- Email: [email protected]
Effects of iron therapy on infant blood lead levels
EFFECTS OF IRON THERAPY ON INFANT BLOOD LEAD LEVELS ABRAHAM W. WOLF, PHD, ELIAS JIMENEZ, MD, AND BETSY LOZOFF, MD
Objectives To determine the effects of iron therapy on blood lead levels in infants with mildly elevated lead levels and varied iron status. Methods Infants from a community-derived sample in Costa Rica were categorized into five groups. Group 1 had irondeficiency anemia with hemoglobin levels #105 g/L. Infants in group 2 were iron-deficient with intermediate hemoglobin levels (between 106-119 g/L). These groups were treated with intramuscular iron or 3 months of oral iron. Group 3 (nonanemic irondeficient) and group 4 (nonanemic iron-depleted) were treated with 3 months of oral iron. Group 5 (iron-sufficient) received oral placebo. Results After 3 months of oral iron therapy, nonanemic iron-depleted infants had the greatest decrease in lead levels, followed by nonanemic iron-deficient infants and iron-deficient infants with hemoglobin levels <120 g/L. Lead levels increased among iron-deficient infants with hemoglobin levels <120 g/L who received intramuscular iron and iron-sufficient nonanemic infants who received placebo. Conclusions Changes in lead levels corresponded closely to changes in iron status and were plausible in terms of absorption mechanisms for lead and iron. Correcting and/or preventing iron deficiency appear to be rapid and effective means of improving infant lead levels, even in nonanemic infants. (J Pediatr 2003;143:789-95)
ron deficiency is thought to be the most common single nutrient deficiency in the world. Lead is considered the most common environmental toxin. Both are global health concerns most prevalent in young children. Although public health interventions in developed countries have dramatically decreased the childhood prevalence of both iron deficiency and blood lead levels,1,2 their worldwide prevalence remains alarming.3,4 Associations between iron status and lead levels are supported by human and animal studies. Children with elevated lead levels are more likely to show evidence of iron deficiency,5,6 and children with iron deficiency are more likely to have elevated blood lead levels even after controlling for age, sex, and socioeconomic status.7 Blood lead levels are inversely related to dietary iron intake.8,9 A mechanism for these interconnections is suggested by the presence of a divalent metal transporter in the gut that is distinct from transferrin and ferritin and competitively binds to iron and lead. If iron deficiency is a risk factor for increased lead absorption, then iron treatment may result in decreased lead levels. In contrast with studies showing that increased dietary iron intake is associated with lower blood lead levels,8 some studies suggest that therapeutic iron may actually interfere with lead excretion and elevate lead levels.10-12 Such findings are especially alarming if giving iron to children in high-risk populations may increase their vulnerability to the effects of lead toxicity. However, the available studies have several methodological difficulties. The distributions of lead values among high-risk children in previous studies have been skewed by markedly elevated values. Also, iron deficiency has been defined in these studies by dichotomizing children as iron-deficient or iron-sufficient, thus obscuring the physiological spectrum of iron deficiency that progresses from iron sufficiency through depletion and deficiency phases to iron-deficiency anemia. It is plausible that the effects of iron therapy on lead levels may depend on the degree of iron deficiency. In a previous study, we used these stages of iron deficiency to evaluate the effects of iron therapy on the development and behavior of young children.13 Because blood lead
I
From the Department of Psychiatry, MetroHealth Medical Center, School of Medicine, Case Western Reserve University, Cleveland, Ohio; the Hospital Nacional de Ninos, University of Costa Rica, San Jose, Costa Rica; and the Center for Human Growth and Development, Department of Pediatrics and Communicable Diseases, University of Michigan, Ann Arbor, Michigan. Supported by NIH grants R01 HD31606-08 and R01-HD33487-06, to B. L., Principal Investigator. Submitted for publication Dec 9, 2002; revision received July 31, 2003; accepted Aug 22, 2003. Reprint requests: Abraham W. Wolf, PhD, Department of Psychiatry, MetroHealth Medical Center, 2500 Metro Health Dr, Cleveland, OH 44109. E-mail: [email protected]. Copyright ª 2003 Mosby, Inc. All rights reserved. 0022-3476/2003/$30.00 + 0 10.1067/S0022-3476(03)00540-7
789
Table. Blood lead levels and iron status measures before and after 3 months of iron therapy by iron status and treatment groups* Pretreatment iron status Anemic iron-deficient Treatment N
Intramuscular iron n = 18
Intermediate iron-deficient
3 Months oral iron n = 30
Intramuscular iron n = 12
3 Months oral iron n = 26
Pretreatment 3-Month Pretreatment 3-Month Pretreatment 3-Month Pretreatment 3-Month 9.9 (0.8) 10.2 (0.5)NS 0.3 (0.4)A,B 0.1 (0.4)A,B
Lead lg/dL Change Change adjusted for pretreatment lead Iron-sufficient{ pretreatment Iron-sufficient after 3 months Hemoglobin, g/L Change Ferritin, lg/L Change Transferrin saturation, % Change Free erythrocyte protoporphyrin, lg/dL of packed red cells Change
10.9 (0.5) 10.3 (0.5)NS 0.6 (0.5)A,B 0.6 (0.3)A,B
11.0 (0.6) 11.5 (0.5)NS 0.5 (0.6)A,B 0.5 (0.5)A,B
11.7 (1.2) 11.2 (0.8)NS 0.5 (0.6)A,B 0.2 (0.4)A,B
0%
0%
0%
0%
6%
57%
17%
77%
93.1 (2.1) 122.6 (2.5)§ 29.5 (3.2)A,B 3.2 (0.4) 9.9 (1.9)NS B,C 6.7 (1.7) 10.1 (1.3) 13.2 (0.9)NS 3.2 (1.7)B 341.9 (47.8) 101.4 (9.1)§
240.5 (41.5)C
96.5 (1.6) 132.2 (1.7)§ 35.7 (2.4)A 3.5 (0.5) 22.3 (3.5)§ A,B 18.8 (3.4) 9.1 (0.4) 25.5 (1.8)§ 16.4 (1.9)A 313.9 (30.0) 85.3 (7.2)§
228.6 (26.8)C
111.8 (1.2) 126.4 (2.6)§ 14.6 (2.3)C,D 3.8 (0.7) 9.8 (2.1)NS B,C 5.9 (2.1) 11.8 (0.8) 20.5 (2.5)z A,B 8.7 (2.2) 161.3 (13.9) 92.9 (14.6)z
68.3 (10.0)A,B
111.2 (0.9) 131.7 (1.8)§ 20.5 (2.12)C,D,E 4.0 (0.5) 25.3 (2.2)§ A,B 21.3 (2.1) 10.6 (0.6) 22.9 (2.2)§ A,B 12.3 (2.1) 156.7 (13.4) 64.5 (4.1)§
92.2 (11.6)B
*Values are means and (SEM). Cells with the same letter do not differ; Tukey test for multiple comparisons used. yp < 0.05. zp < 0.005. §p < 0.001. iTest of interaction of iron status/treatment groups by difference of pretreatment and 3-month measures. {Criterion for iron sufficiency was three measures of iron status in the normal range: serum ferritin $ 12 lg/L, transferrin saturation $ 10%, free erythrocyte protoporphyrin <100 lg/dL.
levels were obtained as part of the hematologic evaluation, the design of that study can also be used to test the following hypothesis: iron therapy will decrease lead levels in infants whose iron status is compromised.
