland, the survival of childhood-onset type 1 diabetic
patients has improved and is only slightly inferior to that of the
background population.