search text   Advanced Search

This Article Abstract Full Text (PDF) P3Rs: Submit a response Alert me when this article is cited Alert me when P3Rs are posted Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Yuan, W. Articles by Lanphear, B. P. Search for Related Content PubMed PubMed Citation Articles by Yuan, W. Articles by Lanphear, B. P. Related Collections Therapeutics & Toxicology

The Impact of Early Childhood Lead Exposure on Brain Organization: A Functional Magnetic Resonance Imaging Study of Language Function

^a Departments of Radiology
^b Pediatrics
^c Cincinnati Children's Environmental Health Center, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio
^d Department of Environmental Health
^e Center for Epidemiology and Biostatistics, University of Cincinnati College of Medicine, Cincinnati, Ohio

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
OBJECTIVES. The purpose of this work was to assess the long-term impact of childhood lead exposure on the neurosubstrate of language function and brain organization.

METHODS. Young adults from the Cincinnati Lead Study were recruited to undergo functional magnetic resonance image scanning while performing a verb generation task. These subjects have been followed from birth through early childhood with extensive documentation of lead exposure, neuropsychology, and behavior. Forty-two subjects provided useful imaging data. The locale, strength, and the correlation between brain language activation and childhood blood lead concentration were studied.

RESULTS. After adjusting for potential confounders, the activation in left frontal cortex, adjacent to Broca's area, and left middle temporal gyrus, including Wernicke's area, were found to be significantly associated with diminished activation in subjects with higher mean childhood blood lead levels, whereas the compensatory activation in the right hemisphere homolog of Wernicke's area was enhanced in subjects with higher blood lead levels.

CONCLUSION. This study indicates that childhood lead exposure has a significant and persistent impact on brain reorganization associated with language function.

Key Words: environmental pollutants • lead poisoning • neuroimaging • language development • pediatric research

Abbreviations: fMRI—functional magnetic resonance imaging • CLS—Cincinnati Lead Study • EPI—echo planar imaging • BOLD—blood oxygen level dependant • SES—socioeconomic status • IQ—intelligence quotient • ROI—region of interest

Childhood lead exposure is inversely associated with intellectual abilities, academic achievement, and psychomotor development.^1–4 Even low-level lead exposure can negatively impact a wide range of cognitive functions, such as attention, language, memory, cognitive flexibility, and visual-motor integration based on various types of neuropsychological and neurobehavioral testing.^5–8 However, because of the nature of the research methodology, previous studies were unable to directly localize functional changes in the brain. The underlying mechanism by which lead disrupts brain function in children, especially for low lead concentrations that do not produce noticeable physical signs, remains poorly understood.

Language, a unique human neurocognitive function, is often the earliest marker for the presence of a developmental or acquired neurologic disorder. The association between lead exposure and language function disorder has been studied extensively in cognitive research.^9–12 In the present study, we used functional MRI (fMRI) and investigated the impact of early childhood lead exposure on the cortical organization of language in young adults who experienced a wide range of lead exposure levels during development. Measurement of language activation patterns in various brain regions of these young adults was expected to help us to localize the deleterious effects on language function and language reorganization associated with lead exposure. We hypothesize that lead exposure during childhood significantly influences brain development and organization of language function. Further, compensatory brain language functional reorganization is expected as a consequence of the subtle brain insult associated with childhood lead exposure.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Subjects
Forty-five young adults with documented childhood lead exposure from the Cincinnati Lead Study (CLS) were recruited for the study. The CLS, a birth cohort recruited from 1979 to 1984, enrolled pregnant women who lived in neighborhoods with high rates of childhood lead poisoning. The pregnant women were excluded if they were known to be addicted to drugs, diabetic, or had any known neurologic or psychiatric malady. Infants were excluded if their birth weight was <1500 g or if genetic or other serious medical issues were present at birth.

