Lead Carbonate Intermediate

Epidemiological data

Administrative data

Endpoint:

epidemiological data

Type of information:

other: Epidemiological-Prospective-Lead Neurotoxicity (Pediatric)

Adequacy of study:

key study

Study period:

1985-2001

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

other: see 'Remark'

Remarks:

The rationale for the reliability scoring is based upon a Lead Study Quality Asessment tool for Pediatric Neurological Effects developed by the Scientific Advisory Panel assembled to provide independent academic review of the Voluntary Risk Assessment for Lead. The Pediatric Assessment assessed study quality on the basis of the following major aspects of study design, data collection, and analysis: Adequacy of Cohort Definition and Size Nature and Extent of Lead Exposure Indices Examiner Training and Procedures for Data Collection and Analysis Precision of Endpoint Definition Extent of Correction for Major Confounding Variables

Materials and methods

Study type:

cohort study (prospective)

Endpoint addressed:

neurotoxicity

Principles of method if other than guideline:

Epidemiologiical-Prospective-Lead Neurotoxicity (pediatric)

Method

Type of population:

other: General Population (Pediatric)

Ethical approval:

confirmed and informed consent free of coercion received

Details on study design:

In the 1980s, researchers realized that lead effects were small relative to other influences upon child development. Variability in study outcome was also thought to potentially arise as a result of factors unrelated to lead exposure. The selection of covariates, and their treatment in data analyses, was judged to be very important. There was also a need to better document lead exposure histories.
In September 1981, a conference was convened in Cincinnati to exchange information about study design and it was decided that the measurement of covariates and confounders, better documentation of the lead exposure variables and more detailed exposure histories could be obtained by undertaking longitudinal studies. In 1984, a second conference was held in Cincinnati and reports were given on longitudinal studies that were being planned or conducted. Ten prospective studies have attempted to detect a relationship between prenatal and postnatal blood lead levels and IQ deficits. Five of these studies, conducted in Boston, Cincinnati, Sydney, Port Pirie and Cleveland were designed similarly and were conducted around the same period. These studies obtained data on blood lead and neurobehavioural measurements from birth, six months or annual intervals up to 7-10 years.
Two other studies, conducted in Mexico City and Yugoslavia, have similar designs to the above but were initiated at later dates. The Glasgow and Nordenham studies are longitudinal, but have different study designs and measurement methodologies. Results have also been recently published from an ongoing study in Rochester, New York. Initiated as a study of relationships between lead in soil, dust and blood, this longitudinal study also differs from earlier studies in aspects of study design. All the longitudinal studies use multivariate analysis to take into account effects of possible confounding covariates upon the outcome measures. The studies have produced varying results that may reflect differences in the covariates used in the analysis, types of statistical measures used and nature of the subject population.
Although the longitudinal studies have not produced results fundamentally different from those obtained in cross-sectional studies, they have produced information at different stages of child development and further defined effects potentially related to exposure at different ages (including prenatal exposure).
This article reports on the results of neurobehavioral studies conducted on 253 children from the Cincinnati Prospective Study at 6.5 years. The authors have previously described the results of cognitive development testing at 4 and 5 years of age (Dietrich, 1991; Dietrich, 1992).
Details of the procedures for subject recruitment and eligibility criteria can be found in earlier reports of the Prospective study (Dietrich, 1991 and 1992). The sociodemographic and biomedoical characteristics of this predominantly African-American, inner-city cohort have also been previously presented (Dietrich, 1991). A total of 253 subjects were excluded from further analyses for medical conditions other than Pb poisoning (Visual handicap, n=1, and Sickle Cell Anemia, n=1).
The base population for this 6.5 year follow-up was the 347 subjects who were initially evaluated postnatally duribng the first two years of life at either 3, 6, 12, or 24 months of age. Two hundred and fifty-three subjects were examined psychologically and medically at approximately 78 months of age (M = 78.7 +/- 2 months).

