Effect on infant mortality of iodination of irrigation water in a severely iodine-deficient area of China

Abstract

Background: Hotien county in Xinjiang province, China, is an area of severe iodine deficiency and has a high infantmortality rate. We investigated whether iodine replacement through iodination of the irrigation water would decrease infant mortality.

Methods: We added potassium iodate to irrigation water over a 2 to 4 week period beginning in 1992 in certain areas of three townships (Tusala, Long Ru, and Bakechi). Logistic regression analysis was used to compare the odds ratios for infant and neonatal mortality in treated and intreated areas.

Findings: The median urinary iodine concentration significantly increased in women of child-bearing age from < 10 micrograms/L to 55 micrograms/L. Infant-mortality rates decreased in the treated areas of Long Ru (mean [SD] 58.2 [4.4] per 1000 births to 28.7 [7.1] per 1000 births), Tusala (47.4 [12.4] per 1000 births to 19.1 [1.5] per 1000 births), and Bakechi (106.2 [9.5] per 1000 births to 57.3 [7.3] per 1000 births). Similar results were also seen for neonatal mortality. On regression analysis iodine treatment and time were significant independent predictors of infant mortality.

Interpretation: Iodine supplementation of irrigation water in areas of severe iodine deficiency decreases neonatal and infant mortality. Iodine replacement has probably been an important factor in the national decrease in infant mortality in China.

PIP: Hotien county in Xinjiang province is an area of severe iodine deficiency which also has a high infant mortality rate (IMR). The authors investigated whether iodine replacement through iodination of irrigation water would decrease the level of infant mortality. Potassium iodine was added to irrigation water over a 2-4 week period beginning in 1992 in certain areas of Long Ru, Tusala, and Bakechi. The median urinary iodine concentration significantly increased in women of child-bearing age from less than 10 mg/l to 55 mg/l, while IMR decreased in Long Ru from a mean of 58.2/1000 births to 28.7/1000, in Tusala from 47.4/1000 to 19.1/1000, and in Bakechi from 106.2/1000 to 57.3/1000. Similar results were also observed for neonatal mortality. Iodine treatment and time were significant independent predictors of infant mortality.