These subjects have been followed from birth through early childhood with extensive documentation of lead exposure, social and medical history, intellectual attainment, neuromotor function, academic achievement, and behavior.^5,13 Blood lead concentrations were determined on a quarterly basis from birth until 5 years of age and again at 5.5, 6.0, and 6.5 years. The mean childhood blood lead concentration, beginning at 3 months and continuing through 78 months of age, was considered to be more representative of the average exposure of the population and, thus, was chosen as our lead level of interest in the study. It should be noted that results similar to that reported later in this study were found in similar brain regions for many of the other age time points. Table 1 briefly summarizes the demographic information of the subjects and the social, family, and individual factors that may potentially confound the association between outcome and exposure in our study. Each subject completed a urine screening for drugs of abuse at the time of imaging. All positive results (N = 22) indicated the presence of cannabinoids. Because marijuana was the only drug detected and used with any frequency in this sample, it was the only drug tested for potential confounding effect, although potentially other drugs might also affect the brain function. All 45 of the subjects finished the verb generation task successfully. However, 3 subjects were removed from the analysis because of excessive head motion (1), data corruption (1), or a significant missing demographic record (1). Therefore, only 42 subjects' results remained for statistical analysis (N = 42; 22 males; mean childhood blood lead level = 14.18 ± 6.52 µg/dL; range: 4.77–31.06 µg/dL).

The Language Task and Experimental Paradigm
The fMRI experimental paradigm followed a standard block periodic design in which "test" period was interleaved with "control" period. The test interval consisted of a 30-second period of verb generation task to collect language activation data. During each test period, a series of nouns were presented to the subject once every 5 seconds. The subject was instructed to silently generate verbs associated with each noun presented. For example, if the noun "ball" was presented, the subject should generate the verbs "throw," "kick," or "hit." The subject was instructed to think about these verbs silently instead of saying them aloud to avoid unnecessary motion artifact. The details of this task have been based on and documented in detail in previous studies.^14–19

Bilateral finger tapping was used as the control task during each 30-second resting period. This task was designed to control for the auditory prompt used in the verb generation task, to distract subject from unintentionally continued engagement in the verb generation task during rest interval, and to provide activation of motor strip as reference data for each subject.

At the beginning of each experiment, a 30-second dummy scan was conducted to allow the magnetization to reach magnetic equilibrium. Five cycles of test and control data were subsequently collected. In each of these 30-second periods, 24 image slices were collected once every 3 seconds. Therefore, there are 110 time points of data collected in 5 minutes and 30 seconds for the language task imaging. The initial 10 images during the dummy scan were discarded with 100 time points of data remaining for the following data analysis.

Image Acquisition
All of the images were acquired on a 3.0 T Bruker Biospec 30/60 MRI scanner (Bruker BioSpin MRI, Inc, Billerica, MA). A 3-plane gradient echo scan was performed to provide a reference for the alignment and localization of the brain followed by a shim procedure to optimize a homogeneous magnetic field. A T2*-weighted gradient-echo echo planar imaging (EPI) sequence was used during the subsequent functional sessions (repetition time/echo time: 3000/38 milliseconds; field of view: 25.6 x 25.6 cm; matrix: 64 x 64; slice thickness: 5 mm; and flip angle: 90°). The whole-brain high-resolution T1-weighted anatomic data set was obtained using a modified driven equilibrium Fourier transform pulse sequence with the following parameters: repetition time/echo time: 15.7/4.3 milliseconds; inversion time: 550 milliseconds; field of view: 19.2 x 25.6 x 16.0 cm; matrix: 256 x 192 x 128; and imaging time: 14 minutes, 40 seconds).

Data Processing and Analysis
Image reconstruction, postprocessing, and group statistical analysis were performed with custom software (Cincinnati Children's Hospital Imaging Processing Software) written in Interactive Data Language (Research Systems Inc, Boulder, CO).

Raw EPI data were reconstructed using the multiecho phase reference scan to correct for Nyquist ghosting and geometrical distortion artifacts.²⁰ The reconstructed EPI-fMRI time series were then corrected for baseline drift using a quadratic baseline correction algorithm on a pixel-by-pixel basis. Motion artifacts were corrected using a pyramid iterative coregistration algorithm.²¹ The images were then transformed into Talairach coordinate²² for the subsequent group analysis.

