The epidemiology of lost residual beta-cell function in short term diabetic children.

Using the country-wide Swedish childhood diabetes register 526 children, who had had diabetes for 6-30 months were traced for measurements of 24-hour urinary C-peptide. Lost beta-cell function was defined as a 24-hour urinary C-peptide excretion per kg body weight less than 10% of the mean for healthy children (less than 0.025 nmol/kg). The estimated cumulative incidence of lost beta-cell function was 0.64 at 30 months. The incidence of lost beta-cell function did not differ by sex. Neither was there any significant variation in season at onset for cases with lost beta-cell function. A significant age dependency was shown for the cumulative incidence of lost beta-cell function with the highest incidence in the young age groups, i.e. a reversed age dependency compared to that of clinical onset. In contrast to the clinical onset of diabetes no significant geographical variation was found for lost beta-cell function when comparing standardized morbidity ratios. The urinary C-peptide excretion was significantly correlated to age at onset but not to degree or duration of ketonuria at onset. It is concluded that there are striking differences when comparing the epidemiology of lost beta-cell function to that of clinical onset in terms of age, sex, seasonal and geographical variations. The timing of clinical onset may thus partly be determined by factors different from those determining the rate of fall in beta-cell function.