TITLE: AUTHORS: ABSTRACT: For children with elevated lead levels, guidelines from the Centers for Disease Control and Prevention recommend evaluating iron status and encouraging iron intake. It is unclear whether there is a physiological association between lead exposure or absorption and iron depletion, or if some other common factor, such as poverty, contributes to both. Identifying and treating iron deficiency anemia in lead-exposed toddlers is crucial because both conditions are believed to contribute to abnormal cognition and development. This study examines whether moderately elevated lead levels are associated with measures of iron status, adjusting for other relevant social variables. The study was conducted at care sites in four urban centers between August 1994 and October 1996. Eligible children were between 11 and 33 months old and had lead levels of 20 to 44 µg/dL confirmed by the CDC laboratory. At their first visit to a TLC clinic, children underwent venipuncture for a complete blood count and determination of ferritin and blood lead levels. Although TLC excluded children with hemoglobin concentrations of less than 9 g/dL from further participation, we included those children in this analysis. Due to over 70% of the subjects in the TLC being African American and the fact that hemoglobin concentration is lower in black children, we limited the analysis to them. METHODS We compared iron status between 2 groups of children. The children with moderate lead exposure came from among those recruited for a clinical trial, the Treatment of Lead-Exposed Children Trial, which used oral chelation to attempt to prevent lead-associated developmental delay. Recruitment for TLC is described elsewhere. We compared the iron status between children 11 to 33 months old with confirmed blood lead levels of 20 to 44 µg/dL and demographically similar children with blood lead levels of <10 µg/dL. There were no differences. Laboratory investigation or empirical treatment for iron deficiency is not justified on the basis of moderately elevated blood lead levels alone. (J Pediatr 1999;135:108-10) can children aged 11 to 33 months who resided in an urban area with lead levels <10 µg/dL. Iron depletion was defined as having a serum ferritin level of less than 10 µg/L or a ferritin level of less than 10 µg/L along with a red cell distribution width (RDW) greater than 14.5%. Iron deficiency was characterized by a ferritin level of less than 10 µg/L and a mean corpuscular volume (MCV) of less than 70 fL. Iron deficiency anemia was defined as having a hemoglobin level of less than 11 g/dL and a ferritin level of less than 10 µg/L, or as having a hemoglobin level of less than 11 g/dL, a ferritin level of less than 10 µg/L, and an RDW greater than 14.5%, or as having a hemoglobin level of less than 11 g/dL, a ferritin level of less than 10 µg/L, and an MCV of less than 70 fL. It is important to note that although ferritin is an acute-phase reactant, there is no reason for variations in ferritin levels not caused by iron status to differ among these groups of children. Statistical Analysis SAS (SAS Institute, Cary, NC) and S-PLUS (Statistical Sciences, Inc, Seattle, Wash) were used to calculate means, frequencies, and Pearson’s correlation coefficient. For continuous variables, a Student t test was used, while for dichotomous variables, a χ2 test was used. To explore the association between FER and blood lead levels, a nonparametric local regression smoothing method was employed. Statistical analysis was conducted using SAS from SAS Institute in Cary, NC, and S-PLUS from Statistical Sciences, Inc. in Seattle, Washington. These tools were employed to calculate means, frequencies, and Pearson’s correlation coefficient. For continuous variables, the Student t test was applied, while the χ2 test was used for dichotomous variables. The relationship between FER and blood lead levels was examined using a nonparametric local regression smoothing method. RESULTS In the data from the TLC, of the 1201 children with eligible blood lead levels of 20 to 44 µg/dL, 1194 were aged 11 to 33 months. Of those, the 787 who were African American constitute the children with moderate lead exposure for this analysis. In the NHANES III data there were 2596 children between 11 to 33 months of age. Of these, 1511 had lead levels <10 µg/dL. Of those, 789 were living in an urban area, defined as an area where 1 million or more persons reside; and of those, the 222 children who were African American constitute the children with low lead levels. Age, sex, and ferritin levels were not significantly different between the moderate and low lead exposure groups. There were no significant differences between the moderate and low lead exposure groups (Table I). The prevalence of iron depletion, iron deficiency, and iron deficiency anemia was similar in both groups (Table II). The correlation coefficient between blood lead and ferritin (FER) levels for children with low lead exposure was –0.04, with a confidence interval of –0.173 to 0.96 and a p-value of 0.57. For children with moderate lead exposure, the correlation coefficient was also –0.04, with a confidence interval of –0.112 to 0.030 and a p-value of 0.26. No significant differences were found when the groups were divided by lead levels (not shown). Our data demonstrate no difference in the prevalence rates of iron depletion, iron deficiency, and iron deficiency anemia in children with moderate lead exposure compared with those with low exposure. Considering that fewer parents in the TLC group had completed high school, fewer were employed, and more received medical assistance, one might expect higher levels of iron deficiency based on these socioeconomic factors alone. However, this is not what the data demonstrated. The study had a power of 80% to detect a 10.3% difference in iron deficiency or iron deficiency anemia between the two groups. Venipuncture is painful, laboratory work is expensive, and iron may be toxic and is a common pediatric ingestion. Thus, laboratory workup or empirical treatment with iron should be reserved for children whose risk derives from factors other than a moderately increased blood lead level.