METHODS Cohort and Sample Selection Information on blood lead levels was obtained in the course of a community-based study of iron deficiency anemia, iron therapy, and infant development. Between 1983 and 1985, families in a lower-middle-class community near San Jose, the capital of Costa Rica, were contacted in a house-tohouse survey for participation in the study. All children between 12 and 23 months old were initially screened by fingerstick for anemia (no family refused to be screened). Infants who met the following inclusion criteria were asked to be brought for a venous blood sample: birth weight 2.5 kg, singleton birth, no major congenital anomalies, no major perinatal complications, no jaundice treated with pho790
Wolf, Jimenez, and Lozoff
totherapy, no hospitalization during the preceding 6 months, no hospital admission at any age for other than an acute uncomplicated problem, no iron therapy after 6 months of age, no intramuscular iron treatment at any age, no iron-containing vitamins or formula within the last month, no known developmental delay, and no major chronic or acute illness. The selection of the final sample was based on a normal pediatric physical examination, venipuncture values for hemoglobin, iron measures, and a lead value <25 lg/dL, the Centers for Disease Control and Prevention cutoff at that time. Hemoglobin values were divided into three levels: nonanemic, >120 g/L; intermediate, between 106 and 119 g/L; and anemic, <105 g/L. Iron status was defined as follows: iron-sufficient, three measures of iron status in the normal range (serum ferritin >12 lg/L, transferrin saturation >10%, free erythrocyte protoporphyrin #100 lg/dL packed red blood cells); iron-depleted, a low serum ferritin level (#12 lg/ L) as the sole indication of iron lack; and iron-deficient, a low serum ferritin level and either a high erythrocyte protoporphyrin (>100 lg/dL packed red blood cells) or a low transferrin saturation (#10%). This grouping represents The Journal of Pediatrics December 2003
Pretreatment iron status Nonanemic iron-deficient
Nonanemic iron-depleted
Iron-sufficient
3 Months oral iron n = 16
3 Months oral iron n = 34
3 Months oral placebo n = 29
Pretreatment
3-Month
Pretreatment
11.4 (0.7)NS
11.9 (0.9)
3-Month
11.6 (0.5) 10.2 (0.3)§ B 1.4 (0.4)
A,B
0.5 (0.4)
0.1 (0.5)A,B
Pretreatment
3-Month
Analysis of variance
10.9 (0.5)y
9.9 (0.4) A
F(6/158) = 3.5zi
1.2 (0.3)B 0%
0.5 (0.3)A 100%
F(6/157) = 2.9z
91%
62%
1.0 (0.3)
0% 88% 136.4 (2.4)§
124.2 (1.3) 12.3 (2.4) 4.3 (0.8)
9.1 (1.3) 32.4 (6.2)
§
28.2 (5.7)A 12.1 (0.9)
E
6.5 (0.4)
41.9 (5.0)
§
35.3 (4.9)A 24.1 (1.9)
§
12.1 (2.0)A,B 123.3 (10.0)
136.1 (1.5)§
127.1 (0.8)
D,E
65.1 (9.1)
y
58.2 (7.7)A,B
§
20.6 (1.1) 31.6 (2.5) 11.07 (2.6)A,B 65.2 (4.3) 42.8 (3.2)NS
22.4 (3.3)A,B
a continuum from least to most iron-deficient and approximates the physiologic progression involved in developing iron-deficiency anemia. Because nonanemic iron-depleted infants were disproportionately common in the original sample, a random selection procedure was instituted so that only one in five such infants was included in the final sample. The proportion of infants in the study sample in each iron status category was as follows: iron-deficient anemic, 17%; intermediate iron-deficient, 21%; nonanemic iron-deficient, 10%; nonanemic iron-depleted, 23%; and nonanemic ironsufficient, (29%). A total of 191 infants met criteria for the iron study and agreed to participate. Two infants had blood lead levels >25 lg/dL and thus did not qualify. For this analysis, however, their data were included. Pretreatment lead levels were available for 184 infants; 165 had lead data both pretreatment and after 3 months. There was no difference between the lead levels for these 165 infants who remained in the study and the 19 whose lead levels were not available at 3 months. Effects of Iron Therapy on Infant Blood Lead Levels
129.4 (1.2) 132.5 (1.3)NS D 3.1 (1.3) 22.9 (1.9) 14.9 (1.0)y C 8.0 (2.2) 21.4 (1.2) 24.5 (1.2)NS B 3.1 (1.4) 61.4 (3.8) 57.4 (3.1)NS
4.0 (3.0)A
F(6/158) = 34.5§ F(6/158) = 17.7§ F(6/158) = 5.6§
F(6/158) = 27.9§
The average age of the 165 infants was 17.3 months (0.3 SEM), ranging from 13 to 24 months; 55.8% (92/165) were male. The weight at birth was 3.37 kg (0.08 SEM), and the mother’s age at delivery was 25.6 years (0.4 SEM). Mothers had an average of 8.8 years of schooling (0.2 SEM) and breastfed their infants an average of 29.1 weeks (1.9 SEM). The infants averaged between the 40th and 50th percentile for weight and length using US standards. In a previous study, we reported that the infant blood lead levels were not related to developmental outcome measures at infancy or in 5-year follow-up.14 The treatment component was a double-blind design with short-term and long-term aspects. In the short-term (1-week) treatment aspect, infants in the anemic and intermediate groups (hemoglobin <120 g/L) were randomly assigned intramuscular iron, oral iron, or oral placebo treatment, and nonanemic infants were randomly assigned oral iron or placebo for 1 week. All parenterally treated infants were given oral placebo drops the first week; all infants who did not receive parenteral iron received a placebo injection. After the 791
first week, the long-term treatment design was applied. Parenterally treated infants and iron-sufficient infants were given placebo drops for 12 weeks. All other infants were treated with oral iron. The doses of intramuscular iron were calculated to correct anemia and increase hemoglobin levels to 125 g/L. To ensure intake of oral medication, project personnel workers visited families daily to administer oral drops. Infants who received an additional week of oral iron in the short-term component did not differ from others treated with oral iron. Therefore, their data were combined with data from the other infants. To assess the effects of intramuscular and oral iron treatment on lead levels at different stages of iron deficiency, seven iron status and treatment groups were formed: (1) anemic iron-deficient infants treated with intramuscular iron, (2) anemic iron-deficient infants treated with oral iron, (3) intermediate iron-deficient infants treated with intramuscular iron, (4) intermediate iron-deficient infants treated with oral iron, (5) nonanemic iron-deficient infants treated with oral iron, (6) nonanemic iron-depleted infants treated with oral iron, and (7) iron-sufficient infants, who received oral placebo.
Measurement of Blood Lead Red blood cells were promptly separated and frozen for future analysis in the United States. The frozen red cells were analyzed in a laboratory at MetroHealth Medical Center that participated in the Centers for Disease Control Maternal and Child Health Resources Development Proficiency Testing Program for blood lead. Lead levels were determined by a graphite furnace method15,16 using a matrix modification procedure (0.5% ammonium phosphate, 0.5% Triton X-100 detergent, Amersham Biosciences Corp, Piscataway, NJ). Samples were diluted by using a Sylva diluter and analyzed by using an autosampler and pyrocoated graphite furnace (Perkin Elmer 560 AA, 2200 furnace, AS-40 Autosampler, Perkin Elmer, Wellesley, Mass). Quality control of the lead analyses was monitored through certified controls obtained from the National Bureau of Standards. Red cell lead values were converted to whole blood lead levels by using the formula of Rosen et al.17
Data Analysis Statistical analyses followed the general linear model for continuous variables. Changes in lead levels and iron status measures within and among iron status and treatment groups were compared by using a mixed model: pretreatment and 3month lead and iron measures were specified as random effects and iron status and treatment groups as fixed effects. Differences between pretreatment and 3-month lead and iron measures were assessed by using paired t tests with the error terms from the mixed model. Multiple comparisons among iron status and treatment groups were assessed by the Tukey test. Changes in lead levels were also assessed by using analysis of covariance to control for the effects of pretreatment levels. Categorical variables were compared by using the v2 test. All analyses were performed with standard computational 792
Wolf, Jimenez, and Lozoff
packages.18 A two-tailed a level of 0.05 was defined for tests of statistical significance.