The fMRI data were first postprocessed with a general linear model²³ to relate the time series with brain activation on a pixel-by-pixel basis. For each subject, a Pearson's correlation coefficient between brain activation and time course was calculated for each pixel, and then a z score was derived from the coefficient by Fisher's z transformation. During the group analysis, a 1-sample t test was performed for each pixel across all of the subjects to test the significance of the pixel in the group. After transforming t statistics to z score, a statistical parametric map (composite z-score map) can be generated to identify brain regions with the most significant contrast between language task and control task for the entire group. This was a random-effect analysis that accounted for the intersubject variance. A cluster method was used to improve specificity and adjust for the inflated from multiple comparisons.²⁴ Monte Carlo simulation performed on the data set showed that a cluster of 3 and nominal z = 3 would lead to a corrected P < .001.

We also conducted a random-effect group analysis to test the association between blood oxygen level–dependant (BOLD) signal activation and mean childhood blood lead levels accounting for likely confounders. This analysis involved the following steps: (1) the brain activation in response to verb generation task was first studied independently for its relationship with childhood lead exposure by pixelwise simple linear regression analysis with a corrected P < .05 (cluster = 9 based on Monte Carlo simulation); (2) because the lead exposure might be associated with other factors that could also contribute to the modification of BOLD signal patterns, 6 such variables, including subjects family's socioeconomic status (SES), subjects' full-scale intelligence quotient (IQ), birth weight, gender, drug (marijuana) usage, and gestational age, were pretested separately for their potential confounding influence to language function. After adding a putative confounder into the otherwise simple linear regression between BOLD and mean childhood blood lead level, the change of regression coefficient (ß) for lead exposure was calculated on a pixel-by-pixel basis to evaluate the influence of the corresponding variable. This testing was conducted within the cortical areas that had been tested to correlate significantly with childhood blood lead level as determined in the previous step. The variable was kept for subsequent final multivariate analysis if adding this variable caused >20% of the pixels within the predetermined ROI to have over a 10% change in the ß value; (3) the final step of this group analysis was to perform a multiple regression analysis to evaluate the main effect of early childhood lead exposure on language function at young adulthood with all of the significant or marginally significant confounders. Because IQ has always been a major index in the literature for lead behavioral studies, we took an extra step to determine if adding IQ to the 4-term regression (BOLD versus blood lead, with birth weight and marijuana as confounders) would appreciably alter the exposure effect. Because the result was negative, IQ, together with gender, SES, and gestational age, was discarded.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Figure 1 presents a statistical parametric map (composite z-score map) that represents the significant language activation (nominal z = 3; cluster = 3; corrected P < .001) across the entire subject group in our study after correction for multiple comparisons. The functional areas for language reported previously using the verb generation task^19,25 are all activated in our CLS cohort: Broca's area (BA 44 and BA 45) at the left inferior frontal gyrus and Wernicke's area (BA 21 and BA 22) at the left superior temporal gyrus and the left middle temporal gyrus, as well as some homologues that are located contralaterally in the right hemisphere. A strong left hemispheric dominance of frontal and temporal language activation is also observed in this composite map.

View larger version (82K): [in this window] [in a new window]   FIGURE 1 Composite fMRI activation map for the verb generation task showing z-score statistics in young adults with childhood lead exposure (n = 42). Most highly activated areas include left inferior frontal gurus, left medial temporal gyrus, and right medial temporal gyrus. The orientation of the images follows radiologic convention.

View larger version (76K): [in this window] [in a new window]   FIGURE 2 Composite partial R map. The pixels with significant correlation between brain activation and lead level with confounders taken into account are coded with color and overlaid on anatomical image. The regions that have significant correlation between activation and lead level without taking confounders into consideration are demonstrated with white lines.

