Changing Trends in the Epidemiology of Pediatric Lead Exposure: Interrelationship of Blood Lead and ZPP Concentrations and a Comparison to the US Population

Summary

Objectives

To determine blood lead and zinc protoporphyrin (ZPP) concentrations in a pediatric population, confirm their interrelationship at low blood lead concentrations, and assess changing trends through comparison of these data with those found in a similar population 10 years earlier and to US national values.

Study Design and Methods

The study was conducted in a large pediatric hospital in the Washington DC area (CNMC) on patient whole blood specimens (n = 4908) (0–17 years) accrued from January 2001 to June 2002. Pediatric blood lead concentrations were determined by atomic absorption spectrophotometry, and ZPP by hematofluorometry. The data were analyzed using a computer adaptation of the Hoffmann approach.

Results and Conclusions

Blood lead level (BLL) means ranged between 2.2 and 3.3 μg/dL, and the median BLL was 3 μg/dL throughout. Mean ZPP concentrations ranged between 21.1 and 26.6 μg/dL and the median concentrations between 21 and 27 μg/dL. In comparison to data obtained from a similar pediatric population at CNMC between 1991 and 1992, pediatric BLLs have significantly declined in the Washington DC area. The current data are also compared with data obtained from the National Health and Nutrition Examination Survey (NHANES III) of the US population. The interrelationship between ZPP and BLLs is examined.

Because of its useful properties and widespread use in industry, gasoline, and household paint during the 1900s, lead is ubiquitous in US urban environments. The US banned leaded paint in 1978 and fully phased out leaded gasoline in 1986, but despite these efforts, many authorities still consider lead poisoning the primary environmental hazard to American children. The principal sources of lead exposure today include old house paint, contaminated soil, lead-contaminated water (because of lead pipes and solder), and food (from the use of pottery glazes and lead containers).¹

Exposure to lead can result in significant adverse health effects to the nervous, hematologic, renal, and reproductive systems. Lead exposure in young children may have significant adverse neurodevelopmental effects. Blood lead level (BLL) is the biologic index most often used by health care providers as an indicator of recent lead exposure. Currently, an elevated blood lead level is defined by the Centers for Disease Control and Prevention (CDC) as ≥10 μg/dL (>483 nmol/L) for boys and girls of all ages, including children younger than 6 years of age.¹

In addition to BLL, other lead exposure indices include free erythrocyte protoporphyrin (FEP) and ZPP; both are precursors of heme, whose levels elevate on moderate to high exposure to lead. Protoporphyrin is the immediate precursor of the heme molecule, requiring iron for the final step of heme synthesis. Because lead inhibits three enzymes in the heme biosynthesis pathway, erythrocyte protoporphyrin accumulates, and increased ZPP and FEP concentrations may be observed.² However, neither FEP nor ZPP is sensitive enough or specific enough to be used as the primary indicator at low lead exposure. Increased blood ZPP and FEP concentrations can occur with iron deficiency, lead poisoning, many erythrocyte disorders, any autosomal recessive porphyria, and sometimes with autosomal dominant acute porphyrias.

ZPP and FEP are not equivalent; the minor nonheme porphyrins in healthy erythrocytes consist of ZPP (approximately 95%), and FEP (which includes ZPP) is approximately 5% greater. In certain diseases these ratios can be significantly altered.^2,3 ZPP determinations are clinically useful in classifying microcytic red blood cells caused by disorders of heme synthesis or by disorders of globin synthesis.⁴ In the assessment of iron status, the ratio of ZPP or FEP to heme is calculated because it correlates well with values of plasma ferritin, plasma iron, transferrin saturation, hemoglobin, and hematocrit as the most sensitive indices of iron status.^5–7 ZPP can be the differential test for the three most common porphyrias and erythropoietic protoporphyria, β-thalassemia minor,⁸ and advanced chronic hepatitis C infection.⁹ ZPP levels can be used in the clinical laboratory as a preliminary, usually out-of-hours, screening test to evaluate the need for a stat blood lead measurement and the possibility of lead poisoning in children.^10,11

Using a large pediatric hospital population (CNMC), we defined specific ranges for pediatric blood lead concentrations¹² and reference ranges for ZPP concentrations.¹³ In this study we investigated the relationship between low blood lead concentrations and ZPP in pediatric patients (0–17 years) and compare these data to a similar pediatric population surveyed 10 years earlier at CNMC (1991–1992)¹⁴ and to the US NHANES III data.