RESULTS The mean blood lead level for the 165 infants before iron treatment was 10.98 lg/dL (0.26 SEM), ranging from 5.8 to 36.9 lg/dL; 57.6% (95/165) had lead levels >10 lg/dL. Pretreatment blood lead levels for the seven iron status and treatment groups did not differ (F = 1.4, df = 6/158, NS). The Table lists the values for blood lead levels and iron status measures at pretreatment and 3 months for the seven iron status and treatment groups. Two groups had significant intragroup changes in lead levels at 3 months. The nonanemic iron-depleted infants (infants whose sole indicator of iron deficiency was a low serum ferritin and who received 3 months of oral iron) decreased lead levels by 1.4 lg/dL, from 11.6 lg/dL to 10.2 lg/dL (P = .0004); after adjusting for pretreatment lead levels, the difference of 1.2 lg/dL remained statistically significant (P = .0002). Infants iron-sufficient at pretreatment, who received 3 months of oral placebo, increased lead levels by 1.0 lg/dL, from 9.9 lg/dL to 10.9 lg/dL (P = .025); after adjusting for pretreatment levels, the value of 0.5 lg/dL was not significant (P = .12). The intergroup comparisons of change in lead levels indicated that nonanemic iron-depleted infants had the greatest decrease in lead levels, whereas ironsufficient infants who received placebo showed the greatest increase in blood lead concentration. These intergroup differences remained significant after adjusting for pretreatment levels and after adjusting for the effects of age and sex. To help understand how iron therapy affected lead levels, we first considered how treatment affected iron status. After 3 months, 93% of all infants treated with intramuscular or oral iron showed an increase in hemoglobin of at least 10 g/ L. The anemic and intermediate iron-deficient infants had the greatest increase in hemoglobin. Iron therapy corrected the anemia in all children but not all three biochemical indicators of iron deficiency. Ten percent (3/30) of children who received intramuscular iron and 77% (82/106) of children who received oral iron became iron-sufficient. In contrast, 62% (18/29) of children who were iron-sufficient at the start of the study developed signs of iron lack after 3 months of oral placebo. We also examined changes in each iron status measure: ferritin, transferrin saturation, and erythrocyte protoporphyrin. These measures improved in all groups receiving iron treatment, but improvement varied considerably by type of treatment and level of iron deficiency. Among anemic and intermediate irondeficient infants, each iron status measure improved more in infants treated with 3 months of oral iron than in infants treated with intramuscular iron. The groups of iron-deficient and iron-depleted nonanemic infants showed even greater levels of improvement. Only one group deteriorated on any measure of iron status: children who started the study as ironsufficient had significantly lower levels of ferritin after 3 months of placebo. The Journal of Pediatrics December 2003
Figure. Mean change from pretreatment to 3-month evaluation of infant blood lead and ferritin levels plotted by iron status and treatment groups. Hatched areas represent 1 SEM of the changes in ferritin and lead levels.
Of the individual measures of iron status, changes in ferritin were closely associated with changes in iron status. The groups with the greatest proportions becoming iron-sufficient showed the most decreases in lead levels. After 3 months of oral iron, nonanemic infants who started the study with a low serum ferritin as the sole abnormal measure of iron status showed the most marked decrease in lead levels, followed by nonanemic infants with iron deficiency (two or more abnormal iron measures at study outset) and iron-deficient infants with hemoglobin levels <120 g/L. The two groups showing an increase in lead levels were ones in which either fewer infants became iron-sufficient or infants showed evidence of developing iron deficiency: iron-deficient infants with hemoglobin levels <120 g/L who received intramuscular iron and iron-sufficient nonanemic infants who were given placebo for 3 months. Changes in ferritin levels were most closely associated with changes in lead levels. The Figure shows the mean changes in blood lead and ferritin levels plotted by iron status and treatment groups. Nonanemic iron-depleted children showed a decrease in lead levels of 1.4 lg/dL and an increase in ferritin of 35.3 lg/L, the largest ferritin increase of any Effects of Iron Therapy on Infant Blood Lead Levels
compromised group. Children iron-sufficient at the start of the study, who received placebo, showed the greatest increase in lead levels, 1.0 lg/dL, and a decrease in ferritin of 8.0 lg/L. This was the only group in which iron status worsened.
DISCUSSION These findings support the hypothesis that iron therapy can decrease lead levels in children whose iron status is compromised. Even over the short period of 3 months, changes in blood lead levels were closely associated with changes in iron status. However, we had not expected that lead levels would increase in this short time in children who were iron-sufficient at the start of study. It was also unexpected that the effect of iron therapy on lead levels was most marked at the earliest stage of iron deficiency. Lead levels decreased to the greatest degree in nonanemic iron-depleted children, that is, children for whom a low ferritin level was the only indication of iron lack. These results indicate that preventing or correcting iron deficiency was associated with lowering of children’s blood lead levels. 793
Our findings appear plausible in terms of the absorption mechanisms for lead and iron. The fact that iron status worsened in the iron-sufficient group treated with placebo indicates the precarious nature of iron nutrition in this population. Because they were on placebo and the local diet was clearly limited in iron, their iron stores reduced over the 3 months, reflected by decreasing ferritin levels. These decreasing iron stores would signal for increased intestinal absorption of divalent metals, and lead would be absorbed along with what little iron was available. The situation for the intramuscular-treated infants was related, in an unanticipated way. The dose of intramuscular iron had been calculated to bring hemoglobin levels to 125 g/L. However, the final hemoglobin level for infants given oral iron was 135 g/L, and a much higher proportion became iron-sufficient, indicating that the intramuscular dose was insufficient. Intramuscular iron, although improving iron status, did not lower blood lead levels, because it did not compete for absorption in the gastrointestinal tract. Because intramuscular iron-treated infants remained iron-deficient throughout the study, intestinal absorption of both iron and lead would also have remained active. Further mechanisms must be invoked to explain differing changes in lead levels among the groups of infants who received oral iron therapy. Although there is a shared transporter for iron and lead, these infants all received comparable doses of oral iron on a per kilogram basis. Thus, differential competition for the divalent metal transporter does not seem to explain between-group differences. We postulate that progressive shutting down of intestinal absorption of iron—and lead—as subjects became more iron-sufficient contributed to the differences. Based on the physiology of iron absorption, one would expect that groups becoming ironsufficient most rapidly would show the fastest shut-down of intestinal absorption, and correspondingly, the most pronounced drop in blood lead levels. Our results are consistent with this expectation. Nonanemic iron-depleted infants, who had the least degree of iron deficiency, probably became ironsufficient earliest, as evidenced by the highest ferritin levels and highest proportion iron-sufficient in this group. Correspondingly, they had the greatest reduction in lead levels. Oral iron-treated infants with greater degrees of iron deficiency also increased iron stores and reduced lead levels, but not as much as the iron-depleted infants. Our results contrast with some previous studies suggesting iron therapy increased lead levels, either by facilitating lead absorption10 or by slowing lead excretion.11 We used a community-derived sample with groups that approximated the physiological progression from iron sufficiency to anemia. Previous study groups had higher pretreatment lead levels—means of 31.1 lg/dL and 28.3 lg/dL, respectively—compared with 10.98 lg/dL in our sample. Furthermore, definitions of iron deficiency in those studies were based solely on a ferritin level, with cut-offs of 8.0 lg/dL and 16.0 lg/dL, respectively, and the number of ‘‘irondeficient’’ children was very small (23 and 20, respectively). Dichotomizing iron status by ferritin alone obscures the 794
Wolf, Jimenez, and Lozoff
process of developing iron deficiency and may fail to detect important differences along that dynamic spectrum. As Pollitt19 states, the several stages of iron-depletion lead to incremental involvement of different systems in the body that activate distinct mechanisms within and among domains. Yet even the identification of iron status along a physiological spectrum will not fully capture lead–iron dynamics. The association between lead levels and iron status was appreciated only with pretreatment and posttreatment measurement. Thus, to understand lead–iron interactions fully, it appears to be essential to examine changes in iron status in addition to the degree of initial iron deficiency by using several hematologic parameters. Our sample was drawn from a study originally designed to evaluate the effects of iron therapy on the cognitive and behavioral development of healthy infants. Therefore, results cannot be generalized to children who are not in good health or who show generalized undernutrition. Further studies are required to assess the effects of iron treatment in children with more severe lead toxicity. Conversely, it is not clear what the effects of iron therapy would be on lead absorption in children with even lower lead levels. Furthermore, the study was conducted in a setting in which compromised iron status in infancy was close to universal. Results may be different if infant iron status is generally good. Mahaffey20 has stressed that lead toxicity is caused by exposure to lead and not by nutritional deficiency. However, the fractional transfer of lead from the environment to target tissue can be modified by nutritional status. Although public health programs to reduce lead toxicity need to include nutritional interventions, such interventions are clearly secondary to the elimination of environmental sources of lead. Our results indicate that correcting iron deficiency may be a rapid and effective means of improving infant lead levels further, even in nonanemic infants. As countries succeed in lowering lead levels by environmental means, improving iron status may help achieve even further reductions.