A mean activation value (mean z score) was calculated across all of the pixels in each of the 4 regions (as defined in Table 2) for each subject. Figure 3 A and B show the scatter plots, along with the partial correlation coefficient (partial R), of the mean brain activation within the corresponding ROI versus the mean childhood blood lead concentration across subjects for ROI_1 (BA46, BA47, and BA10: partial R = –0.32; P < .04) and ROI_2 (BA22 and BA42: partial R = 0.35; P < .03), respectively. The correlation between the mean activation versus mean childhood blood lead concentration is also significant for ROI_3 (BA21: partial R = –0.31; P < .05) but not for ROI_4 (BA46 and BA9: partial R = –0.30; P = .08).

View larger version (20K): [in this window] [in a new window]   FIGURE 3 A, Multivariate linear regression of brain activation in left inferior frontal gyrus (ROI_1) versus childhood mean blood lead level adjusted for confounders. Partial R = –0.328; P = .039. B, Multivariate linear regression of brain activation in right middle temporal gyrus (ROI_2) versus childhood mean blood lead level adjusted for confounders. Partial R = 0.354; P = .025.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

We also found a positive correlation between childhood mean blood lead level with brain activation in the right hemisphere homologs of Wernicke's area, as shown in ROI_2 (Figs 2 and 3B). This suggests that a compensatory mechanism may be at work at the level of the neural circuitry supporting language. As language begins to emerge in early childhood, there is a documented trend toward increasing focus of language-related activity in the left temporal and frontal lobe regions in normally developing children.^19,25 If these regions are not available either because of impaired connectivity or reduced neuronal capacity, the brain deploys new areas to replace the lost capacity for language normally subserved by these preferred neural circuits.²⁶ Regions of positive correlation between lead exposure level and later language activation that occur in the right temporal lobe (ROI_2 in Fig 2) correspond closely with the location of Wernicke's area (BA22) in the left hemisphere. This is consistent with previous neuroimaging studies describing the impact of developmental and/or acquired lesions, for example, tumor,²⁷ epilepsy,²⁸ stroke,²⁹ and dyslexia,³⁰ on language functional development. The neuroplasticity observed in the present study is consistent with our understanding about language function reorganization: the brain attempts to compensate for injury to regions that are structurally and functionally dedicated to language by recruiting support from other cortical regions. The engagement of the analogous right hemisphere temporal lobe in the compensation process provides evidence for a dormant language circuit in the nondominant hemisphere. This is a hypothesis that has been proposed and supported by other fMRI language studies.^31,32

One possible explanation for these findings is that lead exposure during early childhood specifically impedes brain regions that are undergoing rapid development to support nascent language capabilities. This might result in language delays in children exposed to lead during the preschool years. Later in childhood as these children enter kindergarten and are taught to read, write, and form complete sentences, demand for neural support of language skills accelerates, forcing the brain to compensate for the injury suffered during preschool years. Dormant, contralateral language circuits in the right hemisphere with similar cytoarchitecture are capable of subserving language functions when recruited for this purpose when formal education begins in the classroom. However, it should be noted that the compensatory alternative pathway does not necessarily yield equivalent performance to that achieved using the normative cortical circuitry for the same function. The degree to which this compensation mechanism is able to meet the demand for the development of language function is assumed to be associated not only with the type of insults, but also with the timing, duration, and intensity of the insult requiring further investigation.

	CONCLUSIONS

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
This fMRI study offers a new insight into the neural substrates of semantic language function that are selectively influenced by lead exposure during childhood. Elevated childhood lead exposure exerts a substantial influence on the cortical organization of semantic language function in young adulthood, demonstrated by a selective, deleterious effect on normal language areas with concomitant recruitment of contralateral regions, resulting in striking, exposure-dependent patterns of recruitment for language function. These imaging data provide further confirmation of the adverse consequences of environmental lead exposure on cognitive abilities.

