Temporal, seasonal, and geographical incidence patterns of Type I diabetes mellitus in children under 5 years of age in Germany

Abstract Aims/hypothesis. To estimate the national incidence of Type I (insulin-dependent) diabetes mellitus in children under 5 years of age in Germany and to analyse temporal, seasonal, and geographical patterns of the diabetes incidence. Methods. During 1993–1995 newly diagnosed subjects were prospectively registered by the hospital-based ’German Paediatric Surveillance Unit' with monthly inquiries in all paediatric departments in Germany. Level of ascertainment was estimated by capture-recapture-analysis using two independent regional data sources. Results. During 1993–1995 the national incidence was 8.10 (95 %-CI: 7.61, 8.61) per 100 000 person-years, ranging in-between lower rates in west European countries and higher rates in northern Europe. Degree of ascertainment was about 85 %. Male to female ratio was 1.11 (95 %-CI: 0.98, 1.25). Compared with results of previous regional studies in the east and the south-west of Germany a 3- and 1.3-fold incidence increase was observed, respectively. Multivariate Poisson regression analysis showed season, geographical region, and interactions of age at onset with sex and calendar year to be independent significant predictors of the incidence. Incidence variation by age was different between boys and girls. A significant incidence increase by calendar year was found in 3- and 4-year-old children only. In summer and fall the incidence was higher than in winter and springtime, in the northern parts of the country higher than in the southern parts. Conclusion/interpretation. This study reports first national incidence data of Type I diabetes in children under the age of 5 years in Germany. Observed marked temporal, seasonal, and geographical incidence variations strongly support the causal role of environmental factors in disease aetiology. [Diabetologia (1999) 42: 1055–1059] Received: 4 March 1999 and in revised form: 6 May 1999