QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kordas, K. Articles by Stoltzfus, R. J. Search for Related Content PubMed PubMed Citation Articles by Kordas, K. Articles by Stoltzfus, R. J.

---

Community and International Nutrition

Blood Lead, Anemia, and Short Stature Are Independently Associated with Cognitive Performance in Mexican School Children¹

Katarzyna Kordas², Patricia Lopez^*, Jorge L. Rosado, Gonzalo García Vargas^**, Javier Alatorre Rico, Dolores Ronquillo, Mariano E. Cebrián and Rebecca J. Stoltzfus

Division of Nutritional Sciences, Cornell University, Ithaca, NY 14853; ^* Department of Nutritional Physiology, National Institute of Medical Sciences and Nutrition, Mexico City, Mexico; School of Natural Sciences, University of Querétaro, Querétaro, Mexico; ^** School of Medicine, University of Juarez at Durango, Gómez Palacio, Mexico; Department of Psychology, National Autonomous University of Mexico, Mexico City, Mexico; and Department of Toxicology, National Polytechnic Institute-CINVESTAV, Mexico City, Mexico

²To whom correspondence should be addressed. E-mail: kk326{at}cornell.edu.

	ABSTRACT

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Lead exposure and nutritional factors are both associated with cognitive performance. Lead toxicity and nutritional status are also associated with each other. We examined whether nutritional status variables account for part or all of the association between cognitive performance and lead exposure. First-grade children (n = 724) ages 6–8 y, attending Mexican public schools located in the vicinity of a metal foundry were asked to participate and 602 enrolled in the study. Blood lead, iron status, anemia, anthropometry, and cognitive function were assessed. Results from 7 standardized tests are presented here. The mean blood lead concentration was 11.5 ± 6.1 µg/dL (0.56 ± 0.30 µmol/L) and 50% of the children had concentrations >10 µg/dL (0.48 µmol/L). The prevalence of mild anemia (<124 g/L) was low (10%) and stunting (<2 SD) was nonexistent (2.3%). In bivariate analyses, lead was negatively associated with 4 cognitive tests and was also inversely correlated with iron status, height-for-age Z scores, and head circumference. In multivariate models, the association between lead and cognitive performance was not strongly affected by nutritional variables, suggesting that the relation of lead to cognition is not explained by lead’s relation to iron deficiency anemia or growth retardation. In multivariate models, hemoglobin concentration was also positively associated with Peabody Picture Vocabulary Test and Number Sequencing performance, whereas serum ferritin was negatively related to the Coding subscale of the Wechsler Intelligence Scales for Children-Revised Mexican Version (WISC-RM).

KEY WORDS: • iron • anemia • growth • lead exposure • cognitive performance

The link between lead status and cognitive function in children has been clearly demonstrated over the last 20 y. Lead exposure is associated with several aspects of cognitive development and status. An impressive amount of data has been generated from cross-sectional and prospective studies showing a negative relation between childhood IQ and blood (1–5) and dentine lead (6,7). Various studies have also attempted to tease out the effects of lead on school achievement and behavior (2,6,8,9), as well as more functionally specific cognitive domains, such as memory, attention, language, or motor development (10–12).

Lead exposure is likely related to nutritional status in children. For example, lead and micronutrient deficiencies, especially iron deficiency, frequently occur together (13–16). Furthermore, iron deficiency and anemia are related to cognitive performance, especially attention, in preschool and school children. Pollitt et al. (17) found significant differences in IQ between children with iron deficient anemia (IDA)³ and iron-replete children. They also found that vocabulary scores differed significantly among IDA, iron deficient, and iron-replete groups. In a study of iron-deficient adolescent girls, Bruner et al. (18) found improvement in vocabulary performance after 8 wk of iron supplementation, although there were no changes in attention measures. Finally, Seshadri and Gopaldas (19) showed improvements in IQ scores of anemic Indian children 7–8 y old and boys 5–6 y old, compared with the placebo group, after 60 d of iron supplementation. Both iron (20,21) and lead (22–24) have also been associated with short stature, which is independently related to child development and cognition (25–27).

Despite the known links between nutritional variables and cognition, the context of children’s nutritional status has usually not been considered when evaluating the association between cognitive deficits and lead exposure. In studies of cognitive function and lead, other well-known confounders are taken into account, i.e., socioeconomic status (SES), maternal IQ, and child or family characteristics. One longitudinal study did assess both hemoglobin concentrations and lead exposure (28), but only the individual effects of lead and iron on cognition were reported. In a later report from this cohort (29), hemoglobin did not change the estimate of lead parameters, suggesting an independent effect of lead. No other nutritional factors were examined. Thus, it is unclear from the current literature to what extent lead exposure is independently related to cognitive deficits in children, as opposed to being indirectly related via its causal or noncausal association with growth or micronutrient deficiencies.

Iron deficiency and lead toxicity are thought to exert similar effects on brain neurochemistry, i.e., both affect the process of myelination and neurotransmitter metabolism and function (30–35). Furthermore, the functional manifestations of these two conditions overlap to some extent (36,37). Therefore, it is possible that when iron deficiency and lead toxicity occur together, the resultant cognitive deficits stem from their synergistic action. In contrast, iron deficiency and lead toxicity may also affect distinct parts of the brain, in which case the resultant outcomes would be independent of each other.

We propose an analytical model that considers these relations simultaneously, in which lead exposure may be influencing cognition directly or through its relation to iron status and growth retardation, two factors that may also affect cognition independently. In this analysis, we explore the individual and combined links between lead, iron status, stature, and cognitive outcomes in Mexican first-graders with low-level lead exposure to answer the following questions: 1) Is lead exposure related to measures of cognitive function in school children? 2) Is iron deficiency, anemia, or growth related to cognitive function in school children? 3) How much of the lead-cognition relation is explained by iron status, anemia, and growth indicators in this population?