MATERIALS AND METHODS

Patients and Sample Collection

Lead and ZPP levels were determined during routine clinical chemistry testing of whole blood samples. The samples were obtained between January 1, 2001 and June 8, 2002 from a population of outpatient and hospitalized children. Samples were identified only by age and gender. The specimens used for testing were kept refrigerated for no longer than 96 h at 2° to 4°C. No effort was made to exclude subjects from the study based on prior chronic or acute lead exposure or conditions affecting ZPP levels because the analytic method for calculating the reference ranges eliminated abnormal or outlier values resulting from such conditions.

Lead Assay

Blood lead concentrations were determined in 100-μL samples of whole blood containing ethylene-diamine-tetraacetic acid (EDTA) with an electrothermal atomization atomic absorption spectrophotometric assay (AAnalyst 600, Perkin Elmer, Wilton, CT). The assay is based on the rapid Zeeman graphite furnace atomic absorption method and has a high degree of accuracy.^9,10 Our laboratory routinely participates in lead proficiency testing. Three levels of control from Kaulson Laboratory were run before each batch of patient samples. In addition, a QC was run after every 10 samples. The analytic range for this assay extends from the minimum detectable concentration of 1 μg/dL to the concentration value of the highest standard of 39 μg/dL. In the case of subjects with a blood lead concentration greater than 39 μg/dL, the sample was diluted with diluent (zero calibrator obtained from the National Institute of Standards and Technology, Gaithersburg, MD) before analysis. Blood lead concentrations were reported in micrograms per deciliter (for lead multiply by 48.3 to convert to nanomoles per liter).

ZPP Assay

ZPP concentrations were measured in 50-μL anticoagulated whole blood samples containing either heparin or EDTA using a zinc protoporphyrin hematofluorometer (ZP Hematofluorometer Model 206, AVIV Biomedical Inc., Lakewood, NJ). After calibration of the instrument, a drop of blood is placed on the slide. The excitation wavelength is 415 nm, and the measurement wavelength is 596 nm. The average of three or more hematofluorometer readings is reported for each specimen, and readings over 50 μg/dL are corrected for the patient’s hematocrit. Appropriate control samples at the low, medium, and high levels were run after every 10 samples. The ZPP results are reported in micrograms per deciliter and can also be converted to micromoles per mole heme by multiplying by a factor of 1.84. This latter conversion makes the assumption that the hemoglobin concentration is 14 g/dL and hence may introduce an error into the result reported. Nevertheless, it is the recommended NCCLS unit.

Statistical Analysis

BLLs (μg/dL), ZPP concentrations (μg/dL), child age, and gender were entered into a spreadsheet program. The data sets were separated by gender and stratified by age. Abnormal and outlier values were truncated from each individual age category according to the Hoffmann method.¹⁵ Generally, the top and bottom 10 to 20% of the data were discarded, and a line was drawn to the central linear portion of the graph. The remaining data, which was of normal Gaussian distribution, was used to calculate the percentiles for each of the age groups. Percent cumulative frequency versus lead concentrations were plotted (percent cumulative frequency) so that the 10th and 90th percentiles could be calculated. These were used as the final reported blood lead concentrations. Geometric mean BLLs for different groups were compared using the unpaired Student t test with P < 0.05 as statistically significant. The χ² test was used for calculating P value for the difference in the fraction of children falling within the lowest BLL interval between the NHANES III and CNMC samples. This P value remained highly significant under other multiple-comparison adjustments (such as Bonferroni).