REFERENCES 1. Halterman JS, Kaczorowski JM, Aligne CA, Auinger P, Szilagyi PG. Iron deficiency and cognitive achievement among school-aged children and adolescents in the United States. Pediatrics 2001;107:1381-6. 2. CDC. Blood lead levels in young children—US and selected states, 1996-1999. MMWR Morb Mortal Wkly Rep 2000;49:1133-7. 3. Stoltzfus RJ. Iron-deficiency anemia: reexamining the nature and magnitude of the public health problem. Summary: implications for research and programs. J Nutr 2001;131:697S-700S. 4. Wasserman GA, Liu X, Popovac D, Factor-Litvak P, Kline J, Waternaux C, et al. The Yugoslavia Prospective Lead Study: contributions of prenatal and postnatal lead exposure to early intelligence. Neurotoxicol Teratol 2000;22:811-8. 5. Yip R, Norris TN, Anderson AS. Iron status of children with elevated blood lead concentrations. J Pediatr 1981;98:922-5. 6. Clark M, Royal J, Seeler R. Interaction of iron deficiency and lead and the hematologic findings in children with severe lead poisoning. Pediatrics 1988;81:247-54. 7. Wright RO, Shannon MW, Wright RJ, Hu H. Association between iron deficiency and low-level lead poisoning in an urban primary care clinic. Am J Public Health 1999;89:1049-53.
The Journal of Pediatrics December 2003
8. Hammad TA, Sexton M, Langenberg P. Relationship between blood lead and dietary iron intake in preschool children: a cross-sectional study. Ann Epidemiol 1996;6:30-3. 9. Lanphear B, Hornung R, Ho M, Howard C, Eberly S, Knauf K. Environmental lead exposure during early childhood. J Pediatr 2002;140:40-7. 10. Angle CR, Stelmak KL, McIntire MS. Lead and iron deficiencies. In: DD Hemphill, editor. Trace substances in environmental health. Vol IX. Columbia (MO): University of Missouri; 1975. p. 367-86. 11. Ruff HA, Markowitz ME, Bijur PE, Rosen JF. Relationships among blood lead levels, iron deficiency, and cognitive development in two-year-old children. Environ Health Perspect 1996;104:180-5. 12. Wright RO, Hu H, Maher TJ, Amarasiriwardena C, Chaiyakul P, Woolf AD, et al. Effect of iron deficiency anemia on lead distribution after intravenous dosing in rats. Toxicol Ind Health 1998;14:547-51. 13. Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez E, et al. Iron deficiency anemia and iron therapy effects on infant developmental test performance. Pediatrics 1987;79:981-95.
14. Wolf AW, Jimenez E, Lozoff B. No evidence of developmental III effects of low-level lead exposure in a developing country. J Dev Behav Pediatr 1994;15:224-31. 15. Fernandez FJ. Micromethod of lead determination in whole blood by atomic absorption with use of graphic furnace. Clin Chem 1975;21: 558-61. 16. Fernandez FJ, Hillegos D. An improved graphite furnace method for the determination of lead in blood using matrix modification and the L’vov platform. Atom Spectrosc 1982;3:1300-11. 17. Rosen FJ, Zarate-Salvador C, Trinadad EE. Plasma lead levels in normal and lead-intoxicated children. Pediatrics 1974;84:45-8. 18. SAS. SAS System, release 8.02, Microsoft Windows TS02M0. Cary (NC): SAS Institute; 2002. 19. Pollitt E. The developmental and probabilistic nature of the functional consequences of iron-deficiency anemia in children. J Nutr 2001;131:669S-75S. 20. Mahaffey KR. Nutrition and lead: strategies for public health. Environ Health Perspect 1995;103(suppl 6):191-6.
50 Years Ago in The Journal of Pediatrics ACUTE RESPIRATORY DISEASE WITH SPECIAL REFERENCE TO PATHOGENESIS, CLASSIFICATION, AND DIAGNOSIS
Manson J. J Pediatr 1953;43:599-619 In this comprehensive review from 50 years ago, Manson proposes a classification system for acute respiratory illnesses in children based on the anatomic distribution of infections. Hence, he reviews in order viral and bacterial infections involving the larynx, tracheobronchial tree, lungs, bronchii, and bronchioles, emphasizing the not uncommon finding still recognized today of infections involving the upper airway progressing distally to affect the lower respiratory tract. Quite accurately, he describes the risks and consequences of secondary bacterial infections after primary viral respiratory tract disease. He also believes that a viral etiology exists for bronchiolitis but acknowledges the likelihood of ‘‘several causative viral agents.’’ Finally, he laments the lack of information regarding ‘‘viral behavior,’’ and by extension, the difficulties in isolating causative viruses in respiratory tract disease. Manson’s paper reflects a bias doubtless prevalent 50 years ago: that children are merely ‘‘little adults’’ and as such follow a similar clinical course of respiratory infections as adults with susceptibility to microbiologic pathogens identical to those found in adults. In fact, he relies heavily on several sources from the adult literature to describe causative agents of acute respiratory infections in children including a 1944 study commissioned by the Army and three studies by Reimann in adults from the 1940s. Current pediatric pulmonary textbooks1,2 reflect subsequent studies describing age-specific differences in susceptibility patterns to different viral and bacterial pathogens an the subsequent clinical course. An important example of this is disease caused by the respiratory syncytial virus, capable of causing a severe, protracted illness in infants and small children but causing a mild coryzal syndrome in immunocompetent adults. Manson’s discussion of clinical and physical findings in children with respiratory illnesses is instructive. He emphasizes the importance of the physical examination and its daily, changing pattern as essential to diagnostic accuracy. The use of other ‘‘tools,’’ such as radiographs and laboratory studies, are merely suggested as afterthoughts and by implication, of use only as a last resort. His elegant, paragraph-long description of the physical findings in bronchiolitis is particularly evocative and is as accurate today as it was 50 years ago. In our current reliance on sophisticated microbiologic and radiographic techniques, perhaps we’ve forgotten, to some extent, the abundant clues provided by a careful physical examination. Blakeslee E. Noyes, MD, Division of Pulmonary Medicine Cardinal Glennon Children’s Hospital, St Louis University School of Medicine St Louis, MO 63104
YMPD506 10.1067/S0022-3476(03)00547-X
REFERENCES 1. 2.
Kending EL. Kendig’s disorders of the respiratory tract in children. Chernick V, Boat T, eds. Philadelphia (PA): WB Saunders Co; 1998. Pediatric respiratory medicine. Taussig LM, Landau LI, eds. St Louis (MO): Mosby, Inc; 1999.
