IDDM in Hungarian children: population-based clinical characteristic and their possible implication for diabetic health care. Hungarian Childhood Diabetes Epidemiology Study Group.

Following our 10-year retrospective study, prospective registration of all newly diagnosed children has started in 1989. Data of the primary source (hospital records) were validated using the central pharmacy register for insulin. Ascertainment rate was 96%. Clinical characteristics are analysed on the basis of data of 324 children diagnosed in 1989-1990. There were some regional differences in incidence with no relationship between incidence and the degree of urbanization. The duration of clinical symptoms before diagnosis was less than two months in 81% of the cases, two to four months in 17% of the children, and no symptoms were recorded in 2% of the cases. The mean reported weight loss was 3.3 kg (range 0 to 16 kg), blood glucose at diagnosis ranged between 8.5 and 80.0 mmol/l (mean 25.4 mmol/l). Ketosis was noted in 86%, and 40% received infusion therapy at onset. There was no correlation between age, initial, blood glucose, ketosis, severity of the clinical condition and duration of symptoms. IDDM was reported in 4.7% of first degree relatives. Birth order was a significant risk factor, firstborn children were less at risk than subsequent siblings (P less than 0.05); the number of children in diabetic families was similar to those in the general population (2.1 vs. 1.8). The high mean blood glucose and ketosis indicate a relatively severe metabolic decompensation at the time of diagnosis and call for further improvement in the diagnostic acumen of the pediatric community.