The Swedish childhood diabetes study. An analysis of the temporal variation in diabetes incidence 1978-1987.

The Swedish Childhood Diabetes Register combines continuous prospective reporting on incident cases and regular retrospective review of hospital records in all paediatric departments in the country and has, from 1978 to 1987, compiled 3838 cases. A validity test using an independent source has revealed a degree of ascertainment of 100%. When fitting a linear regression model to the natural logarithm of the yearly incidence rates, a significant increase over time was found for girls and for both sexes combined. To analyse the time trend with respect to age, sex, calendar year and birth cohort and their interactions, Poisson regression modelling was used. There was a significant age effect and interaction between age and sex. When these age and sex effects were taken into account a significant increase over time was found. Calendar period had a significant effect when adjusted for cohort effects whereas cohort did not add significantly to the explanation of the variance. There was no significant interaction between sex and calendar period indicating similar time variations for both sexes. The temporal variability was also apparently similar among different age groups. The observed increase of childhood insulin-dependent diabetes mellitus (IDDM) affecting both sexes and all age groups, strongly supports the importance of environmental factors in the aetiology of IDDM.