The age groups were selected to conform to data published in Pediatric Reference Ranges 2003, 4th edition, AACC Press. The relationship between lead and ZPP was fit by a LOWESS (LOcally WEighted Scatterplot Smoother) algorithm. Statistical analyses were performed with SPSS, version 11 (SPSS Inc, Chicago, IL) and R, version 1.6.1 (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Blood lead and ZPP concentrations were determined separately for girls and for boys and stratified by age groups (Table 1). Generally, BLLs were higher for boys in each of the age groups. In both boys and girls, mean lead values starting after age 13–24 months declined with age. The highest values were observed in girls aged 13–24 months, probably because of hand-to-mouth exploratory behavior normal for this age group. The medians for lead in female and male subjects were similar. The geometric means for ZPP were generally higher in girls, and the median values declined with age for both genders. The relationship between BLL and ZPP of pediatric patients (0–17 years) of Children’s National Medical Center (CNMC) (January 2001 to June 2002) is shown in Table 1 and Figure 1.

TABLE 1.

Lead and ZPP concentrations (μg/dL): CNMC January 2001 to June 2002

Age group	Sample size	Geometric mean (95% confidence interval)		10%tile		25%tile		50%tile		75%tile		90%tile
		Lead	ZPP	Pb	ZPP	Pb	ZPP	Pb	ZPP	Pb	ZPP	Pb	ZPP
Females (n = 1,812)
0–11 months	43	3.1 (2.9–3.3)	26.6 (23.9–29.7)	3	17	3	21	3	27	3	34	4	44.4
1 year	250	3.3 (3.1–3.4)	26.1 (24.9–27.4)	3	17	3	20.8	3	25	3.3	32	5	41
2–4 years	1260	3.1 (3.0–3.2)	24.1 (23.6–24.7)	1	16	2	19	3	23	4	29	7	37
5–9 years	222	2.8 (2.6–3.0)	23 (22.1–24)	1	15.3	2	18	3	23	4	28	5	34
10–17 years	37	2.2 (1.9–2.5)	25.4 (21.3–30.4)	1	14.6	2	19	3	22	3	31	3	60.8
Males (n = 3,096)
0–11 months	142	3.2 (3.1–3.4)	25.8 (24.5–27.2)	3	18	3	21	3	25	3	30	4	37
1 year	586	3.2 (3.1–3.3)	25.7 (24.9–26.6)	2	16	3	20	3	25	4	32	5.3	41.3
2–4 years	1820	3.2 (3.1–3.3)	24.4 (23.9–24.8)	1	16	2	19	3	24	4	29	7	38
5–9 years	491	3.1 (3.0–3.3)	21.1 (20.6–21.7)	1	15	2	17	3	21	4	26	7	31
10–17 years	57	2.6 (2.2–3.1)	21.7 (19.4–24.2)	1	13	2	17	3	21	3	26	5	40.4

Conversion: blood lead concentrations (μg/dL), multiply by 48.3 to convert to nmol/L; ZPP (μg/dL), multiple by 1.84 to convert to μmol/mol heme.

FIGURE 1.

The relationship between BLL and ZPP (μg/dL) of pediatric patients (0 to 17 years) of Children’s National Medical Center (CNMC) (January 2001 to June 2002). Box is 25th to 75th centile; bar within box is median.

We were interested in comparing low BLLs to the ZPP levels of the same patients, detected from the same blood sample. As illustrated in Figure 1, up to BLL 20μg/dL there is no increase in ZPP concentrations. However, as the BLLs increase, ZPP levels increased concomitantly.

A LOWESS curve was fitted to the data in an attempt to find the ZPP values for BLLs (Fig. 2). A LOWESS curve does not have a “formula” in the usual sense; it is just a smooth curve that passes reasonably close to the points. The values for the curve are listed in Table 2.

FIGURE 2.

The relationship between BLL and ZPP concentrations (μg/dL) of pediatric patients (0 to 17 years) at Children’s National Medical Center (January 2001 to June 2002). The graph shows the CNMC data with a “LOWESS” fitted curve. A LOWESS curve does not have a “formula” in the usual sense; it is just a smooth curve that passes reasonably close to the points. The values for the curve are listed in Table 2.