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
    Participants. The study took place in Torreón, a highly industrialized city in the north of Mexico, where the source of lead exposure is localized to a metal foundry. Torreón is located in a semiarid area at an elevation of 1060 m above sea level and has an average monthly rainfall of 17.4 mm. These dry conditions contribute to lead exposure among children because lead particles settle in the dust and can be picked up on hands or inhaled. Children eligible for the study attended 9 public elementary schools located within a 3.5-km radius of the foundry, considered the area of greatest effect of lead contamination. All 724 children from these schools regularly attending first grade in early January 2001 were asked to participate. Investigators and school officials organized informational meetings with parents at which the study, its risks, and benefits were explained and consent forms were distributed. Overall, 602 children were enrolled in the study. The sample size was designed to detect clinically significant differences in blood lead in a randomized supplementation trial, whose results will be described elsewhere. Baseline data collection consisted of venous blood sample, anthropometry, and cognitive assessment and was conducted between January and April of 2001. During cognitive assessment, both standardized and nonstandardized measures were used. This study considers only the standardized tests or their adaptations. The study was approved by human subject committees at Johns Hopkins University Bloomberg School of Public Health and the Institute of Medical Sciences and Nutrition in Mexico.

    Laboratory methods. Venous blood samples from fasting subjects were obtained. Whole blood was collected in 5-mL sodium heparin vacutainer tubes (Becton Dickinson). Hemoglobin (Hb) was analyzed at the schools, using a Hemocue Photometer (Hemocue AB). Zinc protoporphyrin (ZPP) is a measure of both iron status and lead exposure and it has been used previously in studies of lead exposure. It was tested in whole blood at the school with ZP Hematofluorometer (AVIV Biomedical). Blood samples were transported to the laboratory on ice and processed the same day. Whole blood and serum were divided into aliquots and stored at -80°C until analysis. Serum ferritin (SF) was analyzed using an immunoradiographic method (Coat-A-Count Ferritin IRMA). C-reactive protein (CRP) was analyzed using the LINCON coagulation kit (Laboratorios Lincon). This qualitative assay indicates the presence or absence of infection, but not CRP concentrations. Blood lead (PbB) analysis was performed at the National Polytechnic Institute in Mexico, using atomic absorption spectrophotometry (Zeeman 5100, Perkin Elmer) (38). Samples were analyzed in duplicate and those with a CV > 5% were reanalyzed. Lead in bovine blood (SRM 955b, NIST) was used as a standard reference. This laboratory participates in two quality control programs, the Trace Elements External Quality Assessment Scheme at University of Surrey, UK and the Inter Laboratory Program of Quality Control at Zaragoza, Spain.

    Anthropometry. Measurements were conducted according to standard methods (39). A single person performed all weight and height measurements. Children took off their shoes but wore school uniforms for both weight and height assessments. Weight was measured using an upright scale (Torino, Model Express Plus) with the capacity to weigh 160 kg in 100-g increments. Height was taken using a measuring board (Martín, University of Guadalajara). Finally, head circumference (HC) was measured using a flexible tape (Rotary R-280).

    Cognitive outcomes. The testing battery comprised 14 paper-and-pencil or computer-based tests covering various aspects of cognitive functioning. The tests were divided into two groups and administered on two different days. Group A consisted of Cognitive Abilities Test (CAT), the Wechsler Intelligence Scale for Children-Revised Mexican Version (WISC-RM) Coding, Digit Retention and Arithmetic; Number and Letter Sequencing, Prueba de Habilidades Cognitivas (a computer-based Test of Cognitive Abilities with 5 subtests), which were always administered in this order on d 1 of testing. On d 2, Group B, consisting of a curriculum-based Math Achievement Test (MAT), a Test of Visual-Spatial Abilities and the Peabody Picture Vocabulary Test (PPVT-Spanish Edition), was administered in the above order. The WISC-RM and the PPVT were both validated with Mexican-American populations; the rest of the tasks were piloted among 1st and 2nd graders in a public elementary school in Mexico City before the project began. The entire WISC-RM battery was not administered due to time constraints and the number of other tasks included in the testing battery. The Coding, Arithmetic and Digit Span tasks were used previously in lead studies among children whose performance could be compared with children in our sample.

Ten Mexican psychologists evaluated the children. They were divided into teams of 2; within the team, each psychologist administered one part of the battery. Each psychologist saw 3 or 4 children every day of testing. All testers underwent a 3-wk training by a child psychologist (J.A.R.) and a standardization session at an elementary school in a neighboring town. During this period, the psychologists worked with 1st and 2nd-graders and paid attention to the use of testing materials and giving clear instructions. Each tester was observed and evaluated until standardization was achieved.

Each day’s testing lasted for 65–70 min. An effort was made to administer the 2 groups of tests closely in time and there was usually no more than a week’s lapse between Group A and B tests for any child. Testing took place at each school in an isolated room, with as little noise and distraction as possible. The testers were unaware of children’s lead or micronutrient status.

The Peabody Picture Vocabulary Test-Revised, Spanish Version (40) was used to test receptive vocabulary skills and comprehension of spoken Spanish. The test consists of 111 pages, with 4 black-and-white sketches and one stimulus word per page. Children were asked to name or point to the picture that best described the given word. The raw score was converted into an age-standardized score based ona Mexican-American sample and used in statistical analyses. The test was designed with a mean and SD of 100 ± 15 points; the mean and SD of this sample was 103.2 ± 15.7.

    WISC-RM Arithmetic, Coding and Digit Span (41). WISC-RM Arithmetic is a subtest used to determine numerical reasoning skills of school children, and their ability to concentrate and integrate spoken instructions with arithmetic knowledge. Children were tested on counting, addition, subtraction, and multiplication. WISC-RM Coding tests learning of unfamiliar tasks in the form of matching two sets of symbols or numbers and symbols. The Digit Span subtest measures short-term auditory attention and memory. Raw scores for each subscale were converted into final scores based on age-specific Spanish-language norms and used in analyses. The mean ± SD of this sample was 7.4 ± 3.6 for Arithmetic, 9.1 ± 3.6 for Digit Span and 11.0 ± 2.8 for Coding. The highest score possible was 15 for Coding and 19 for Arithmetic and Digit Span.