Effects of Iron Therapy on Infant Blood Lead Levels
795
Objectives To determine the effects of iron therapy on blood lead levels in infants with mildly elevated lead levels and varied iron status. Methods Infants from a community-derived sample in Costa Rica were categorized into five groups. Group 1 had irondeficiency anemia with hemoglobin levels #105 g/L. Infants in group 2 were iron-deficient with intermediate hemoglobin levels (between 106-119 g/L). These groups were treated with intramuscular iron or 3 months of oral iron. Group 3 (nonanemic irondeficient) and group 4 (nonanemic iron-depleted) were treated with 3 months of oral iron. Group 5 (iron-sufficient) received oral placebo. Results After 3 months of oral iron therapy, nonanemic iron-depleted infants had the greatest decrease in lead levels, followed by nonanemic iron-deficient infants and iron-deficient infants with hemoglobin levels <120 g/L. Lead levels increased among iron-deficient infants with hemoglobin levels <120 g/L who received intramuscular iron and iron-sufficient nonanemic infants who received placebo. Conclusions Changes in lead levels corresponded closely to changes in iron status and were plausible in terms of absorption mechanisms for lead and iron. Correcting and/or preventing iron deficiency appear to be rapid and effective means of improving infant lead levels, even in nonanemic infants. (J Pediatr 2003;143:789-95)
ron deficiency is thought to be the most common single nutrient deficiency in the world. Lead is considered the most common environmental toxin. Both are global health concerns most prevalent in young children. Although public health interventions in developed countries have dramatically decreased the childhood prevalence of both iron deficiency and blood lead levels,1,2 their worldwide prevalence remains alarming.3,4 Associations between iron status and lead levels are supported by human and animal studies. Children with elevated lead levels are more likely to show evidence of iron deficiency,5,6 and children with iron deficiency are more likely to have elevated blood lead levels even after controlling for age, sex, and socioeconomic status.7 Blood lead levels are inversely related to dietary iron intake.8,9 A mechanism for these interconnections is suggested by the presence of a divalent metal transporter in the gut that is distinct from transferrin and ferritin and competitively binds to iron and lead. If iron deficiency is a risk factor for increased lead absorption, then iron treatment may result in decreased lead levels. In contrast with studies showing that increased dietary iron intake is associated with lower blood lead levels,8 some studies suggest that therapeutic iron may actually interfere with lead excretion and elevate lead levels.10-12 Such findings are especially alarming if giving iron to children in high-risk populations may increase their vulnerability to the effects of lead toxicity. However, the available studies have several methodological difficulties. The distributions of lead values among high-risk children in previous studies have been skewed by markedly elevated values. Also, iron deficiency has been defined in these studies by dichotomizing children as iron-deficient or iron-sufficient, thus obscuring the physiological spectrum of iron deficiency that progresses from iron sufficiency through depletion and deficiency phases to iron-deficiency anemia. It is plausible that the effects of iron therapy on lead levels may depend on the degree of iron deficiency. In a previous study, we used these stages of iron deficiency to evaluate the effects of iron therapy on the development and behavior of young children.13 Because blood lead
I
From the Department of Psychiatry, MetroHealth Medical Center, School of Medicine, Case Western Reserve University, Cleveland, Ohio; the Hospital Nacional de Ninos, University of Costa Rica, San Jose, Costa Rica; and the Center for Human Growth and Development, Department of Pediatrics and Communicable Diseases, University of Michigan, Ann Arbor, Michigan. Supported by NIH grants R01 HD31606-08 and R01-HD33487-06, to B. L., Principal Investigator. Submitted for publication Dec 9, 2002; revision received July 31, 2003; accepted Aug 22, 2003. Reprint requests: Abraham W. Wolf, PhD, Department of Psychiatry, MetroHealth Medical Center, 2500 Metro Health Dr, Cleveland, OH 44109. E-mail: [email protected]. Copyright ª 2003 Mosby, Inc. All rights reserved. 0022-3476/2003/$30.00 + 0 10.1067/S0022-3476(03)00540-7
789
Table. Blood lead levels and iron status measures before and after 3 months of iron therapy by iron status and treatment groups* Pretreatment iron status Anemic iron-deficient Treatment N
Intramuscular iron n = 18
Intermediate iron-deficient
3 Months oral iron n = 30
Intramuscular iron n = 12
3 Months oral iron n = 26
Pretreatment 3-Month Pretreatment 3-Month Pretreatment 3-Month Pretreatment 3-Month 9.9 (0.8) 10.2 (0.5)NS 0.3 (0.4)A,B 0.1 (0.4)A,B
Lead lg/dL Change Change adjusted for pretreatment lead Iron-sufficient{ pretreatment Iron-sufficient after 3 months Hemoglobin, g/L Change Ferritin, lg/L Change Transferrin saturation, % Change Free erythrocyte protoporphyrin, lg/dL of packed red cells Change
10.9 (0.5) 10.3 (0.5)NS 0.6 (0.5)A,B 0.6 (0.3)A,B
11.0 (0.6) 11.5 (0.5)NS 0.5 (0.6)A,B 0.5 (0.5)A,B
11.7 (1.2) 11.2 (0.8)NS 0.5 (0.6)A,B 0.2 (0.4)A,B
0%
0%
0%
0%
6%
57%
17%
77%
93.1 (2.1) 122.6 (2.5)§ 29.5 (3.2)A,B 3.2 (0.4) 9.9 (1.9)NS B,C 6.7 (1.7) 10.1 (1.3) 13.2 (0.9)NS 3.2 (1.7)B 341.9 (47.8) 101.4 (9.1)§
240.5 (41.5)C
96.5 (1.6) 132.2 (1.7)§ 35.7 (2.4)A 3.5 (0.5) 22.3 (3.5)§ A,B 18.8 (3.4) 9.1 (0.4) 25.5 (1.8)§ 16.4 (1.9)A 313.9 (30.0) 85.3 (7.2)§
228.6 (26.8)C
111.8 (1.2) 126.4 (2.6)§ 14.6 (2.3)C,D 3.8 (0.7) 9.8 (2.1)NS B,C 5.9 (2.1) 11.8 (0.8) 20.5 (2.5)z A,B 8.7 (2.2) 161.3 (13.9) 92.9 (14.6)z
68.3 (10.0)A,B
111.2 (0.9) 131.7 (1.8)§ 20.5 (2.12)C,D,E 4.0 (0.5) 25.3 (2.2)§ A,B 21.3 (2.1) 10.6 (0.6) 22.9 (2.2)§ A,B 12.3 (2.1) 156.7 (13.4) 64.5 (4.1)§
92.2 (11.6)B
*Values are means and (SEM). Cells with the same letter do not differ; Tukey test for multiple comparisons used. yp < 0.05. zp < 0.005. §p < 0.001. iTest of interaction of iron status/treatment groups by difference of pretreatment and 3-month measures. {Criterion for iron sufficiency was three measures of iron status in the normal range: serum ferritin $ 12 lg/L, transferrin saturation $ 10%, free erythrocyte protoporphyrin <100 lg/dL.
levels were obtained as part of the hematologic evaluation, the design of that study can also be used to test the following hypothesis: iron therapy will decrease lead levels in infants whose iron status is compromised.
