Drinking water composition and childhood-onset Type 1 diabetes mellitus in Devon and Cornwall, England


Dr B. A. Millward, Department of Molecular Medicine, Plymouth Postgraduate Medical School, Tamar Science Park, ITTC Building, Room S5, 1 Davy Road, Derriford, Plymouth, Devon, PL6 8BX, UK.


Aims  Previous studies have reported inconsistent results on the association between some compositions (e.g. nitrate) in domestic water and the risk of childhood-onset Type 1 diabetes mellitus. This study aimed to examine the relationship between nitrate, zinc and magnesium in drinking water and the risk of childhood-onset Type 1 diabetes mellitus.

Methods  The study covers the Cornwall and the former Plymouth Health Authority Regions in the far south-west of England. Five hundred and seventeen children, aged 0–15 years, diagnosed with Type 1 diabetes mellitus between 1975 and 1996, were identified for inclusion in the study. Domestic water data (nitrate, Zn, Mg, Cu, Al, Ca, Fe and Mn) between 1993 and 1997 were provided by South-west Water Plc, UK, for each of the 40 Water Supply Zones in which the subjects had been resident at the time of diagnosis. The standardized incidence ratio (SIR) of the disease was calculated for each Water Supply Zone using the UK 1991 census population data. The relationship between the SIR of the disease and the water quality indicators in thirds (three strata of low, medium and high concentrations) was examined by χ2 test for trend and Poisson regression analysis.

Results  The initial analyses by χ2 test for trend on the relation of SIRs and drinking water compositions suggested that copper, magnesium and nitrate might have some protective effects, but Poisson regression analyses showed that only zinc and magnesium were significant factors. The data suggest that the incidence rate of childhood diabetes is significantly lower when the concentrations of zinc and magnesium in the domestic drinking water are in the range 22.27–27.00 µg/l (incidence rate ratio (IRR), 0.76; 95% CI, 0.59–0.97) and greater than 2.61 mg/l (IRR, 0.72; 95% CI, 0.58–0.91), respectively.

Conclusions  Our findings suggest evidence of a possible association between zinc and magnesium in the domestic drinking water and childhood diabetes in the far south-west of England. However, these possible protective effects of zinc and magnesium in domestic drinking water warrant further confirmation.

Diabet. Med. 18, 709–717 (2001)