The computer-based Cognitive Abilities Test, Stimulus Discrimination (CAT SD) Subtest (42) was used to measure stimulus encoding and retrieval. CAT SD is a task consisting of a probe and test stimuli. The stimuli are squares divided into a matrix of smaller fields, filled to create unique patterns. The stimuli are chosen randomly, both for the probe and for 5 "distracters," from among 24 possible matrices. During testing, children use a touch-screen to match one of 6 test stimuli to a given probe, after the probe disappears from the screen. After 4 practice trials, the test continues, without time limits, until the child completes 72 correct matches. Two variables were analyzed from this task, i.e., log₁₀-transformed SD of decision time and errors (specifically, whether children committed errors or not).

Number and Letter Sequencing is a task adapted from the Trail Making Test, Trails A (43); it is thought to measure psychomotor skills but has not been formally validated. In Number Sequencing, children were asked to connect as many of the 15 encircled numbers (4 through 18) randomly distributed across a page as possible in 60 s. In Letter Sequencing, children had to connect 15 randomly distributed letters (D-P, including the Ñ and Q). The first 3 numbers and letters in each subtest were used for task explanation and practice. The number of correct responses was used in analyses.

In a post-hoc analysis, the task was moderately correlated with WISC-RM Coding, which measures psychomotor speed (Numbers: Spearman = 0.24, P < 0.001; Letters: Spearman = 0.223, P < 0.001). Performance on this task might have been affected by children’s familiarity with numbers and letters. Most participants (93.6%) were able to count from 1 to 14 when administered WISC-RM Arithmetic, but performance on Number Sequencing was statistically lower for children who could not correctly answer the counting question (4.5 ± 3.0 vs. 6.0 ± 4.0 points, P = 0.014). Because the test was not standardized and correlates with other psychomotor and visual-spatial tests only moderately, the exact cognitive domain it is measuring is not clear. However, performance on WISC-RM Coding and a Visual-Spatial task, as well as teacher’s ratings of attention problems were significantly associated with Number Sequencing outcome (P < 0.001) and account for 21% of variability in scores, suggesting that the test does measure some aspects of visual-spatial or visual-motor abilities.

We do not have any measures of familiarity with letters, but children’s performance on Letter Sequencing was poorer with higher ratings of reading problems on the Conners Rating Scales for Teachers (44) (ANOVA, P < 0.001), a questionnaire in which teachers are asked to rate children on frequency of certain problem behaviors. It is also possible that suboptimal Letter Sequencing precedes poor reading ability. In multivariate regression analysis, performance on the Visual-Spatial task and especially the ability to pay attention (as rated by teachers) were both significantly associated with Letter Sequencing score (P < 0.001) and accounted for 13% of variability in scores. WISC-RM Coding was not associated with Letter Sequencing, suggesting that the motor aspect of the task was not an important determinant of performance.

    Demographic data, clinical history, and socioeconomic status. Parents were asked to attend an informational meeting at which they filled out a questionnaire concerning family health history, parental education and occupations, and family’s SES. The questionnaires were self-administered but the research staff assisted parents and clarified questions. Home visits were made for parents not in attendance at the meetings.

A family’s SES index was computed on the basis questions about possessions, housing conditions, crowding, and caregiver education. Points given for individual questions were summed into a scale ranging between 5 and 12 points. Three SES levels were developed: low (<33% of total range), medium (33–66% of total range), and high (above 66% of total range).

    Statistical methods. Height-for-age (HAZ) and height-for-weight Z-scores (WHZ) were computed using EpiInfo (CDC). Data were analyzed in STATA 6. Three outliers were removed from analyses: lead level of 47.9 µg/dL,⁴ ZPP level of 452 µmol/mol heme, and serum ferritin of 432.1 µg/L. For Letter Sequencing, the majority of children were not able to obtain any correct responses. Only 1 child of 594 tested completed the entire test correctly. Therefore, Letter responses were converted into a dichotomous variable (0 corresponded to 0 correct responses and 1 corresponded to any correct response) and modeled in logistic regression. Similarly, CAT SD errors were modeled in logistic regression to predict the likelihood of errors.

Means and SD were calculated for all cognitive tests according to the most commonly used lead categories: <10 µg/dL, 10–14.9 µg/dL, 15–19.9 µg/dL and 20 µg/dL. Only tests statistically associated with lead were used in subsequent analyses. Linear models were used for simplicity of interpretation but a quadratic term was also significant, indicating some degree of nonlinearity of relations. The quadratic term did not alter the basic inference about the strength of the relation between lead and cognitive function. In Model 1, the lead-cognition relation was adjusted for gender, age, SES, and school. In models 2–6, Hb, ZPP, SF, HAZ, and HC were added sequentially to Model 1. The degree of influence of nutrition variables on the lead-cognition relation was assessed by monitoring changes in the lead regression coefficient as nutrition variables were added into the model.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Of the children eligible for participation, 602 were enrolled for baseline evaluations and 595 were available to complete the 2-d cognitive battery during testing at the schools. The sample included 46% girls and children’s ages ranged from 6.2 to 8.5 y (51% 6-y-olds; 47% 7-y-olds; 2% 8-y-olds) (Table 1). The mean blood lead concentration for this population was 11.5 ± 6.1 µg/dL and did not differ between boys and girls. Lead concentrations were 10 µg/dL in 51% of the children and 20% had values 15 µg/dL. Ten percent of the children fell below the altitude-adjusted cut-off value for anemia of 124 g/L (45) but anemia was generally mild, with the lowest value at 100 g/L. Stunting was not a significant problem, with only 2.3% of children having HAZ less than -2 SD, whereas a greater proportion of children would be considered overweight, with WHZ > 2 SD (15.4%).

A negative association was found between blood lead and both Number and Letter Sequencing. Lead exposure was associated with the final score on Number Sequencing, and lead categories alone predicted 1.4% variability in performance (P = 0.003). Letter Sequencing is a more difficult test and many children could only attempt it, i.e., the median final score was 0 points. Thus, children were less likely to perform the task correctly in each higher lead category (P = 0.006).

    Question 2: Are iron status, anemia and stature related to cognitive outcomes? The central question of this paper concerns the degree to which children’s nutritional status influences the lead-cognition relation. Therefore, only those cognitive outcomes significantly associated with lead exposure were further examined for associations with nutritional variables.