METHODS Cohort and Sample Selection Information on blood lead levels was obtained in the course of a community-based study of iron deficiency anemia, iron therapy, and infant development. Between 1983 and 1985, families in a lower-middle-class community near San Jose, the capital of Costa Rica, were contacted in a house-tohouse survey for participation in the study. All children between 12 and 23 months old were initially screened by fingerstick for anemia (no family refused to be screened). Infants who met the following inclusion criteria were asked to be brought for a venous blood sample: birth weight 2.5 kg, singleton birth, no major congenital anomalies, no major perinatal complications, no jaundice treated with pho790
Wolf, Jimenez, and Lozoff
totherapy, no hospitalization during the preceding 6 months, no hospital admission at any age for other than an acute uncomplicated problem, no iron therapy after 6 months of age, no intramuscular iron treatment at any age, no iron-containing vitamins or formula within the last month, no known developmental delay, and no major chronic or acute illness. The selection of the final sample was based on a normal pediatric physical examination, venipuncture values for hemoglobin, iron measures, and a lead value <25 lg/dL, the Centers for Disease Control and Prevention cutoff at that time. Hemoglobin values were divided into three levels: nonanemic, >120 g/L; intermediate, between 106 and 119 g/L; and anemic, <105 g/L. Iron status was defined as follows: iron-sufficient, three measures of iron status in the normal range (serum ferritin >12 lg/L, transferrin saturation >10%, free erythrocyte protoporphyrin #100 lg/dL packed red blood cells); iron-depleted, a low serum ferritin level (#12 lg/ L) as the sole indication of iron lack; and iron-deficient, a low serum ferritin level and either a high erythrocyte protoporphyrin (>100 lg/dL packed red blood cells) or a low transferrin saturation (#10%). This grouping represents The Journal of Pediatrics December 2003
Pretreatment iron status Nonanemic iron-deficient
Nonanemic iron-depleted
Iron-sufficient
3 Months oral iron n = 16
3 Months oral iron n = 34
3 Months oral placebo n = 29
Pretreatment
3-Month
Pretreatment
11.4 (0.7)NS
11.9 (0.9)
3-Month
11.6 (0.5) 10.2 (0.3)§ B 1.4 (0.4)
A,B
0.5 (0.4)
0.1 (0.5)A,B
Pretreatment
3-Month
Analysis of variance
10.9 (0.5)y
9.9 (0.4) A
F(6/158) = 3.5zi
1.2 (0.3)B 0%
0.5 (0.3)A 100%
F(6/157) = 2.9z
91%
62%
1.0 (0.3)
0% 88% 136.4 (2.4)§
124.2 (1.3) 12.3 (2.4) 4.3 (0.8)
9.1 (1.3) 32.4 (6.2)
§
28.2 (5.7)A 12.1 (0.9)
E
6.5 (0.4)
41.9 (5.0)
§
35.3 (4.9)A 24.1 (1.9)
§
12.1 (2.0)A,B 123.3 (10.0)
136.1 (1.5)§
127.1 (0.8)
D,E
65.1 (9.1)
y
58.2 (7.7)A,B
§
20.6 (1.1) 31.6 (2.5) 11.07 (2.6)A,B 65.2 (4.3) 42.8 (3.2)NS
22.4 (3.3)A,B
a continuum from least to most iron-deficient and approximates the physiologic progression involved in developing iron-deficiency anemia. Because nonanemic iron-depleted infants were disproportionately common in the original sample, a random selection procedure was instituted so that only one in five such infants was included in the final sample. The proportion of infants in the study sample in each iron status category was as follows: iron-deficient anemic, 17%; intermediate iron-deficient, 21%; nonanemic iron-deficient, 10%; nonanemic iron-depleted, 23%; and nonanemic ironsufficient, (29%). A total of 191 infants met criteria for the iron study and agreed to participate. Two infants had blood lead levels >25 lg/dL and thus did not qualify. For this analysis, however, their data were included. Pretreatment lead levels were available for 184 infants; 165 had lead data both pretreatment and after 3 months. There was no difference between the lead levels for these 165 infants who remained in the study and the 19 whose lead levels were not available at 3 months. Effects of Iron Therapy on Infant Blood Lead Levels
129.4 (1.2) 132.5 (1.3)NS D 3.1 (1.3) 22.9 (1.9) 14.9 (1.0)y C 8.0 (2.2) 21.4 (1.2) 24.5 (1.2)NS B 3.1 (1.4) 61.4 (3.8) 57.4 (3.1)NS
4.0 (3.0)A
F(6/158) = 34.5§ F(6/158) = 17.7§ F(6/158) = 5.6§
F(6/158) = 27.9§
The average age of the 165 infants was 17.3 months (0.3 SEM), ranging from 13 to 24 months; 55.8% (92/165) were male. The weight at birth was 3.37 kg (0.08 SEM), and the mother’s age at delivery was 25.6 years (0.4 SEM). Mothers had an average of 8.8 years of schooling (0.2 SEM) and breastfed their infants an average of 29.1 weeks (1.9 SEM). The infants averaged between the 40th and 50th percentile for weight and length using US standards. In a previous study, we reported that the infant blood lead levels were not related to developmental outcome measures at infancy or in 5-year follow-up.14 The treatment component was a double-blind design with short-term and long-term aspects. In the short-term (1-week) treatment aspect, infants in the anemic and intermediate groups (hemoglobin <120 g/L) were randomly assigned intramuscular iron, oral iron, or oral placebo treatment, and nonanemic infants were randomly assigned oral iron or placebo for 1 week. All parenterally treated infants were given oral placebo drops the first week; all infants who did not receive parenteral iron received a placebo injection. After the 791
first week, the long-term treatment design was applied. Parenterally treated infants and iron-sufficient infants were given placebo drops for 12 weeks. All other infants were treated with oral iron. The doses of intramuscular iron were calculated to correct anemia and increase hemoglobin levels to 125 g/L. To ensure intake of oral medication, project personnel workers visited families daily to administer oral drops. Infants who received an additional week of oral iron in the short-term component did not differ from others treated with oral iron. Therefore, their data were combined with data from the other infants. To assess the effects of intramuscular and oral iron treatment on lead levels at different stages of iron deficiency, seven iron status and treatment groups were formed: (1) anemic iron-deficient infants treated with intramuscular iron, (2) anemic iron-deficient infants treated with oral iron, (3) intermediate iron-deficient infants treated with intramuscular iron, (4) intermediate iron-deficient infants treated with oral iron, (5) nonanemic iron-deficient infants treated with oral iron, (6) nonanemic iron-depleted infants treated with oral iron, and (7) iron-sufficient infants, who received oral placebo.
Measurement of Blood Lead Red blood cells were promptly separated and frozen for future analysis in the United States. The frozen red cells were analyzed in a laboratory at MetroHealth Medical Center that participated in the Centers for Disease Control Maternal and Child Health Resources Development Proficiency Testing Program for blood lead. Lead levels were determined by a graphite furnace method15,16 using a matrix modification procedure (0.5% ammonium phosphate, 0.5% Triton X-100 detergent, Amersham Biosciences Corp, Piscataway, NJ). Samples were diluted by using a Sylva diluter and analyzed by using an autosampler and pyrocoated graphite furnace (Perkin Elmer 560 AA, 2200 furnace, AS-40 Autosampler, Perkin Elmer, Wellesley, Mass). Quality control of the lead analyses was monitored through certified controls obtained from the National Bureau of Standards. Red cell lead values were converted to whole blood lead levels by using the formula of Rosen et al.17
Data Analysis Statistical analyses followed the general linear model for continuous variables. Changes in lead levels and iron status measures within and among iron status and treatment groups were compared by using a mixed model: pretreatment and 3month lead and iron measures were specified as random effects and iron status and treatment groups as fixed effects. Differences between pretreatment and 3-month lead and iron measures were assessed by using paired t tests with the error terms from the mixed model. Multiple comparisons among iron status and treatment groups were assessed by the Tukey test. Changes in lead levels were also assessed by using analysis of covariance to control for the effects of pretreatment levels. Categorical variables were compared by using the v2 test. All analyses were performed with standard computational 792
Wolf, Jimenez, and Lozoff
packages.18 A two-tailed a level of 0.05 was defined for tests of statistical significance.
