Abstract

In February and May 1986, weekly samples of whole pasteurised milk were collected from the 24 dairies supplying seven British towns. A random sample of 96 women aged 25-64 was drawn from general practitioners' lists in each town, and catch specimens of early morning urine were collected by post from 194 subjects in February and from 186 subjects in May. Median milk iodine concentration was significantly greater in February (235 micrograms/l) than in May (119 micrograms/l). The median urine iodine concentration (expressed per g of creatinine) was also greater in February (106 micrograms/g) than in May (78 micrograms/g). There was a strong and statistically significant correlation between milk and urine iodine concentrations in the seven towns in February (Spearman's r = 0.79, p = 0.04). Within the towns, the change in milk iodine levels between February and May was positively associated with the change in the iodine:creatinine ratio over the same period. There is concern that an excess of dietary iodine may be harmful to some individuals. Should it prove desirable to reduce iodine intakes at the community level, the present results suggest that this could be achieved by a reduction in milk iodine levels, which can be readily brought about by adjusting the levels of iodine in cattle feed.