Anemia was related to two tests of cognitive performance (Table 3). First, children with Hb <124 g/L performed more poorly on Number Sequencing than children with higher Hb concentrations (nonanemic score: 6.0 ± 3.0 vs. anemic score: 4.8 ± 2.8; P = 0.004 in a 2-tailed t test). Children in the anemic group also tended to commit more errors of commission on the test than nonanemic children (P = 0.056, data not shown). Second, the anemic group tended to give fewer correct answers than the nonanemic group on Letter Sequencing (P = 0.069). Similarly, children with ZPP concentrations 100 µmol/mol heme, the most abnormal category, tended to have poorer performance in Letter Sequencing than children with better iron status (Table 3; P = 0.056). Serum ferritin was associated only with WISC-RM Coding final score in which the performance of iron-replete children was worse than that of iron-deficient children (P = 0.012).

    Are lead and nutritional factors associated with each other? Anemia was not significantly associated with iron deficiency in this sample. The prevalence of elevated ZPP levels was substantial (31.8% children with ZPP 70 µmol ZP/mol heme), but the proportion of anemia did not differ among children with ZPP 70 µmol ZP/mol heme vs. children with ZPP < 70 µmol ZP/mol heme (12.6 vs. 8.8%; P = 0.149). The prevalence of iron deficiency, as measured by both ferritin < 15 µg/L and ZPP 70 µmol ZP/mol heme, tended to be higher in anemic than nonanemic children but this difference was not significant (16.4 vs. 7.7%, P = 0.161).

Iron status was correlated with blood lead. Among children with lead concentrations <10 µg/dL, 27.8% had elevated ZPP (70 µmol ZP/mol heme), compared with 35.5% of children with lead 10 µg/dL (P = 0.042). Low ferritin (<15 µg/L) was also more prevalent in children with PbB above than below 15 µg/dL (33 vs. 18.4%; P = 0.001).

Anthropometric measures were associated with both lead and Hb. HAZ decreased with increasing lead concentrations (Spearman = -0.16, P < 0.01) but were positively correlated with HC ( = 0.35, P < 0.01). Both HAZ and HC were also correlated with Hb and increased as Hb concentrations increased ( = 0.13 and = 0.12, respectively; both P < 0.01).

    Question 3: Does nutritional status explain the association between lead and cognitive outcomes? Lead was modeled as an independent predictor of cognitive test scores, with age, gender, SES, and school as covariates to obtain the adjusted lead coefficient. Anemia, iron deficiency, HAZ, and HC were then entered sequentially into the regression model as independent predictors of cognitive function. Detailed model building is shown in Table 4 for the Peabody Picture Vocabulary Test. Regression analysis for the other outcomes was performed in the same manner, with the results described in the text.

For WISC-RM Coding, the adjusted score difference between children with lead concentrations below and above 10 µg/dL was 0.6 points on a 15-point scale (P = 0.014). Hemoglobin, ZPP, HAZ, or HC were not associated with Coding performance and did not change lead’s relation with Coding in the regression model. SF, however, strengthened the association of lead with Coding score (lead ß coefficient = -0.68, P = 0.011). SF was also independently associated with WISC-RM Coding in the regression model, i.e., for each 1 µg/L increase in SF concentration, the final score decreased by 0.02 points (P = 0.004).

The adjusted association between lead and Number Sequencing was significant (P = 0.045) and remained so after all nutritional variables were added to the model (P = 0.034). In the final model (all variables included), for every 1 µg/dL increase in lead, the Number Sequencing score decreased by 0.06-point (on a 15-point scale). Anemia was also independently associated with Number Sequencing. After adjusting for lead and confounders, there was a 1.28-point difference in performance between anemic and nonanemic children, with the anemic group scoring lower (P = 0.002). With additional nutritional variables in the model, the difference between anemic children and the rest of the sample increased further (anemia ß coefficient = -1.426, P = 0.001).

The likelihood of obtaining any correct responses on the Letter Sequencing task was associated with blood lead in an unadjusted logistic regression (Table 2) but after adjusting for covariates, this relation was not significant (odds ratio = 0.967, P = 0.057). The nutritional variables of interest did not further change this relation.

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Lead exposure was independently associated with performance on 4 cognitive tasks, i.e., Peabody Picture Vocabulary Test, WISC-RM Coding, and Number and Letter Sequencing, which measure receptive vocabulary skills and visual spatial abilities. Tests assessing aspects of memory and attention (WISC-RM Digit Span and Arithmetic) were not associated with lead in this population, possibly because they were not sensitive enough to capture lead’s effects. This study is important because negative relations between lead, attention, mental flexibility and perceptual skills have been reported in children previously, but here the strongest finding is the negative association between lead exposure and verbal recognition.

The 3 WISC tasks described in this report were used in previous studies of lead exposure in children. Stiles and Bellinger (3) examined the association between postnatal blood lead concentrations and cognitive function of 10-y-old children using the Wechsler Intelligence Scale for Children-Revised (n = 148). The means ± SD for the Arithmetic, Digit Span and Coding subscales in the Stiles and Bellinger (3) study were higher (12.1 ± 2.8, 10.9 ± 2.7 and 11.4 ± 3.0, respectively) than in the present sample, possibly because their subjects were older. In addition, Arithmetic and Digit Span, but not Coding performance, were negatively associated with blood lead at 24 mo in adjusted analysis (P = 0.01 and 0.056, respectively). Tong et al. (4) also administered the same three tasks to school children (11–13 y old) and found that only Arithmetic was negatively related with lead exposure (P = 0.01) in adjusted linear regression. Finally, Hansen et al. (7) found no differences on the 3 WISC subtests between matched pairs of first-grade Danish children with high and low dentine lead.

The associations between lead and cognitive performance found in this study were remarkably robust and not strongly affected by the nutritional factors we measured. Stature slightly attenuated the lead-PPVT relation, lowering lead’s regression coefficient by 13.3% (Table 4). The addition of anemia and iron variables actually increased the lead coefficient by 7.1% in the WISC-RM Coding regression. The association between lead and Number Sequencing was also stronger (25% increase in lead’s coefficient) when ZPP and ferritin were added into the model. Finally, the association of lead with Letter Sequencing did not change with the addition of nutritional status indicators.