TABLE 2.

The values for the “LOWESS curve” in Figure 2: the relationship between BLL and ZPP (μg/dL) (0–17 years)

Lead (μg/dL)	ZPP (μg/dL)	Lead (μg/dL)	ZPP (μg/dL)	Lead (μg/dL)	ZPP (μg/dL)
1	25	15	24	29	71
2	24	16	26	30	75
3	23	17	28	31	80
4	23	18	31	32	85
5	23	19	34	33	90
6	22	20	37	34	94
7	23	21	40	38	113
8	23	22	43	39	118
9	22	23	47	43	136
10	23	24	50	45	146
11	23	25	54	47	155
12	24	26	58	49	165
13	23	27	62	103	432
14	24	28	66

We compared our results with a population of children at CNMC tested in 1991–1992. The children were 9 months to 3 years old and inner city subjects.¹⁴

The distribution of the number of subjects across blood lead groups changes from 1991–1992 to 2001–2002, with a very pronounced tendency for a greater proportion of the subjects to be in the lowest group in 2002, <10 μg/dL, compared with 1992 (Tables 3 and 4). This change is highly significant (χ² = 521.6, df = 7, P < 0.0001).

TABLE 3.

Changes in CNMC data from 1993 to 2002

Blood lead groups (μg/dL)	Number of subjects		Blood lead concentration (μg/dL)		ZPP concentration (μg/dL)
	1993 (n = 4,217)	2002 (n = 4,906)	1993	2002	1993	2002
<10	3,437	4,715	4.7 ± 1.9	3.3 ± 1.6	25.7 ± 9.3	26.0 ± 13.5
10–14	504	107	11.5 ± 1.4	11.5 ± 1.3	23.9 ± 13.5	26.4 ± 14.4
15–19	144	32	16.6 ± 1.4	16.8 ± 1.6	29.9 ± 19.2	40.1 ± 32.3
20–24	61	20	21.7 ± 1.4	22.0 ± 1.2	31.8 ± 19.6	37.9 ± 24.4
25–29	25	19	27.0 ± 1.4	26.6 ± 1.2	56.8 ± 35.9	42.0 ± 41.0
30–34	13	6	31.6 ± 1.3	32.2 ± 1.5	64.5 ± 11.1	94.5 ± 45.5
35–39	12	2	36.9 ± 1.3	38.5 ± 0.7	91.0 ± 133.4	162.5 ± 34.6
≥40	21	5	58.8 ± 16.2	57.4 ± 25.6	100.0 ± 134.9	169.8 ± 137.4

TABLE 4.

Distribution of blood lead groups in 1993 and 2002

Blood lead groups (μg/dL)	Number (%) of subjects
	1993	2002
<10	3437 (81.5)	4715 (96.2)*
10–14	504 (11.9)	107 (2.1)
15–39	255 (6.1)	79 (1.6)
≥40	21 (0.5)	5 (0.1)
Total	4,217	4,906

A greater proportion of the subjects are in the lowest group in 2002, compared with 1993.

^*This change is highly significant (chi square = 521.6, df = 7, P < 0.0001).

DISCUSSION

Epidemiologic research has been consistent at showing relationships between lead exposure and lowered IQ, cognitive problems, behavioral impairments, and emotional status.¹ Lead exposure adversely affects the neurocognitive development and behavior of young children. Lead toxicity is not only a function of the developmental vulnerability of the child; it is also a function of the exposure dose, the timing, duration, the route of exposure, and the nutritional status of the child. The current CDC reference cutoff for children aged less than 6 years is <10 μg/dL.¹ Children are particularly vulnerable to lead exposure because they absorb lead more readily. In addition, their central nervous system is more susceptible to lead insults because it is still developing and serves as a “sink” for absorbed lead. In particular, children living in urban areas and in older homes are at a greater exposure risk because of elevated environmental lead levels.¹⁴ Growing evidence indicates that no detectable threshold exists for the adverse effects of lead exposure on neurodevelopment.^16,17 Lead monitoring is especially important in areas where children may be at a higher risk for lead exposure, such as those of lower socioeconomic status, minority groups, and those dwelling in houses painted before the lead paint ban (1987).