RESULTS The mean blood lead level for the 165 infants before iron treatment was 10.98 lg/dL (0.26 SEM), ranging from 5.8 to 36.9 lg/dL; 57.6% (95/165) had lead levels >10 lg/dL. Pretreatment blood lead levels for the seven iron status and treatment groups did not differ (F = 1.4, df = 6/158, NS). The Table lists the values for blood lead levels and iron status measures at pretreatment and 3 months for the seven iron status and treatment groups. Two groups had significant intragroup changes in lead levels at 3 months. The nonanemic iron-depleted infants (infants whose sole indicator of iron deficiency was a low serum ferritin and who received 3 months of oral iron) decreased lead levels by 1.4 lg/dL, from 11.6 lg/dL to 10.2 lg/dL (P = .0004); after adjusting for pretreatment lead levels, the difference of 1.2 lg/dL remained statistically significant (P = .0002). Infants iron-sufficient at pretreatment, who received 3 months of oral placebo, increased lead levels by 1.0 lg/dL, from 9.9 lg/dL to 10.9 lg/dL (P = .025); after adjusting for pretreatment levels, the value of 0.5 lg/dL was not significant (P = .12). The intergroup comparisons of change in lead levels indicated that nonanemic iron-depleted infants had the greatest decrease in lead levels, whereas ironsufficient infants who received placebo showed the greatest increase in blood lead concentration. These intergroup differences remained significant after adjusting for pretreatment levels and after adjusting for the effects of age and sex. To help understand how iron therapy affected lead levels, we first considered how treatment affected iron status. After 3 months, 93% of all infants treated with intramuscular or oral iron showed an increase in hemoglobin of at least 10 g/ L. The anemic and intermediate iron-deficient infants had the greatest increase in hemoglobin. Iron therapy corrected the anemia in all children but not all three biochemical indicators of iron deficiency. Ten percent (3/30) of children who received intramuscular iron and 77% (82/106) of children who received oral iron became iron-sufficient. In contrast, 62% (18/29) of children who were iron-sufficient at the start of the study developed signs of iron lack after 3 months of oral placebo. We also examined changes in each iron status measure: ferritin, transferrin saturation, and erythrocyte protoporphyrin. These measures improved in all groups receiving iron treatment, but improvement varied considerably by type of treatment and level of iron deficiency. Among anemic and intermediate irondeficient infants, each iron status measure improved more in infants treated with 3 months of oral iron than in infants treated with intramuscular iron. The groups of iron-deficient and iron-depleted nonanemic infants showed even greater levels of improvement. Only one group deteriorated on any measure of iron status: children who started the study as ironsufficient had significantly lower levels of ferritin after 3 months of placebo. The Journal of Pediatrics December 2003
Figure. Mean change from pretreatment to 3-month evaluation of infant blood lead and ferritin levels plotted by iron status and treatment groups. Hatched areas represent 1 SEM of the changes in ferritin and lead levels.
Of the individual measures of iron status, changes in ferritin were closely associated with changes in iron status. The groups with the greatest proportions becoming iron-sufficient showed the most decreases in lead levels. After 3 months of oral iron, nonanemic infants who started the study with a low serum ferritin as the sole abnormal measure of iron status showed the most marked decrease in lead levels, followed by nonanemic infants with iron deficiency (two or more abnormal iron measures at study outset) and iron-deficient infants with hemoglobin levels <120 g/L. The two groups showing an increase in lead levels were ones in which either fewer infants became iron-sufficient or infants showed evidence of developing iron deficiency: iron-deficient infants with hemoglobin levels <120 g/L who received intramuscular iron and iron-sufficient nonanemic infants who were given placebo for 3 months. Changes in ferritin levels were most closely associated with changes in lead levels. The Figure shows the mean changes in blood lead and ferritin levels plotted by iron status and treatment groups. Nonanemic iron-depleted children showed a decrease in lead levels of 1.4 lg/dL and an increase in ferritin of 35.3 lg/L, the largest ferritin increase of any Effects of Iron Therapy on Infant Blood Lead Levels
compromised group. Children iron-sufficient at the start of the study, who received placebo, showed the greatest increase in lead levels, 1.0 lg/dL, and a decrease in ferritin of 8.0 lg/L. This was the only group in which iron status worsened.
DISCUSSION These findings support the hypothesis that iron therapy can decrease lead levels in children whose iron status is compromised. Even over the short period of 3 months, changes in blood lead levels were closely associated with changes in iron status. However, we had not expected that lead levels would increase in this short time in children who were iron-sufficient at the start of study. It was also unexpected that the effect of iron therapy on lead levels was most marked at the earliest stage of iron deficiency. Lead levels decreased to the greatest degree in nonanemic iron-depleted children, that is, children for whom a low ferritin level was the only indication of iron lack. These results indicate that preventing or correcting iron deficiency was associated with lowering of children’s blood lead levels. 793
Our findings appear plausible in terms of the absorption mechanisms for lead and iron. The fact that iron status worsened in the iron-sufficient group treated with placebo indicates the precarious nature of iron nutrition in this population. Because they were on placebo and the local diet was clearly limited in iron, their iron stores reduced over the 3 months, reflected by decreasing ferritin levels. These decreasing iron stores would signal for increased intestinal absorption of divalent metals, and lead would be absorbed along with what little iron was available. The situation for the intramuscular-treated infants was related, in an unanticipated way. The dose of intramuscular iron had been calculated to bring hemoglobin levels to 125 g/L. However, the final hemoglobin level for infants given oral iron was 135 g/L, and a much higher proportion became iron-sufficient, indicating that the intramuscular dose was insufficient. Intramuscular iron, although improving iron status, did not lower blood lead levels, because it did not compete for absorption in the gastrointestinal tract. Because intramuscular iron-treated infants remained iron-deficient throughout the study, intestinal absorption of both iron and lead would also have remained active. Further mechanisms must be invoked to explain differing changes in lead levels among the groups of infants who received oral iron therapy. Although there is a shared transporter for iron and lead, these infants all received comparable doses of oral iron on a per kilogram basis. Thus, differential competition for the divalent metal transporter does not seem to explain between-group differences. We postulate that progressive shutting down of intestinal absorption of iron—and lead—as subjects became more iron-sufficient contributed to the differences. Based on the physiology of iron absorption, one would expect that groups becoming ironsufficient most rapidly would show the fastest shut-down of intestinal absorption, and correspondingly, the most pronounced drop in blood lead levels. Our results are consistent with this expectation. Nonanemic iron-depleted infants, who had the least degree of iron deficiency, probably became ironsufficient earliest, as evidenced by the highest ferritin levels and highest proportion iron-sufficient in this group. Correspondingly, they had the greatest reduction in lead levels. Oral iron-treated infants with greater degrees of iron deficiency also increased iron stores and reduced lead levels, but not as much as the iron-depleted infants. Our results contrast with some previous studies suggesting iron therapy increased lead levels, either by facilitating lead absorption10 or by slowing lead excretion.11 We used a community-derived sample with groups that approximated the physiological progression from iron sufficiency to anemia. Previous study groups had higher pretreatment lead levels—means of 31.1 lg/dL and 28.3 lg/dL, respectively—compared with 10.98 lg/dL in our sample. Furthermore, definitions of iron deficiency in those studies were based solely on a ferritin level, with cut-offs of 8.0 lg/dL and 16.0 lg/dL, respectively, and the number of ‘‘irondeficient’’ children was very small (23 and 20, respectively). Dichotomizing iron status by ferritin alone obscures the 794
Wolf, Jimenez, and Lozoff
process of developing iron deficiency and may fail to detect important differences along that dynamic spectrum. As Pollitt19 states, the several stages of iron-depletion lead to incremental involvement of different systems in the body that activate distinct mechanisms within and among domains. Yet even the identification of iron status along a physiological spectrum will not fully capture lead–iron dynamics. The association between lead levels and iron status was appreciated only with pretreatment and posttreatment measurement. Thus, to understand lead–iron interactions fully, it appears to be essential to examine changes in iron status in addition to the degree of initial iron deficiency by using several hematologic parameters. Our sample was drawn from a study originally designed to evaluate the effects of iron therapy on the cognitive and behavioral development of healthy infants. Therefore, results cannot be generalized to children who are not in good health or who show generalized undernutrition. Further studies are required to assess the effects of iron treatment in children with more severe lead toxicity. Conversely, it is not clear what the effects of iron therapy would be on lead absorption in children with even lower lead levels. Furthermore, the study was conducted in a setting in which compromised iron status in infancy was close to universal. Results may be different if infant iron status is generally good. Mahaffey20 has stressed that lead toxicity is caused by exposure to lead and not by nutritional deficiency. However, the fractional transfer of lead from the environment to target tissue can be modified by nutritional status. Although public health programs to reduce lead toxicity need to include nutritional interventions, such interventions are clearly secondary to the elimination of environmental sources of lead. Our results indicate that correcting iron deficiency may be a rapid and effective means of improving infant lead levels further, even in nonanemic infants. As countries succeed in lowering lead levels by environmental means, improving iron status may help achieve even further reductions.