The range of lead in this population was wide, and 51% of children had concentrations 10 µg/dL, allowing for the examination of cognitive outcomes at various exposures. The level of iron deficiency and anemia, however, was relatively low and most children were very well off in terms of iron status (Table 1). Moreover, stunting was nonexistent, with the percentage of children less than -2 SD HAZ comparable to normal height variation in a well-nourished population. It is possible, therefore, that the low degree of influence of nutritional factors on the lead-cognition relation was due to the small variability in these indicators rather than to the absence of such effects. Stronger interrelations are plausible in a population in which higher degrees of iron deficiency, anemia, and stunting are present.

The results of this study do not indicate that either iron deficiency or anemia is essential for the causal pathway between lead and cognitive deficits; in this population, lead exerted its negative effects on cognition independently of nutritional status. Other cross-sectional studies have also failed to show a clear relation between lead and iron status in children (46,47). However, a recent longitudinal assessment revealed that young children who were iron deficient on 2 consecutive visits were more likely to be lead exposed than children who became iron replete or who were never deficient (48). Because iron is thought to compete with lead for absorption (49,50), a supplementation trial may be necessary to clarify the association between iron and lead. It is possible that iron supplementation would also improve cognitive outcomes, both through improving iron status and lowering lead levels.

Although nutritional factors did not explain the association between lead and test performance, some of them were independently linked to cognition. Hemoglobin concentration was significantly associated with two outcomes, Peabody Picture Vocabulary Test and Number Sequencing. For the PPVT, every unit (g/L) increase in Hb was associated with 0.16-point increase on the final test score, when adjusted for all other covariates and nutritional variables (Table 4, Model 1). Furthermore, compared with nonanemic children, the mildly anemic group scored 1.28 points lower on Number Sequencing, after controlling for gender, age, SES, and school (P = 0.002, data not shown). This relation became even stronger with lead and nutritional variables in the model and remained highly significant (anemia ß coefficient = -1.43, P = 0.001).

Serum ferritin was negatively associated with WISC-RM Coding. Why iron deficient children should perform better on this task than iron-replete children is not clear. Approximately 7% of children had elevated CRP levels (indicating infection) around the time of cognitive testing, which could have caused iron-deficient children to appear replete due to iron sequestration by storage molecules. However, adjusting the regression model for the presence of elevated CRP levels did not change the outcome (data not shown), suggesting that the negative relation between WISC-RM Coding and iron status is not due to underestimation of iron deficiency in the sample. Hemochromatosis was also unlikely because all children had SF concentrations within normal range. This negative relation was small (0.2 points for every 10 µg/L increase in SF) in magnitude and isolated, i.e., SF was not significantly associated with any other outcomes. Thus, the biological importance of this observation is unclear.

Also of interest is the significant association of stature with PPVT and Letter Sequencing seen in simple linear regression. The association between height and cognitive outcomes was reported previously (51). One possible explanation for this relation is that, to some degree, stature reflects mental maturation (52,53). Additional analysis of the present population revealed that HAZ was highly correlated with HC (r² = 0.35, P < 0.001), an indicator of brain growth (54). Another explanation is that variation in height reflects variability in SES (55) and that social disadvantages and adverse events, such as illness and malnutrition, crowding, loss of sleep or stress, which often go hand-in-hand with fewer opportunities for learning, influence both the physical (56) and the mental development of children. In our sample, the association of stature with PPVT became attenuated, once lead, nutritional factors, and psychosocial confounders were accounted for in regression analysis. Furthermore, in one study, height alone accounted for only a small percentage of variability in cognitive performance, compared with social class variables (51).

Finally, the association of HAZ with cognitive performance might be explained to some extent by parent and teacher perceptions of ability in taller children. In one study boys’ size (height controlled for age) was positively related to teacher ratings of academic, athletic, and social competence as well as independence (57). It is possible that parents (57) and teachers (58) behave differently toward children they consider more mature, gifted, or likely to do well in school. Studies of teacher expectancy effects suggest that children expected to have better intellectual gains do perform better in school than controls (58). Thus, if parents see their taller children as more mature, they may use richer vocabulary or decide to teach them letters and numbers sooner than shorter children. If teachers perceive taller children as more mature or likely to achieve, they may give these children extra help, responsibility, attention, or praise. Such behavior may also contribute to taller children’s better motivation or belief in their abilities.

The study’s cross-sectional design is its major limitation. It does not permit knowledge of lead exposure history or nutritional deficiencies of the study population. Although the children in our study were well nourished, we cannot rule out previous periods of malnutrition and its effects on cognition, separate from lead effects. The majority of our children lived in Torreón for >5 y (Table 1). Blood lead, nutritional status, or cognitive performance (with the exception of Number Sequencing: = 0.123, P < 0.01 and HC; = 0.103, P = 0.013) were not correlated with the length of time in Torreón (data not shown). Previous reports from our study area showed higher lead exposures in school children (59), suggesting a trend toward decreasing blood leads. Thus, relating lead concentrations to cognitive performance at only one time point may not reflect a true lead-related deficit. In some prospective studies, the highest associations of cognitive outcomes in school children were with lead concentrations at younger ages (3).

In conclusion, in this study sample, the relation of lead to cognitive performance was not explained by lead’s association with iron deficiency, anemia, or stature. We conclude, therefore, that IDA is not an essential part of the causal pathway between lead and cognitive deficits and that lead effects on cognition are direct and independent of nutritional status. Because this was a cross-sectional analysis, it is not a definitive assessment of the causal web relating lead, nutritional status, and cognitive outcomes. It remains to be demonstrated whether nutritional interventions, such as iron supplementation, can improve cognitive function in lead-exposed children, and if so, whether iron acts by attenuating the adverse effects of lead or through independent mechanisms.