	ACKNOWLEDGMENTS

 
This work was supported by grants from the National Institutes of Health (National Institute on Environmental Health Sciences PO1-ES011261, National Institute on Environmental Health Sciences R21 ES013524, National Institute of Child Health and Human Development R01 HD38578, National Cancer Institute R01 CA112182) and the US Environmental Protection Agency (R82938901).

	FOOTNOTES

 
Accepted Mar 29, 2006.

Address correspondence to Kim M. Cecil, PhD, Cincinnati Children's Hospital Medical Center, Department of Radiology/MLC 5031, 3333 Burnet Ave, Cincinnati, OH, 45229. E-mail: Kim.Cecil{at}cchmc.org

The authors have indicated they have no financial relationships relevant to this article to disclose.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

1. Lanphear BP, Hornung R, Khoury J, et al. Low-level environmental lead exposure and children's intellectual function: an international pooled analysis. Environ Health Perspect. 2005;113 :894 –899[ISI][Medline]
2. Dietrich KN, Berger OG, Succop PA, Hammond PB, Bornschein RL. The developmental consequences of low to moderate prenatal and postnatal lead exposure: intellectual attainment in the Cincinnati Lead Study Cohort following school entry. Neurotoxicol Teratol. 1993;15 :37 –44[CrossRef][ISI][Medline]
3. Bellinger DC, Stiles KM, Needleman HL. Low-level lead exposure, intelligence and academic achievement: a long-term follow-up study. Pediatrics. 1992;90 :855 –861[Abstract/Free Full Text]
4. Needleman HL, Schell A, Bellinger D, Leviton A, Allred EN. The long-term effects of exposure to low doses of lead in childhood. An 11-year follow-up report. N Engl J Med. 1990;322 :83 –88[Abstract]
5. Ris MD, Dietrich KN, Succop PA, Berger OG, Bornschein RL. Early exposure to lead and neuropsychological outcome in adolescence. J Int Neuropsychol Soc. 2004;10 :261 –270[CrossRef][ISI][Medline]
6. Baghurst PA, McMichael AJ, Tong S, Wigg NR, Vimpani GV, Robertson EF. Exposure to environmental lead and visual-motor integration at age 7 years: the Port Pirie Cohort Study. Epidemiology. 1995;6 :104 –109[ISI][Medline]
7. Bellinger D, Hu H, Titlebaum L, Needleman HL. Attentional correlates of dentin and bone lead levels in adolescents. Arch Environ Health. 1994;49 :98 –105[ISI][Medline]
8. Stiles KM, Bellinger DC. Neuropsychological correlates of low-level lead exposure in school-age children: a prospective study. Neurotoxicol Teratol. 1993;15 :27 –35[CrossRef][ISI][Medline]
9. Wang CL, Chuang HY, Ho CK, et al. Relationship between blood lead concentrations and learning achievement among primary school children in Taiwan. Environ Res. 2002;89 :12 –18[Medline]
10. Fergusson DM, Horwood LJ. The effects of lead levels on the growth of word recognition in middle childhood. Int J Epidemiol. 1993;22 :891 –897[Abstract/Free Full Text]
11. Mayfield SA. Language and speech behaviors of children with undue lead absorption: a review of the literature. Speech Hear Res. 1983;26 :362 –368
12. Campbell TF, Needleman HL, Riess JA, Tobin MJ. Bone lead levels and language processing performance. Dev Neuropsychol. 2000;18 :171 –186[CrossRef][ISI][Medline]
13. Dietrich, K., Bearer, CF, Kahn, R, et al. Long-term adverse neurobehavioral consequences of low-level exposure to environmental toxins: an update of the Cincinnati Children's Environmental Health Center [abstract]. Epidemiology. 2004;15 :S90
14. Binder JR. Neuroanatomy of language processing studied with functional MRI. Clin Neurosci. 1997;4 :87 –94[CrossRef][ISI][Medline]
