Purpose and Previous U.S. Preventive Services Task Force Recommendation

This report was commissioned by the U.S. Preventive Services Task Force (USPSTF) in order to update its 2006 recommendations on screening and supplementation for iron deficiency anemia in young children.¹

The USPSTF previously concluded that the evidence was insufficient to recommend for or against routine screening for iron deficiency anemia in asymptomatic children ages 6 to 12 months (I statement) and that the evidence was insufficient to recommend for or against routine iron supplementation for asymptomatic children ages 6 to 12 months who are at average risk for iron deficiency anemia (I statement). These recommendations were based on a lack of evidence that screening results in improved health outcomes, and poor and conflicting evidence on the benefit of iron supplementation in children who are not at increased risk for iron deficiency anemia.

The USPSTF also previously recommended routine iron supplementation for asymptomatic children ages 6 to 12 months who are at increased risk for iron deficiency anemia (B recommendation). This recommendation was based on evidence that iron supplementation may improve neurodevelopmental outcomes in children who are at increased risk for iron deficiency anemia, a possible benefit that outweighed any potential harms.

Condition Definition

Iron is required in the production of hemoglobin, an essential protein found in red blood cells. Over time, iron is stored in the body for use in hemoglobin production. Iron deficiency occurs when the level of stored iron becomes depleted. Iron deficiency anemia occurs when iron levels are sufficiently depleted to produce anemia, characterized by hypochromic and microcytic red blood cells.²

For infants and children, iron deficiency anemia is defined as iron deficiency with hemoglobin levels less than 110 g/L.^3,4 When used alone, hemoglobin is not the most accurate test in this age group.⁵ Hemoglobin is a sensitive test for iron deficiency anemia, but it has low specificity because iron deficiency anemia accounts for less than half of all cases of anemia in toddlers in the United States.⁶ Serum ferritin—ideally assessed in the absence of infection or inflammatory disease and in combination with another hematological measure, such as C-reactive protein, transferritin saturation, or reticulocyte hemoglobin—is a useful laboratory measure of iron status, with a low value being diagnostic of iron deficiency.^6-8 Infants in the United States with iron deficiency are usually asymptomatic.

Prevalence

Iron deficiency among infants and toddlers in the United States has a prevalence of about 8 percent in the general population,^9-11 although estimates based on a model of body iron stores (as opposed to using the more conventional ferritin model) are much higher at 14 percent for children ages 1 to 2 years.⁸ Only about one third of children who are iron deficient have associated anemia.^6,9,12 The prevalence of iron deficiency anemia in children ages 1 to 5 years is estimated to be about 1 to 2 percent in the United States.^6,13 Prevalence in children from low-income families is estimated to be slightly higher, around 3 percent for boys and 4 percent for girls, based on one study of 432 children ages 1 to 3 years residing in California.¹¹ Current evidence on the prevalence of iron deficiency anemia in infants younger than age 1 year in the United States is lacking, although estimates for low-risk infants in other developed countries range from 2 to 4 percent.^14,15

Prevalence of iron deficiency anemia in the United States varies according to race/ethnicity. Based on National Health and Nutrition Examination Survey (NHANES) data from 1999–2002, rates of iron deficiency anemia among 1- to 3-year-olds were lowest in Mexican Americans (0.9%) and highest in non-Hispanic whites (2%), with rates for non-Hispanic blacks falling between the two (1.6%).⁶ These data stand in contrast to older NHANES data (1988–1994), which found reversed prevalence trends in the same age group (Mexican Americans, 5.5%; blacks, 3.5%; and whites, 1.2%), although the overall prevalence of iron deficiency anemia (2.6%) was similar to the later data.¹⁶ The reason for these differences is unclear. Definitions of iron deficiency anemia were the same at both timepoints; however, estimates of prevalence of iron deficiency anemia based on NHANES data have been noted to be statistically unreliable.^6,17

Risk factors associated with iron deficiency anemia are discussed below in Contextual Question 1.

Burden of Disease/Illness

Iron deficiency anemia can present a significant burden of disease in infancy and childhood. A narrative review of 15 studies found that infants with iron deficiency anemia had cognitive test scores 6 to 15 points lower and motor test scores 9 to 15 points lower than iron-sufficient infants.¹⁸ In addition, behavioral characteristics, such as diminished activity and social interaction levels, have consistently been observed in infants with iron deficiency anemia.^11,18

The effect of iron deficiency anemia in infancy and childhood has been reported in few well-designed long-term controlled studies. As described in the 2006 USPSTF report, numerous cohort and case-control studies have reported a correlation between presence of iron deficiency with or without anemia and impaired neurodevelopment in older children.¹⁷ However, many of these studies had methodological flaws; for example, the outcomes examined varied and were not clearly clinically important, and the effect of iron deficiency anemia in infancy and childhood has been reported in few well-designed long-term controlled studies. A more recent nonsystematic review of primarily observational studies determined that presence of iron deficiency anemia was consistently associated with cognitive and behavioral delays in children.¹⁹

This association was more pronounced when iron deficiency anemia was present in children younger than age 2 years than in those older than age 2 years, suggesting that the negative long-term effects of iron deficiency anemia may be more avoidable when it is identified and treated at a young age. However, the same review found that the evidence was insufficient to prove causation in these studies, primarily because of the difficulty in assessing the effect of potential confounders among the studies, including nutritional disparities between groups, socioeconomic factors, and clinical effects of anemia that are not directly related to cognitive function.