	ACKNOWLEDGMENTS

 
We give special thanks to Ernesto Pollitt for advice on study design and to Joanne Katz for help with analysis. We thank Arturo Cebrian, Grissel Concha, Brenda Gamez, Julio Gaviño, Magdalena Gutiérrez, Gabriel León, Alicia Luna, Francisco Marentes, Rosa Isela Morales, Carina Sosa, and Griselda Torres for their dedication to quality data collection and success of the project. We also thank the project nurse, Remedios Sánchez, for her special touch in taking blood and help with the study. We thank Bertha Castellanos from the Coahuila Ministry of Health for approving the study protocol and the children, parents, and teachers participating in the study for the opportunity to look at issues related to lead exposure.

	FOOTNOTES

 
¹ Funded by the Spencer Foundation, Chicago, IL.

³ Abbreviations used: CAT SD, Cognitive Abilities Test, Stimulus Discrimination task; CRP, C-reactive protein; HAZ, height-for-age Z-score; Hb, hemoglobin; HC, head circumference; IDA, iron deficiency anemia; MAT, Math Achievement Test; PbB, blood lead concentration; PPVT, Peabody Picture Vocabulary Test; SES, socioeconomic status; SF, serum ferritin; WHZ, weight-for-height Z-score; WISC-RM, Wechsler Intelligence Scale for Children-Revised Mexican Version; ZPP, zinc protoporphyrin.

⁴ To convert lead concentration from µg/dL to µmol/L, multiply by a factor of 0.0483.

Manuscript received 4 August 2003. Initial review completed 26 September 2003. Revision accepted 4 November 2003.

	LITERATURE CITED

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 

1. Bellinger, D. C., Stiles, K. M. & Needleman, H. L. (1992) Low-level lead exposure, intelligence and academic achievement: a long-term follow-up study. Pediatrics 90:855-861.[Abstract/Free Full Text]

2. Muñoz, H., Romieu, I., Palazuelos, E., Mancilla-Sanchez, T., Meneses-Gonzalez, F. & Hernandez-Avila, M. (1993) Blood lead level and neurobehavioral development among children living in Mexico City. Arch. Environ. Health 48:132-139.[Medline]

3. Stiles, K. M. & Bellinger, D. C. (1993) Neuropsychological correlates of low-level lead exposure in school-age children: a prospective study. Neurotoxicol. Teratol. 15:27-35.[Medline]

4. Tong, S., Baghurst, P., McMichael, A., Sawyer, M. & Mudge, J. (1996) Lifetime exposure to environmental lead and children’s intelligence at 11–13 years: the Port Pirie cohort study. Br. Med. J. 312:1569-1575.[Abstract/Free Full Text]

5. Wasserman, G. A., Liu, X., Lolacono, N. L., Factor-Litvak, P., Kline, J. K., Popovac, D., Morina, N., Musabegovic, A., Vrenezi, N., Capuni-Paracka, S., Lekic, V., Preteni-Redjepi, E., Hadzialjevic, S., Slavkovic, V. & Graziano, J. H. (1997) Lead exposure and intelligence in 7-year-old children: the Yugoslavia prospective study. Environ. Health Perspect. 105:956-962.[Medline]

6. Needleman, H. L., Gunnoe, C., Leviton, A., Reed, R., Peresie, H., Maher, C. & Barrett, P. (1979) Deficits in psychologic and classroom performance of children with elevated dentine lead levels. N. Engl. J. Med. 300:689-695.[Abstract]

7. Hansen, O. N., Trillingsgaard, A., Beese, I., Lyngbye, T. & Grandjean, P. (1989) A neuropsychological study of children with elevated dentine lead level: assessment of the effect of lead in different socioeconomic groups. Neurotoxicol. Teratol. 11:205-213.[Medline]

8. Silva, P. A., Hughes, P., Williams, S. & Faed, J. M. (1988) Blood lead, intelligence, reading attainment and behaviour in eleven year old children in Dunedin, New Zeland. J. Child Psychol. Psychiatry 29:43-52.[Medline]

9. Bellinger, D., Leviton, A., Allred, E. & Rabinowitz, M. (1994) Pre- and postnatal lead exposure and behavior problems in school-aged children. Environ. Res. 66:12-30.[Medline]

10. Dietrich, K. N., Berger, O. G. & Succop, P. A. (1993) Lead exposure and the motor developmental status of urban six-year-old children in the Cincinnati Prospective Study. Pediatrics 91:301-307.[Abstract/Free Full Text]

11. Minder, B., Das-Smaal, E. A., Brand, E.F.J.M. & Orlebeke, J. F. (1994) Exposure to lead and specific attentional problems in schoolchildren. J. Learn. Disabil. 27:393-399.

12. Campbell, T. F., Needleman, H. L., Riess, J. A. & Tobin, M. J. (2000) Bone lead levels and language processing performance. Dev. Neuropsychol. 18:171-186.[Medline]

13. Bushnell, P. J. & Levin, E. D. (1983) Effects of zinc deficiency on lead toxicity in rats. Neurobehav. Toxicol. Teratol. 5:283-288.[Medline]

14. Hashmi, N. S., Kachru, D. N. & Tandon, S. K. (1989) Interrelationship between iron deficiency and lead intoxication (Part 1). Biol. Trace Elem. Res. 22:287-297.[Medline]

15. Hammad, T. A., Sexton, M. & Langenberg, P. (1996) Relationship between blood lead and dietary iron intake in preschool children. Ann. Epidemiol. 6:30-33.[Medline]

16. Goyer, R. A. (1997) Toxic and essential metal interactions. Annu. Rev. Nutr. 17:37-50.[Medline]

17. Pollitt, E., Hathirat, P., Ktchabhakdi, N. J., Missel, L. & Valyasevi, A. (1989) Iron deficiency and educational achievement in Thailand. Am. J. Clin. Nutr. 50:687-697.

18. Bruner, A. B., Joffe, A., Duggan, A. K., Casella, J. F. & Brandt, J. (1996) Randomized study of cognitive effects of iron supplementation in non-anemic iron-deficient adolescent girls. Lancet 348:992-996.[Medline]

19. Seshadri, S. & Gopaldas, T. (1989) Impact of iron supplementation on cognitive functions in preschool and school-aged children: the Indian experience. Am. J. Clin. Nutr. 50:675-686.