Exposure assessment:

measured

Details on exposure:

Prenatal-8.4ug/dl
Neonatal 4.8ug/dl
2 yr PbB: 21.2ug/dl
Mean lifetime average PbB Conc. at 7 years:
Quartile 1: (low): 10.8ug/dl
Quartile 2: 15.7ug/dl
Quartile 3: 19.7ug/dl
Quartile 4 (high): 24.8ug/dl
Mean lifetime average PbB concentration 0-13 years: 14ug/dl
Mean PbB levels used in this ananlysis at six years:
3-12 months: 10.6ug/dl
15-24 months: 17.1ug/dl
27-36 months: 16.3 ug/dl
39-48 months: 14.0 ug/dl
51-60 months: 11.8 ug/dl
Blood lead was also sampled at 66, 72, and 78 months. Postnatal Pb exposures were expressed as the mean PbB concentrations of the 1st through 6th year of life (Mean-PbB1-6). Average lifetime blood Pb level was defined as the mean of 20 quarterly PbB assessments and the PbBs at 66 and 72 months of age (MPbBLife).

Statistical methods:

Data analyses employed multiple linear regression statistical techniques that examined the relationship between blood indices of prenatal/postnatal Pb exposure and cognitive developmental status at 6.5 years. Covariables were pretested for their confounding potential. Any cofactor associated with the primary developmental outcome variables at p< 0.15 was considered in the multivariate regressions. Following both backward and forward elimination procedures, any cofactor still associated with developmental outcomes at p<0.10 was retained in the model. Forward and backward elimination prodecures yielded the same models. Two potential effect modifiers, child sex, and social class were also examinied, Such interactions have been observed in previous studies and in this cohort in early infancy and the late preschool years. All p values cited in this report are two-tailed.

Results and discussion

Results:

This study found no prenatal effects on Global IQ measures. There was an effect on performance IQ only at 6 years of age in children with life-time average exposure blood leads in excess of 20 µg/dL (Dietrich 1993a).

Confounding factors:

fetal distres, and growth, perinatal complications, prenatal maternal substance abuse, sociodemographic characteristics, postnatal indices of health and nutritional status, maternal IQ, features of the Home environment known to be associated with adequate social and cognitive development.

Any other information on results incl. tables

Performance on the Wechsler Intelligence Scale For Children-Revised

Wechsler Subscale	Mean	SD	Lowest	Highest
Full Scale IQ (N=251)	86.9	11.3	50	118
Performance IQ (N=251)	88.5	12.6	46	120
Verbal IQ (N=252)	87.4	11.5	58	125

n varies for Wechsler subtests due to the presence of incomplete protocols for a few subjects

Covariate-Adjusted Regression Coefficients (And SEs) for PbB Indices and WISC-R FSIQ, PIQ and VIQ

Blood Lead	FSIQ (Full Scale)	PIQ (Performance)	VIQ (Verbal)
PrePbB	0.15	0.06	0.16
NeonPbB	-0.03	-0.22	0.11
Mean PbB1	0.01	-0.16	0.12
Mean PbB2	-0.02	-0.10	0.02
MeanPbB3	-0.13	-0.22*	-0.04
MeanPbB4	-0.15	-0.27*	-0.02
MeanPbB5	-0.23*	-0.38*	-0.07
MeanPbB6	-0.338	-0.52**	-0.02
MeanPbBLif2	-0.13	-0.26*	-0.01

*p</= 0.05 **p</= ).001 Regression coefficients adjusted for HOME score, maternal IQ, birth weight, birth length, child sex, and cigarette consumption during pregnancy. N= 199 for PrePbB analyses and 231 for postnatal PbB indices due to missing data for some covariates (e.g., maternal IQ, 3 yer H.O.M.E. scores).

Applicant's summary and conclusion

Conclusions:

These statistical analyses suggest that avaeraged lifetime blood lead concentrations in excess of 20ug/dl were associated with deficits in Performance IQ in the order of approximately 7 points when compared to children with mean concentrations less or equal to 10ug/dl.

Executive summary:

In a follow-up study of the Cincinnati Lead Study Cohort, 253 children were administered the Wchsler Intelligence Scale for Children-Revised (WISC-R) at approximately 6.5 years of age. Postnatal blood lead concentrations were inversely associated with Full-Scale (FSIQ) and Performance IQ (PIQ). Following statistical adjustment for developmental co-factors such as maternal IQ and an assessment of the quality of the caretaking in the home environment, a statistically significant relationship remained between averaged lifetime blood lead concentrations and six year PIQ. Further statistical analyses suggested that averaged lifetime blood lead concentrations in excess of 20ug/dl were associated with deficits in PIQ on the order of approximately 7 points when compared to children with mean concentrations less or equal to 10ug/dl.