This study compares the interrelationship between blood lead and ZPP concentrations in 4908 whole blood samples collected in the Washington, DC area at Children’s National Medical Center. In addition, this study compares the recent 2002 findings to those of an earlier study of the pediatric population at CNMC by Rifai et al in 1993¹⁴ in an attempt to gauge any changes that have occurred in BLLs and ZPP concentrations over the past 10 years.

The CNMC 2002 data disclose similar blood lead and ZPP levels between boys and girls in age groups 0–11 months, 1 year, and 2–4 years. In age groups 5–9 years and 10–17 years, boys have higher blood lead levels and lower ZPP levels than girls; this difference grows more pronounced at progressively higher percentiles. From Table 4 we can see that 3.8% of the children in 2001–2002 (CNMC) have mean BLLs ≥10 μg/dL compared with 18.5% of the children tested in the 1991–1992 CNMC study.¹⁴ This is a very significant improvement. It is clear from these data that efforts to decrease environmental lead exposure are responsible for this improvement.

CDC’s National Health and Nutrition Examination surveys (NHANES) are an ongoing series of national examinations of the health and nutritional status of the civilian, noninstitutionalized population.^18–20 NHANES III (1988–1994) and NHANES 1999 (results published in 2002)²¹ indicate that BLLs in the US population continued to decrease. Gender and age were two principal risk factors found for excessive BLLs in children and among children aged 1 to 5 years. BLLs were more likely to be elevated among those who were poor, non-Hispanic black, living in large metropolitan areas, or living in older housing. The pediatric population in our study is mostly urban, and a high percentage of the population is nonwhite (approximately 60%). Our study is in accord with the NHANES III data, which revealed a decline in prevalence of elevated BLLs (≥10 μg/dL) from 77.8% in 1976 to 4.4% (1999), although many of the children tested within the CNMC lead study are at risk for high lead levels because of socioeconomic status and associated ethnic variation. The overall mean BLLs for the US population of age 1 year and older (including all adults tested) was 2.3 μg/dL.²⁰ In our study the geometric mean for female subjects 0–17 years was 3.0 μg/dL and for male subjects 0–17 years 3.2 μg/dL. However, when the results were stratified by age, the geometric mean for 10-to 17-year-olds (both genders) was 2.4 μg/dL.

NHANES III data show differences between urban and rural populations among both boys and girls, with BLLs higher in both gender groups in urban areas at the 90th percentile. The 90th percentile (upper end of both lead and ZPP ranges) decreases with age in both males and females. The only notable exception to this is the 0 to 11-month 90th percentile being lower than the 1- to 4-year 90th percentiles. This is true both for CNMC and for the US NHANES III data. NHANES III data demonstrate that boys display higher exposure levels than girls and, in the over-18-year age group had a significantly higher 90th percentile than the 12- to 18-year-old age group. This is probably a result of exposure to lead through an occupation or a hobby in the over-18-year-old age group.

The dose–response curves for the population relationship between ZPP and blood lead concentrations are nonlinear for adults and for children.²² As expected, ZPP concentrations were steady at blood lead levels below 18 μg/dL; however, beginning at around BLL 18 to 20 μg/dL, ZPP concentrations rapidly increased, disclosing a sharp concentration–response relationship that was especially apparent when represented by the LOWESS fitted curve. The calculated values for ZPP and BLL are indicated in Table 2.

In summary, BLLs in the United States have declined consistently since 1976. The Washington, DC-area tendency parallels the US national data, indicating that exposure to lead has declined considerably in the past 10 years. Despite these regional and national declines, lead’s pernicious health effects militate toward continued monitoring as lead exposure still lingers as a significant pediatric health concern.