20. Angeles, I. T., Schuttink, W. J., Matulessi, P., Gross, R. & Sastroamidjojo, S. (1993) Decreased rate of stunting among anemic Indonesian preschool children through iron supplementation. Am. J. Clin. Nutr. 58:339-342.[Abstract/Free Full Text]

21. Dangour, A. D., Hill, H. L. & Ismail, S. J. (2002) Height, weight and haemoglobin status of 6 to 59-month-old Kazakh children living in Kzyl-Orda region, Kazakhstan. Eur. J. Clin. Nutr. 56:1030-1038.[Medline]

22. Schwartz, J., Angle, C. & Pitcher, H. (1986) Relationship between childhood blood lead levels and stature. Pediatrics 77:281-288.[Abstract/Free Full Text]

23. Frisancho, A. R. & Ryan, A. L. (1991) Decreased stature associated with moderate blood lead concentrations in Mexican-American children. Am. J. Clin. Nutr. 54:516-519.[Abstract/Free Full Text]

24. Kafourou, A., Touloumi, G., Markopoulos, V., Loutradi, A., Papanagiotou, A. & Hatzakis, A. (1997) Effects of lead on the somatic growth of children. Arch. Environ. Health 52:377-383.[Medline]

25. Powell, C. A., Walker, S. P., Himes, J. H., Fletcher, P. D. & Grantham-McGregor, S. M. (1995) Relationships between physical growth, mental development and nutritional supplementation in stunted children: the Jamaican study. Acta Paediatr. 84:22-29.[Medline]

26. Mendez, M. A. & Adair, L. S. (1999) Severity and timing of stunting in the first two years of life affect performance on cognitive tests in late childhood. J. Nutr. 129:1555-1562.[Abstract/Free Full Text]

27. Berkman, D. S., Lescano, A. G., Gilman, R. H., Lopez, S. L. & Black, M. M. (2002) Effects of stunting, diarrhoeal disease, and parasitic infection during infancy on cognition in late childhood: a follow-up study. Lancet 359:564-571.[Medline]

28. Wasserman, G., Graziano, J. H, Factor-Litvak, P., Popovac, D., Morina, N., Musabegovic, A., Vrenezi, N., Capuni-Paracka, S., Lekic, V., Preteni-Redjepi, E., Hadzialjevic, S., Slavkowich, V., Kline, J., Shrout, P. & Stein, Z. (1992) Independent effects of lead exposure and iron deficiency anemia on developmental outcome at age 2 years. J. Pediatr. 121:695-703.[Medline]

29. Wasserman, G., Graziano, J. H., Factor-Litvak, P., Popovac, D., Morina, N., Musabegovic, A., Vrenezi, N., Capuni-Paracka, S., Lekic, V., Preteni-Redjepi, E., Hadzialjevic, S., Slavkowich, V., Kline, J., Shrout, P. & Stein, Z. (1994) Consequences of lead exposure and iron supplementation on childhood development at age 4 years. Neurotoxicol. Teratol. 16:233-240.[Medline]

30. Silbergeld, E. K. (1982) Neurochemical and ionic mechanisms of lead neurotoxicity. Prasad, K. N. Vernadakis, A. eds. Mechanisms of Action of Neurotoxic Substances 1982:1-23 Raven Press New York, NY. .

31. Costa, L. G. (1998) Signal transduction in environmental neurotoxicity. Annu. Rev. Pharmacol. Toxicol. 38:21-43.[Medline]

32. Patrick, G. W. & Anderson, W. J. (2000) Dendritic alterations of cerebellar Purkinje neurons in postnatally lead-exposed kittens. Dev. Neurosci. 22:320-328.[Medline]

33. Beard, J. L. (2001) Iron biology and immune function, muscle metabolism and neuronal functioning. J. Nutr. 131:568S-580S.[Abstract/Free Full Text]

34. Costa, L. G., Guizzetti, M., Lu, H., Bordi, F., Vitalone, A., Tita, B., Palmeri, M., Valeri, P. & Silvestrini, B. (2001) Intracellular signal transduction pathways as targets for neurotoxicants. Toxicology 160:19-26.[Medline]

35. Erikson, K. M., Jones, B. C., Hess, E. J., Zhang, Q. & Beard, J. L. (2001) Iron deficiency decreases dopamine D1 and D2 receptors in rat brain. Pharmacol. Biochem. Behav. 69:409-418.[Medline]

36. Grantham-McGregor, S. & Ani, C. (2001) A review of studies on the effect of iron deficiency on cognitive development of children. J. Nutr. 131:649S-668S.[Abstract/Free Full Text]

37. Tang, H. W., Huel, G., Campagna, D., Hellier, G., Boissinot, C. & Blot, P. (1999) Neurodevelopmental evaluation of 9-month-old infants exposed to low levels of lead in utero: involvement of monoamine neurotransmitters. J. Appl. Toxicol. 19:167-172.[Medline]

38. Miller, D. T., Paschal, D. C., Gunter, E. W., Stroud, P. E. & D’Angelo, J. (1987) Determination of lead in blood using electrothermal atomisation atomic absorption spectrometry with a L’vov platform and matrix modifier. Analyst 112:1697-1699.[Medline]

39. Habicht, J. P. (1974) Estandarizacón de métodos epidemiológicos cuantitativos sobre el terreno. (Standardization of quantitative epidemiologic methods for the field). Bol. Oficina Sanit. Panam. 76:375-84.[Medline]

40. Dunn, L. M., Lugo, D. E., Padilla, E. R. & Dunn, L. M. (1986) Test de Vocabulario en Imágenes Peabody 1986 American Guidance Service Circle Pines, MN.

41. Wechsler, D., Gómez Palacio, M., Padilla, E. R. & Roll, S. (1984) Escala de Inteligencia Para Nivel Escolar Wechsler-RM 1984 El Manual Moderno, México México.

42. Detterman, D. K. (1988) Cognitive Abilities Test 1988 Case Western Reserve University Cleveland, OH.

43. Reitan, R. M. & Wolfson, D. (1992) Neuropsychological Evaluation of Older Children 1992 Neuropsychology Press Tucson, AZ.

44. Conners, C, K. (1997) Conners’ Rating Scales-Revised: Technical Manual 1997 Multi-Health Systems North Tonawanda, NY.

45. Dirren, H., Logman, M.H.G.M., Barclay, D. V. & Freire, W. B. (1994) Altitude correction for hemoglobin. Eur. J. Clin. Nutr. 48:625-632.[Medline]

46. Wolf, A. W., Jimenez, E. & Lozoff, B. (1994) No evidence of developmental ill effects of low-level lead exposure in a developing country. Dev. Behav. Pediatr. 15:224-231.[Medline]

47. Serwint, J. R., Damokosh, A. I., Berger, O. G., Chisolm, J. J., Gunter, E. W., Jones, R. L., Rhoads, G. G. & Rogan, W. (1999) No difference in iron status between children with low and moderate lead exposure. J. Pediatr. 135:108-110.[Medline]

48. Wright, R. O., Shirng-Wern, T., Schwartz, J., Wright, R. J. & Hu, H. (2003) Association between iron deficiency and blood lead level in a longitudinal analysis of children followed in an urban primary care clinic. J. Pediatr. 142:9-14.[Medline]

49. Barton, J. C., Conrad, M. E., Nuby, S. & Harrison, L. (1978) Effects of iron on the absorption and retention of lead. J. Lab. Clin. Med. 92:536-547.[Medline]

50. Bannon, D. I., Portnoy, M. E., Olivi, L., Lees, P. S. J., Culotta, V. C. & Bressler, J. P. (2002) Uptake of lead and iron by divalent metal transporter 1 in yeast and mammalian cells. Biochem. Biophys. Res. Commun. 295:978-984.[Medline]

51. Stathis, S. L., O’Callaghan, M. J., Williams, G. M., Najman, J. M., Andersen, M. J. & Bor, W. (1999) Behavioural and cognitive associations of short stature at 5 years. J. Paediatr. Child. Health 35:562-567.[Medline]

52. Tanner, J. M. (1961) Education and Physical Growth: Implication of the Study of Children’s Growth for Educational Theory and Practice 1961 University of London Press London, UK.

53. Pollitt, E., Mueller, W. & Leibel, R. L. (1982) The relation of growth to cognition in a well-nourished preschool population. Child Dev. 53:1157-1163.[Medline]

54. Lohman, T. G., Roche, A. F. & Martorell, R. (1988) Anthropometric Standardization Reference Manual 1988 Human Kinetics Books Champaign, IL.

55. Mascie Taylor, C. G. (1991) Biosocial influences on stature: a review. J. Biosoc. Sci. 23:113-128.[Medline]

56. Nystrom Peck, M. & Lundberg, O. (1995) Short stature as an effect of economic and social conditions in childhood. Soc. Sci. Med. 41:733-738.

57. Villimez, C., Eisenberg, N. & Carroll, J. L. (1986) Sex differences in the relation of children’s height and weight to academic performance and others’ attributions of competence. Sex Roles 15:667-681.

58. Rosenthal, R. (2002) Covert communication in classrooms, clinics, courtrooms, and cubicles. Am. Psychol. 57:839-849.[Medline]

59. García Vargas, G. G., Rubio Andrade, M., Del Razo, L. M., Borja Aburto, V., Vera Aguilar, E. & Cebrián, M. E. (2001) Lead exposure in children living in a smelter community in region Lagunera, Mexico. J. Toxicol. Environ. Health 62:417-429.

This article has been cited by other articles:

				I. Al-Saleh, N. Shinwari, M. Nester, A. Mashhour, L. Moncari, G. El Din Mohamed, and A. Rabah Longitudinal Study of Prenatal and Postnatal Lead Exposure and Early Cognitive Development in Al-Kharj, Saudi Arabia: A Preliminary Results of Cord Blood Lead Levels J Trop Pediatr, October 1, 2008; 54(5): 300 - 307. [Abstract] [Full Text] [PDF]

				P. M. Suratt, J. T. Barth, R. Diamond, L. D'Andrea, M. Nikova, V. A. Perriello Jr, M. A. Carskadon, and C. Rembold Reduced Time in Bed and Obstructive Sleep-Disordered Breathing in Children Are Associated With Cognitive Impairment Pediatrics, February 1, 2007; 119(2): 320 - 329. [Abstract] [Full Text] [PDF]

				P. M. Suratt, M. Peruggia, L. D'Andrea, R. Diamond, J. T. Barth, M. Nikova, V. A. Perriello Jr, and M. L. Johnson Cognitive Function and Behavior of Children With Adenotonsillar Hypertrophy Suspected of Having Obstructive Sleep-Disordered Breathing Pediatrics, September 1, 2006; 118(3): e771 - e781. [Abstract] [Full Text] [PDF]

				J. A. Jamieson, C. G. Taylor, and H. A. Weiler Marginal Zinc Deficiency Exacerbates Bone Lead Accumulation and High Dietary Zinc Attenuates Lead Accumulation at the Expense of Bone Density in Growing Rats Toxicol. Sci., July 1, 2006; 92(1): 286 - 294. [Abstract] [Full Text] [PDF]

				J. L. TARLETON, R. HAQUE, D. MONDAL, J. SHU, B. M. FARR, and W. A. PETRI JR. Cognitive effects of diarrhea, malnutrition, and entamoeba histolytica infection on school age children in dhaka, bangladesh. Am J Trop Med Hyg, March 1, 2006; 74(3): 475 - 481. [Abstract] [Full Text] [PDF]

				J. A. Rico, K. Kordas, P. Lopez, J. L. Rosado, G. G. Vargas, D. Ronquillo, and R. J. Stoltzfus Efficacy of Iron and/or Zinc Supplementation on Cognitive Performance of Lead-Exposed Mexican Schoolchildren: A Randomized, Placebo-Controlled Trial Pediatrics, March 1, 2006; 117(3): e518 - e527. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kordas, K. Articles by Stoltzfus, R. J. Search for Related Content PubMed PubMed Citation Articles by Kordas, K. Articles by Stoltzfus, R. J.