15. Binder JR, Frost JA, Hammeke TA, Cox RW, Rao SM, Prieto T. Human brain language areas identified by functional magnetic resonance imaging. J Neurosci. 1997;17 :353 –362[Abstract/Free Full Text]
16. Petersen SE, Fox PT, Posner MI, Mintun M, Raichle ME. Positron emission tomographic studies of the cortical anatomy of single-word processing. Nature. 1988;331 :585 –589[CrossRef][Medline]
17. Benson RR, Logan WJ, Cosgrove GR, et al. Functional MRI localization of language in a 9-year-old child. Can J Neurol Sci. 1996;23 :213 –219[ISI][Medline]
18. Benson RR, FitzGerald DB, LeSueur LL, et al. Language dominance determined by whole brain functional MRI in patients with brain lesions. Neurology. 1999;52 :798 –809[Abstract/Free Full Text]
19. Holland SK, Plante E, Byars AW, Strawsburg RH, Schmithorst VJ, Ball WS Jr. Normal fMRI brain activation patterns in children performing a verb generation task. Neuroimage. 2001;14 :837 –843[CrossRef][ISI][Medline]
20. Schmithorst VJ, Dardzinski BJ, Holland SK. Simultaneous correction of ghost and geometric distortion artifacts in EPI using a multiecho reference scan. IEEE Trans Med Imaging. 2001;20 :535 –539[CrossRef][ISI][Medline]
21. Thevenaz P, Unser M. A pyramid approach to subpixel Registration based on intensity. IEEE Trans Image Process. 1998;7 :27 –41[Medline]
22. Talairach J, Tournoux P. Co-Planar Sterrotaxic Atlas of the Human Brain. New York, NY: Thieme Medical Publishers, Inc;1988
23. Worsley KJ, Liao CH, Aston J, et al. A general statistical analysis for fMRI data. Neuroimage. 2002;15 :1 –15[CrossRef][ISI][Medline]
24. Xiong J, Gao JH, Lancaster JL, Fox PT. Clustered pixels analysis for functional MRI activation studies of the human brain. Hum Brain Mapp. 1995;3 :287 –301[CrossRef][ISI]
25. Szaflarski JP, Holland SK, Schmithorst VJ, Byars AW. fMRI study of language lateralization in children and adults. Hum Brain Mapp. 2006; 27: 202–212[CrossRef][ISI][Medline]
26. Lidsky TI, Schneider JS. Lead neurotoxicity in children: basic mechanisms and clinical correlates. Brain. 2003;126 :5 –19[Abstract/Free Full Text]
27. DeVos KJ, Wyllie E, Geckler C, Kotagal P, Comair Y. Language dominance in patients with early childhood tumors near left hemisphere language areas. Neurology. 1995;45 :349 –356[Abstract]
28. Saltzman-Benaiah J, Scott K, Smith ML. Factors associated with atypical speech representation in children with intractable epilepsy. Neuropsychologia. 2003;41 :1967 –1974[CrossRef][ISI][Medline]
29. Thulborn KR, Carpenter PA, Just MA. Plasticity of language-related brain function during recovery from stroke. Stroke. 1999;30 :749 –754[Abstract/Free Full Text]
30. Shaywitz SE, Shaywitz BA, Pugh KR, et al. Functional disruption in the organization of the brain for reading in dyslexia. Proc Natl Acad Sci USA. 1998;95 :2636 –2641[Abstract/Free Full Text]
31. Schmithorst VJ, Holland SK, Plante E. Cognitive modules utilized for narrative comprehension in children: a functional magnetic resonance imaging study. Neuroimage. 2006; 29: 254–266[CrossRef][ISI][Medline]
32. McClelland JL, Rogers TT. The parallel distributed processing approach to semantic cognition. Nat Rev Neurosci. 2003;4 : 310–322[CrossRef][ISI][Medline]

This Article Abstract Full Text (PDF) P3Rs: Submit a response Alert me when this article is cited Alert me when P3Rs are posted Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Yuan, W. Articles by Lanphear, B. P. Search for Related Content PubMed PubMed Citation Articles by Yuan, W. Articles by Lanphear, B. P. Related Collections Therapeutics & Toxicology

	Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved.