Iodine is a required component of thyroid hormones and is necessary for growth and development. Because of iodine’s critical role in fetal and early childhood neurocognitive development, pregnant and lactating women and children less than 2 years of age are the primary groups targeted by efforts to ensure iodine sufficiency (1). Since the introduction of voluntary salt iodization programs in the 1920s, the overall iodine status of the U.S. population has generally been considered sufficient or even in excess of requirements (2). Urine iodine concentration (UIC) from spot urine samples is the most common indicator used for assessing the iodine status of populations (1). To monitor the iodine status of the U.S. population, the National Health and Nutrition Examination Survey (NHANES) measures UIC among a representative sample of residents aged >6 y. From NHANES I (1971–1974) to the present the iodine levels in NHANES have decreased by approximately 50% (2, 3). Although the median iodine level in the U.S. population is still considered sufficient despite this drop, some data suggest that iodine intake among pregnant women may be insufficient, whereas iodine intake among school-age children (6–12 y) may be above requirements (3, 4). The Recommended Dietary Allowance (RDA) of iodine is 90 μg/d for children 1–8 y and 120 μg/d for children 9–13 y (5).

Sources of iodine in the U.S. include dairy products, due to the use of iodine-containing cleaning products used in the milking process and iodine added to animal feed; grains and breads, as iodine can occur naturally in crops grown in iodine-rich soils or be added through the use of iodate-dough conditioners; table salt, approximately 70% of which is estimated to be iodized by voluntary iodization programs; marine fish, other seafood, and some seaweeds; and some dietary supplements (6–8). However, estimating people’s iodine intake is complicated by substantial variation in the iodine content of foods. Because of this variation, the U.S. Department of Agriculture food composition tables, which are frequently used to estimate nutrient intake on the basis of reported food intake, do not contain data on the iodine content of U.S. foods (9). The only national data that can be used to estimate U.S. iodine intake are from the U.S. Food and Drug Administration’s Total Diet Study (TDS), which measures iodine in samples of more than 250 foods from three locations in each of four regions. TDS data for 2003–2004 showed that dairy products were the single largest contributor to total iodine intake in all age-sex groups examined, other than infants, accounting for 70% of iodine intake among children aged 6 y and 10 y, and that grains accounted for approximately 15% of iodine intake among these children. However, TDS data do not reflect iodine intake from table salt or dietary supplements (10).

Previously we described dairy products as an important contributor to iodine status among pregnant and reproductive age women in the U.S. (4). Public health interventions that aim to ensure iodine sufficiency among pregnant and reproductive age women must also be aware of the potential for excess intake of iodine among children. In this study, we sought to describe contributors to iodine intake among U.S. children aged 6–12 y, the group whose iodine status the World Health Organization (WHO) has recommended for monitoring (1).

Slightly more than half of children were non-Hispanic white, 22.0% Hispanic, and 14.1% non-Hispanic black (Table 1). Most children had consumed some dairy products in the previous 24 h (87.6%), while nearly all had consumed some grain products (99.2%). Among those who had consumed dairy products, median dairy intake was 331 g, and among those who had consumed grain products, median grain intake was 270 g. Use of a supplement in the previous 24 h that was identified as containing iodine was reported by 16.5% of the sample, and 28.3% reported “occasionally/very often” adding salt to their food at the table.

The estimated median UIC among all children 6–12 y was 211 μg/L (Table 2). Results of our bivariate analyses showed that the median UIC varied significantly by dairy intake, supplement use, race/ethnicity, and age, but not by grain intake, table salt use, or sex. For nearly all subgroups analyzed, median UIC was adequate or above requirements; no subgroups had median UIC values categorized as insufficient, and only the group taking a supplement containing iodine in the previous 24 h had median UIC just above the cut-off for excess intake (median UIC=303 μg/L). The proportion of children with UIC below 50 μg/L was 5.4%. This proportion was higher in females, those not taking a supplement containing iodine, and children 9–12 y. Among all subgroups categorized this proportion was never higher than 10%.

Median UIC increased by quartile of dairy product intake, from 157 μg/L in the lowest quartile to 278 μg/L in the highest quartile (p<0.001; Figure 1a). However, median UIC did not vary by quartile of grain intake (p=0.2; Figure 1b).

In linear regression analyses adjusting for all primary predictors of interest and covariates (Table 3), dairy product intake and supplement use were positively associated with UIC, while grain product intake and being non-Hispanic black were inversely associated with UIC. Based on this model, for every 100 g/d increase in dairy product intake, there is an increase in UIC of 5%, controlling for all other variables; on the other hand, for every 100 g/d increase in grain intake, UIC decreased by 3%. Taking a supplement containing iodine in the previous 24 h increased UIC by 23%, controlling for all other variables.

Our results indicate it is unlikely there is iodine deficiency among children 6–12 y in the U.S. Among all of the sub-groups examined, median UIC was consistently categorized as adequate or above adequate, and the proportion of children with UIC <50 μg/L was always less than 10%. As we found dairy products and dietary supplements to be important sources of iodine among children, we additionally explored the iodine status among children who had consumed neither in the previous 24 h: this included 10.5% of children in our sample, who had adequate iodine status (median UIC=150 μg/L). This is reassuring, for while data are limited, some evidence suggests that even mild iodine deficiency may impair a child’s cognitive development. A recent iodine supplementation trial in New Zealand showed that mildly iodine-deficient children improved their scores on picture concepts and matrix reasoning tests after they began taking iodine supplements (15).

We found several subgroups of children whose median UIC was categorized as above requirements. This meaning of this term is difficult to interpret from the public health perspective. This range of median UIC 200–299 μg/L is generally considered to increase risk of iodine-induced hyperthyroidism among populations with long-standing iodine deficiency who then experience a rapid increase in iodine intake (1). Given that the U.S. population has had an adequate iodine status since the 1920s, this situation is not applicable in the U.S. Otherwise, median UIC values up to 300 μg/L are generally considered safe (1, 16). Results of a study of children aged 6–12 from five continents showed that although median UIC values >500 μg/L were associated with increased thyroid volume, levels up to 500 μg/L were generally well tolerated by healthy children (17).

The only subgroup of children with a median UIC in the excessive range were those who had consumed a supplement containing iodine during the previous day, and median UIC in this group was just above the cut-off for excess (median UIC 303 μg/L; excess is >300 μg/L). We found that 16.5% of children had consumed a supplement containing iodine in the previous 24 h. We were surprised at this relatively high proportion, which likely would have been even higher with a longer recall period. An analysis of NHANES data for 2003–2006 showed that 43% of children 4–8 y and 29% of children 9–13 y had taken some type of supplement in the previous 30 d and that 32% and 20%, respectively, had taken a multivitamin-multimineral supplement (18). This analysis did not report the type of supplements that children were taking (e.g. children’s vitamin, general multivitamin). Among children who had consumed a supplement containing iodine in the previous day in our analysis, median intake was 66 μg, representing 73% of the RDA for children 6–8 y and 55% of the RDA for children 9–12 y. Given current dietary intake patterns of children, particularly dairy consumption, and the use of iodine-containing cleaning agents in the dairy production process, few children in the U.S. are likely to need a supplement containing iodine.

In bivariate analyses grain product intake was not associated with UIC, while in the adjusted model grain product intake was inversely associated with UIC. It is unclear why we found this inverse association. One possible explanation could be that high grain product consumption displaced other better sources of iodine. Additionally, the amount of iodine in grain products can vary substantially (8), and grain product intake from 24-hr dietary recalls may not be a good measure of iodine intake. Regular use of table salt was not associated with UIC. Information on salt use did not necessarily indicate recent intake, and we were unable to determine whether the table salt used was iodized, which may explain why we did not find an association between salt use and iodine status. Iodization of salt in the U.S. is voluntary, and while it has been estimated that approximately 70% of table salt in the U.S. is iodized, it is not clear how valid this estimate is. Additionally, the majority of the salt consumed in the U.S. is from processed and restaurant foods, not from table salt, and this salt is typically not iodized (19–21).

This analysis has several limitations. UIC values and single 24-hour dietary recalls can both be highly variable, and do not capture usual intake. However, since UIC is reflective of recent iodine intake, it was important to use only the 24-hour recall conducted at the time urine samples were collected in order to assess the association between dairy intake and iodine status. Since iodine is not included in the USDA nutrient database we are unable to quantify iodine consumption, so instead quantified consumption of food groups known to be important contributors to iodine intake. However, the iodine content of foods included in these groups can be variable (8), and it is unknown how well these aggregate food groups capture iodine intake. UIC is generally not used as an indicator of individual iodine status, but here we used it as such in regression models in order to describe dietary factors associated with iodine intake. This was done in order to be able to adjust for other potential contributors to iodine status and demographic characteristics. Findings from the regression models were consistent with findings from bivariate analyses which used UIC as a population level indicator. The strengths of the analysis include that we analyzed recent, nationally representative data of U.S. children, and we were able to describe dietary factors that are associated with iodine status.

We found that the consumption of dairy products and the use of dietary supplements were each positively associated with UIC among U.S. children. Iodine containing supplements are likely not needed by most schoolchildren in the U.S. because dietary iodine intake is adequate in this age group; whether iodine should be included in supplements that are commonly consumed by children should be evaluated. Currently there is no monitoring or regulation of the use of iodine sanitizing agents in the dairy industry, and changes in the dairy production process could lead to changes in iodine levels of dairy products, and consequentially in the iodine status of the population (22). Continued monitoring of iodine status and better understanding of the iodine content in dairy products is warranted.