New Statistical Approach to Apportion Dietary Sources of Iodine Intake: Findings from Kenya, Senegal and India

Abstract

Progress of national Universal Salt Iodization (USI) strategies is typically assessed by household coverage of adequately iodized salt and median urinary iodine concentration (UIC) in spot urine collections. However, household coverage does not inform on the iodized salt used in preparation of processed foods outside homes, nor does the total UIC reflect the portion of population iodine intake attributable to the USI strategy. This study used data from three population-representative surveys of women of reproductive age (WRA) in Kenya, Senegal and India to develop and illustrate a new approach to apportion the population UIC levels by the principal dietary sources of iodine intake, namely native iodine, iodine in processed food salt and iodine in household salt. The technique requires measurement of urinary sodium concentrations (UNaC) in the same spot urine samples collected for iodine status assessment. Taking into account the different complex survey designs of each survey, generalized linear regression (GLR) analyses were performed in which the UIC data of WRA was set as the outcome variable that depends on their UNaC and household salt iodine (SI) data as explanatory variables. Estimates of the UIC portions that correspond to iodine intake sources were calculated with use of the intercept and regression coefficients for the UNaC and SI variables in each country's regression equation. GLR coefficients for UNaC and SI were significant in all country-specific models. Rural location did not show a significant association in any country when controlled for other explanatory variables. The estimated UIC portion from native dietary iodine intake in each country fell below the minimum threshold for iodine sufficiency. The UIC portion arising from processed food salt in Kenya was substantially higher than in Senegal and India, while the UIC portions from household salt use varied in accordance with the mean level of household SI content in the country surveys. The UIC portions and all-salt-derived iodine intakes found in this study were illustrative of existing differences in national USI legislative frameworks and national salt supply situations between countries. The approach of apportioning the population UIC from spot urine collections may be useful for future monitoring of change in iodine nutrition from reduced salt use in processed foods and in households.

Keywords: household salt; native dietary iodine; population iodine intake apportionment; processed foods; universal salt iodization.

Figure 1

Estimation of UIC portions corresponding with dietary sources of iodine intake of WRA. Marker symbols: Kenya triangles, Senegal squares, India circles. Open symbols indicate the location of average UNaC findings in each country. Their corresponding UIC levels are shown by the horizontal dotted lines. Closed symbols show the total UIC findings and UIC portion estimates attributable to iodine intakes from household salt are illustrated by square brackets.

Figure 2

UIC ¹ portions (geometric mean, 95% CI) by dietary source of iodine intake of WRA in Kenya, Senegal and India. ¹ UIC = Urinary iodine concentration, WRA = Women of reproductive age.