REFERENCES 1. Halterman JS, Kaczorowski JM, Aligne CA, Auinger P, Szilagyi PG. Iron deficiency and cognitive achievement among school-aged children and adolescents in the United States. Pediatrics 2001;107:1381-6. 2. CDC. Blood lead levels in young children—US and selected states, 1996-1999. MMWR Morb Mortal Wkly Rep 2000;49:1133-7. 3. Stoltzfus RJ. Iron-deficiency anemia: reexamining the nature and magnitude of the public health problem. Summary: implications for research and programs. J Nutr 2001;131:697S-700S. 4. Wasserman GA, Liu X, Popovac D, Factor-Litvak P, Kline J, Waternaux C, et al. The Yugoslavia Prospective Lead Study: contributions of prenatal and postnatal lead exposure to early intelligence. Neurotoxicol Teratol 2000;22:811-8. 5. Yip R, Norris TN, Anderson AS. Iron status of children with elevated blood lead concentrations. J Pediatr 1981;98:922-5. 6. Clark M, Royal J, Seeler R. Interaction of iron deficiency and lead and the hematologic findings in children with severe lead poisoning. Pediatrics 1988;81:247-54. 7. Wright RO, Shannon MW, Wright RJ, Hu H. Association between iron deficiency and low-level lead poisoning in an urban primary care clinic. Am J Public Health 1999;89:1049-53.
The Journal of Pediatrics December 2003
8. Hammad TA, Sexton M, Langenberg P. Relationship between blood lead and dietary iron intake in preschool children: a cross-sectional study. Ann Epidemiol 1996;6:30-3. 9. Lanphear B, Hornung R, Ho M, Howard C, Eberly S, Knauf K. Environmental lead exposure during early childhood. J Pediatr 2002;140:40-7. 10. Angle CR, Stelmak KL, McIntire MS. Lead and iron deficiencies. In: DD Hemphill, editor. Trace substances in environmental health. Vol IX. Columbia (MO): University of Missouri; 1975. p. 367-86. 11. Ruff HA, Markowitz ME, Bijur PE, Rosen JF. Relationships among blood lead levels, iron deficiency, and cognitive development in two-year-old children. Environ Health Perspect 1996;104:180-5. 12. Wright RO, Hu H, Maher TJ, Amarasiriwardena C, Chaiyakul P, Woolf AD, et al. Effect of iron deficiency anemia on lead distribution after intravenous dosing in rats. Toxicol Ind Health 1998;14:547-51. 13. Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez E, et al. Iron deficiency anemia and iron therapy effects on infant developmental test performance. Pediatrics 1987;79:981-95.
14. Wolf AW, Jimenez E, Lozoff B. No evidence of developmental III effects of low-level lead exposure in a developing country. J Dev Behav Pediatr 1994;15:224-31. 15. Fernandez FJ. Micromethod of lead determination in whole blood by atomic absorption with use of graphic furnace. Clin Chem 1975;21: 558-61. 16. Fernandez FJ, Hillegos D. An improved graphite furnace method for the determination of lead in blood using matrix modification and the L’vov platform. Atom Spectrosc 1982;3:1300-11. 17. Rosen FJ, Zarate-Salvador C, Trinadad EE. Plasma lead levels in normal and lead-intoxicated children. Pediatrics 1974;84:45-8. 18. SAS. SAS System, release 8.02, Microsoft Windows TS02M0. Cary (NC): SAS Institute; 2002. 19. Pollitt E. The developmental and probabilistic nature of the functional consequences of iron-deficiency anemia in children. J Nutr 2001;131:669S-75S. 20. Mahaffey KR. Nutrition and lead: strategies for public health. Environ Health Perspect 1995;103(suppl 6):191-6.
50 Years Ago in The Journal of Pediatrics ACUTE RESPIRATORY DISEASE WITH SPECIAL REFERENCE TO PATHOGENESIS, CLASSIFICATION, AND DIAGNOSIS
Manson J. J Pediatr 1953;43:599-619 In this comprehensive review from 50 years ago, Manson proposes a classification system for acute respiratory illnesses in children based on the anatomic distribution of infections. Hence, he reviews in order viral and bacterial infections involving the larynx, tracheobronchial tree, lungs, bronchii, and bronchioles, emphasizing the not uncommon finding still recognized today of infections involving the upper airway progressing distally to affect the lower respiratory tract. Quite accurately, he describes the risks and consequences of secondary bacterial infections after primary viral respiratory tract disease. He also believes that a viral etiology exists for bronchiolitis but acknowledges the likelihood of ‘‘several causative viral agents.’’ Finally, he laments the lack of information regarding ‘‘viral behavior,’’ and by extension, the difficulties in isolating causative viruses in respiratory tract disease. Manson’s paper reflects a bias doubtless prevalent 50 years ago: that children are merely ‘‘little adults’’ and as such follow a similar clinical course of respiratory infections as adults with susceptibility to microbiologic pathogens identical to those found in adults. In fact, he relies heavily on several sources from the adult literature to describe causative agents of acute respiratory infections in children including a 1944 study commissioned by the Army and three studies by Reimann in adults from the 1940s. Current pediatric pulmonary textbooks1,2 reflect subsequent studies describing age-specific differences in susceptibility patterns to different viral and bacterial pathogens an the subsequent clinical course. An important example of this is disease caused by the respiratory syncytial virus, capable of causing a severe, protracted illness in infants and small children but causing a mild coryzal syndrome in immunocompetent adults. Manson’s discussion of clinical and physical findings in children with respiratory illnesses is instructive. He emphasizes the importance of the physical examination and its daily, changing pattern as essential to diagnostic accuracy. The use of other ‘‘tools,’’ such as radiographs and laboratory studies, are merely suggested as afterthoughts and by implication, of use only as a last resort. His elegant, paragraph-long description of the physical findings in bronchiolitis is particularly evocative and is as accurate today as it was 50 years ago. In our current reliance on sophisticated microbiologic and radiographic techniques, perhaps we’ve forgotten, to some extent, the abundant clues provided by a careful physical examination. Blakeslee E. Noyes, MD, Division of Pulmonary Medicine Cardinal Glennon Children’s Hospital, St Louis University School of Medicine St Louis, MO 63104
YMPD506 10.1067/S0022-3476(03)00547-X
REFERENCES 1. 2.
Kending EL. Kendig’s disorders of the respiratory tract in children. Chernick V, Boat T, eds. Philadelphia (PA): WB Saunders Co; 1998. Pediatric respiratory medicine. Taussig LM, Landau LI, eds. St Louis (MO): Mosby, Inc; 1999.
Effects of Iron Therapy on Infant Blood Lead Levels
795