Epidemiological studies of childhood onset type 1 diabetes in Devon and Cornwall
A high quality register of children with type 1 diabetes, The Cornwall and Plymouth Children's Diabetes Register (CPCDR), has been successfully established in the far South West of England for this thesis. Children aged 0-15 years of age who were diagnosed with type I diabetes in the study area have been systematically registered on the database since 01 /01/1975 until the present day. Children are now registered prospectively, to provide a database of the disease in this region and to allow epidemiological studies. Basic descriptive studies on the occurrence of the disease based on the CPCDR between 1975-1996 have been carried out for this thesis. In addition, space-time clustering analysis of the disease has been conducted, as well as an exploration of the association of some chemical contents in domestic drinking water, serum minerals and birth weight to the risk of developing the disease. The incidence study demonstrated that the overall incidence rate of childhood onset type 1 diabetes was 14.9 cases/100 000/year in this area during the 22-year study period (1975-1996). The case ascertainment was 94.4% for the whole register. A significant increase (2.49% per year) of overall incidence has been observed, mainly due to a significant increase in the 0-4-year age group (6.32% per year). The incidence significantly differed among the 22 years with peak incidences seen in the years 1977, 1983, 1988, 1990 and 1993-94. Incidence increased with age, with a peak age of 12 years for girls and 14 years for boys. Girls had a significantly higher incidence than boys. The significant seasonal variations at diagnosis were detected with the peak incidence appearing in autumn and winter. Significant space-time clustering was found by the Knox test in the study region in the following combinations of critical cutoff values, 25, 35 and 50 km and 270 or 360 days (P values < 0.05), and 50 km and 90 days (P < 0.05) with the highest level of significance found at 35 km and 360 days (P < 0.01). Stronger clustering was found in the younger children (0-4 years). Therefore, there is evidence of space-time clustering in the onset of childhood diabetes in the far South West of England. These results lend some support to the hypothesis that viral infections in early life or other unknown environmental factor(s) may have a role to play in the development of childhood diabetes. The relationship between childhood diabetes and drinking water quality has been explored in the study area. The initial analysis with the Spearman's rank test and χ² test for trend on the tertiles of the dataset suggest that copper, magnesium and nitrate have some protective effect on children for developing type 1 diabetes. However, Poisson regression analyses revealed that zinc and magnesium were the main possible protective chemicals. The concentration ranges of zinc (22.27-27.00 µg/L) and magnesium (>= 2.61 mg/L) in domestic drinking water would significantly decrease the risk of childhood diabetes. The results indicated that zinc and magnesium in drinking water are protective factors against the development of childhood diabetes. No significant associations were detected between birth weight or serum minerals and the risk of developing the disease.