The Effects of Iodized Salt on Cognitive Development in Ethiopia

MATERIALS AND METHODS Sample size was estimated on the basis of the Bayley Scales of Infant and Toddler Development III (Bayley) where the mean is 100 and the standard deviation is 15. Using an effect size of 0.20 standard deviation, a beta of .05 and power of .80, and accommodating for clustering within villages and attrition, the investigators require a sample size of 800 for each group (intervention and control). Severely anemic children (< 8 mg/L) will be referred for treatment and not included in the study.

In this longitudinal study the investigators will recruit 3200 6-mo old infants and 1600 children from each of two other age groups: 18-, and 60-months (inclusion criteria) and their mothers. The exclusion criteria: severe anemia (hemoglobin < 8g/dL). If children are too disabled to take a cognitive test, this will be noted but no cognitive testing administered.

From six zones around Bahir Dar, the investigators will randomly select 60 districts. Half will be assigned randomly to start receiving iodized salt in July 2011; the other half will receive iodized salt as it becomes available in the market. From each district, 1-2 villages will be randomly selected for recruiting 30 children from each age group. It is assumed that with a pregnancy rate of 3.7%, approximately 166 children are born every year in a village of 4500, and 13 are born every month. So within 3 months, starting in May 2011, 30 children of each age could be recruited.

With the help of Health Extension Workers, trained field workers will identify the participants : children of eligible age. Mothers will provide written consent for themselves and their children to participate.

Mothers will be interviewed at baseline and endline concerning the family's economic and demographic information, child's health and diet, and opportunities for stimulation of children (Home Observation for Measurement of the Environment Inventory; Center for Epidemiologic Studies Depression Scale symptom scores). Children's nutritional status will be measured with a food frequency questionnaire, length/height, weight, blood indicators for thyroid function (thyroglobulin, thyroxin, triiodothyronine) and iron status (hemoglobin, ferritin, soluble transferrin receptor), and inflammation (C-reactive protein), goiter, and urinary iodine. A maximum of 10 mL will be collected for blood by venipuncture. Finally, trained assistants will administer the Bayley Scales of Infant and Toddler Development III (cognitive, language, and fine motor scales) to children under-3 years and two Wechsler Preschool and Primary School Intelligence test subscales and School Readiness test for 60-month olds. Salt will be collected at various places of purchase and in households to test for iodine content (first qualitatively, then quantitatively if iodized) at baseline and at the end of the study. All data collection will be done at home or a convenient site in the village.

Conversion of diet into iodine content will be done based on food composition table. Anthropometric data will be converted to standardized scores using the new World Health Organization reference curves, using World Health Organization's software Anthro v. 2 for children up to 60 months of age.

Analyses will compare the nutritional and developmental outcomes of two groups who differ in iodized salt exposure, covarying confounders such as mother's education, family assets, child's sex, and adjusting for clustering. Developmental scores will be derived for the Bayley subtests, and the Wechsler Preschool and Primary School Intelligence test. These tests typically assign 1 point for every correct answer and 0 points for incorrect answers; the summed continuous scores are then used for analysis. Potential confounds include the Home Observation for Measurement of the Environment Inventory and Center for Epidemiologic Studies Depression Scale symptom scores, which also are calculated as the sum of individual item points. Laboratory analyses will be conducted to determine hemoglobin, ferritin, soluble transferrin receptor, thyroglobulin, thyroxin, and triiodothyronine from blood samples, and iodine excretion from urine samples. Analyses of salt will also indicate whether salt consumed by participants in the intervention districts has the required ppm of iodine and more than salt consumed in control districts. If households in the control sites have higher than expected levels of iodine, this will be recorded and included in the analyses.