One of the longest and most frequently cited studies on the effect of iron deficiency anemia in infancy and long-term developmental outcomes followed 114 healthy and relatively well-nourished Costa Rican children from ages 12 to 23 months to age 19 years.^17,20-23 All children with iron deficiency anemia in this study received treatment in infancy, resulting in resolution of anemia within 3 months. Despite this, as noted in the prior USPSTF report, at ages 11 to 14 years the children who were treated for iron deficiency in infancy scored lower on intelligence tests (Wechsler Intelligence Scale for Children–Revised Verbal Intelligence Quotient, 101.8 ± 2.0 vs. 104.6 ± 1.3) and on a variety of cognitive function tests than children who were iron sufficient in infancy.^17,20 Previously iron-deficient children were also more likely to have repeated a grade in school (26% vs. 12%; p=0.04) and to have been referred for special education services (tutoring, 21% vs. 7%; p=0.02). Parents and teachers of children in the iron-deficient group were also more likely to report concerns regarding behavior compared with iron-sufficient children. Intelligence and cognitive test scores were adjusted for a number of potential confounders, including age, sex, and mother's Intelligence Quotient score. Longer followup of the same cohort, published since the 2006 USPSTF report, found that measures of cognitive function—using subscales of the Wechsler Intelligence Scale for Adults and measures of mathematic ability—remained lower in previously iron-deficient infants at age 19 years compared with previously iron-sufficient infants (mean score, 98.2 vs. 107.6 points; mean difference, 9.4 points [95% confidence interval (CI), 7.0 to 11.0 points]).²³ This effect was magnified in children from families with low socioeconomic status (mean score, 70.4 vs. 95.3 points; mean difference, 24.9 points [95% CI, 20.6 to 29.4 points]). Pattern recognition and other tests of neurocognitive function were also lower in the previously iron-deficient group at age 19 years.²⁴

Presence of iron deficiency anemia has also been associated with ischemic stroke in otherwise healthy children in case reports.²⁵ A case-control study conducted in Canada found that among 15 cases of stroke in otherwise healthy children ages 12 to 38 months, eight had iron deficiency anemia (53%) compared with 13 of 143 (9%) matched healthy controls.²⁶ After controlling for platelet count, the adjusted odds ratio (OR) was 10 (95% CI, 3 to 33), suggesting that children with iron deficiency anemia may be at increased risk for stroke compared with similarly aged children without iron deficiency anemia.

Rationale for Screening/Screening Strategies

Screening young children for iron deficiency anemia may lead to earlier identification and therefore earlier treatment, which has the potential to prevent serious negative health outcomes. Strategies for screening can include either routine screening or targeted screening based on established risk factors or risk-assessment instruments and diagnostic tests. Routine screening for all children could occur when they present for pediatric health visits.

Current Clinical Practice

Recommendations of Other Groups

Contextual Question 1. What Risk Factors Are Associated With Iron Deficiency Anemia, and How Well Do Risk-Assessment Tools Identify Children at Increased Risk for Iron Deficiency Anemia?

The CDC has identified children from families with low socioeconomic status (including migrant workers and recent immigrants to the United States) as having a high risk for developing iron deficiency anemia.⁴ Prematurity or low birthweight also increases risk for developing iron deficiency anemia, as does use of nonfortified formula in the first year of life (unlikely to occur in the United States), exclusive breastfeeding with no or erratic iron supplementation after age 6 months, and the introduction of cow's milk before the age of 1 year.¹⁷ As with the prior report, we found no studies reporting quantified data on these risk factors and their association with iron deficiency anemia.¹⁷ One small study of 68 children younger than age 3 years found that those with food insecurity were more likely than food-secure children to have iron deficiency anemia (19% [12/65] vs. 10% [56/561]; adjusted OR, 2.4 [95% CI, 1.1 to 5.2]).⁴⁰

The relationship between specific risk factors and presence of iron deficiency can be estimated based on several analyses of NHANES data on children ages 1 to 3 years^10,16,41 and one study⁴² of similarly aged, low-income, primarily Hispanic children (Table 1). Based on these data, weight for height at or greater than the 95th percentile was consistently associated with presence of iron deficiency, although the risk estimate was not significant in one study. Bottle-feeding beyond the first year of life, having a mother who is currently pregnant, and residence in an urban area were also all significantly associated with increased iron deficiency risk based on one study. Hispanic ethnicity was a significant predictor of iron deficiency in two studies, but two others found no significant association. Results from two studies were also mixed on the association between risk for iron deficiency and male sex. Birthweight less than 2,500 g and family income below the Federal poverty level were associated with a nonsignificant trend toward presence of iron deficiency in these studies.

Table 1

Association Between Demographic Characteristics and Iron Deficiency.

Risk for anemia was reported separately in one study.⁴² Having a mother who is pregnant was a significant predictor of anemia in this population (OR, 3.5 [95% CI, 1.4 to 8.9]). Hispanic children had moderately increased risk for anemia compared with non-Hispanic children (OR, 1.3), although the result was not statistically significant (95% CI, 0.3 to 5.9). Children who participated in the Women, Infants, and Children Special Supplemental Nutrition Program (WIC) were less likely to be anemic than children who did not participate (adjusted OR, 0.34 [95% CI, 0.1 to 0.9]). Children with a higher rate of weight gain were also marginally less likely to be anemic than slower growing children (adjusted OR, 0.995 [95% CI, 0.990 to 0.999]). There was no significant association between either age or sex and increased risk for anemia.

We identified no studies that assessed how well risk-assessment tools identify children at increased risk for iron deficiency anemia.