News Release

‘Don’t Brand Individuals With Large-Population Studies’

Peer-Reviewed Publication

Center for Advancing Health

Studies of large groups of children suggest that low levels of lead can have statistically significant effects on intelligence test scores, but these studies tell almost nothing about how individual children who are exposed to lead will fare, says a leading pediatric researcher.

Writing in the February issue of Developmental and Behavioral Pediatrics, Holly A. Ruff, Ph.D., of the Albert Einstein College of Medicine in New York City, cites the debate over low-level lead exposure and cognitive development as but one example of how people who make public policy or testify in legal proceedings erroneously attempt to apply population-based findings to explain individual behavior.

"With lead, as with any risk factor, some children or subgroups of children will be more vulnerable to the effects than others," Ruff writes. "Vulnerability or resilience in human beings is a complex phenomenon that involves both the biological characteristics of individual children and their physical and social environments. The sources of variation are virtually unlimited."

The most common measure of cognitive development, the intelligence quotient (IQ), has critical limitations, Ruff says.

* A score from one testing of IQ does not represent a stable trait; scores can be affected by a host of factors, including hunger, fatigue, and anxiety. * IQ tests may measure only a fraction of the skills necessary to function well in daily life.

* Most studies estimate that low to moderate lead exposure produces average declines of only a few points in IQ, well within the 5-point margin of error for individual scores.

"If the IQ is erroneously thought of as a perfectly reliable measure of a fixed trait that is an accurate predictor of performance in cognitive and academic tasks, then we are more likely to think of lead as having a permanent and significant effect on the individual's functioning," Ruff says.

Citing analyses that find exposure to lead explains only a small fraction of the range of differences in children's IQ scores, some of them as little as 2 percent, Ruff says most of the variation has to be attributed to other factors.

Another problem stems from the public's understanding of statistics, she says. "Statisticians use the term 'significant' to mean a particular finding is reliable or likely to be repeated. The public is likely to interpret 'significant' as 'important' -- a very different concept," she says.

"Thus, low levels of blood lead are significantly related to, but actually explain little of, the observed variation in IQ," she says.

Although the effects of low levels of lead are small over a large population of children, certain children do remain more vulnerable than others, Ruff says. Poor children from disadvantaged backgrounds are not only more likely to be exposed to lead, they appear more vulnerable to its effects. These children typically face other risk factors, such as poor nutrition and disease, and may lack certain "protective" factors, such as strong family support and a stimulating environment.

Since epidemiological studies report average results from large groups, they are most useful in developing public policy affecting large segments of the population, Ruff advises.

When considering individual children with low levels of exposure to lead or other risks, pediatricians and others should take care to evaluate -- the child's general health, the skill of the parents, the quality of nutrition, the nature of the physical environment, and many other factors," she says.

###

The Journal is published bi-monthly by the Society for Developmental and Behavioral Pediatrics. For information about the Journal, contact: Mary Sharkey at 212-460-9112.

Posted by the Center for the Advancement of Health http://www.cfah.org. For information about the Center, call Richard Hebert, 202-387-2829 rhebert@cfah.org.



Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.