Commentary: On Lead and IQ

If you wish to catch up, Chuck Dinerstein's article can be found here. 

In this article, I will address two issues: 1) Does this study represent an example of using a group of studies to boot-strap to a conclusion? (boot-strapping means to advance an idea to achieve a specific objective), 2) How reliable is IQ as a measure of intelligence?     

IQ and Lead Study

The current study of the effect of lead on cognition concludes that a total of 824,097,690 million IQ points were lost because of childhood lead exposure among the US population by 2015, an average of 2.6 IQ points per person.

The study estimated the number of people living in the US in 2015 who had various blood lead levels as young children, before, during, and after the era of leaded gasoline. The estimate was made using data from the US Census, statistics on leaded gasoline consumption, and data from the National Health and Nutrition Examination Survey (NHANES) from 1976 to 2016. The result showed that, for children born between 1976 and 2016, banning lead in gasoline resulted in a significant decline in childhood blood lead levels. For adults born between 1950 and 1980, the researchers estimated greater than 90% of the population had blood lead levels > 5 µg/dL, the current “acceptable” standard.      

The study then linked the data on blood lead levels with loss of IQ points using data from a 2005 study. This analysis grouped results from seven international studies on IQ and blood lead levels. The study examined 1,333 children from the US, Mexico, Australia, and Yugoslavia and concluded that

• an increase in blood lead levels from 2.4 to 10 µg/dL resulted in a decline of 3.9 IQ points
• curiously, increases in blood lead levels from 10 to 20 µg/dL resulted in a decline of only 1.9 IQ points and increases in blood lead levels from 20 to 30 µg/dL resulted in a lesser reduction of 1.1 IQ points  

This result violates one of the key premises of toxicology – that increasing the dose increases the response. As no explanation was provided for this, it questions the validity of the study.

Study Details

The seven international studies were quite dated, most being done earlier, during the 1980s. I had great difficulty trying to match up the results of each of the studies, as given in their individual reports, with the summary analysis of the 2005 study. For example, the study set in Rochester, New York, reported that the children were tested for IQ at 3 and 5 years of age, while the 2005 table said that the children were tested at 6 years of age. It was the only US study to show a decrease in IQ at blood lead concentrations below 10 µg/dL - an estimated overall loss in IQ for each increase in the lifetime average lead concentration of 1 µg/dL of -1.37 points. There were other issues

• The study set in Yugoslavia did not examine important confounding factors, such as maternal smoking or alcohol use. Although mentioned in the study, the authors did not factor in the consequences of the United Nations embargo of Serbia in the spring of 1992 that deprived the children of adequate food and other necessities. 
• One study in Boston, MA, did not find a statistically significant relationship between blood lead and IQ. The results were not consistent in direction when relevant confounding factors were considered.
• Another study in Cincinnati, OH, found associations between blood lead levels greater than 20 µg/dL and IQ decrements compared to children with blood lead levels less than 10 µg/dL. Since the children with blood lead levels less than 10 µg/dL were used as a control group, it is not possible to say whether there were IQ decrements at blood lead levels less than 10 µg/dL.  

Intelligence Quotient (IQ) – what does it mean?

An IQ is a total score derived from a series of standardized tests designed to test human intelligence.

The following presents different categories based on IQ scores, with the % of the population in each category:

• 130 IQ, very gifted, 2.1% of population
• 121-130 IQ, gifted, 6.4% of population
• 111-120 IQ, above average intelligence, 15.7% of population
•  90-110 IQ, average intelligence, 51.6% of population
• 80-89 IQ, below-average intelligence, 15.7% of population
• 70-79 IQ, cognitively impaired, 6.4% of population.

Most tests provide both a specific IQ score and an IQ range. That is because your score may vary based on an error of measurement or circumstances under which you took the test, but the average IQ score is always 100, and your personal score compares your IQ score with other people who took the same test.

Shortcomings of Lead-IQ studies

There are several shortcomings in studies that equate lead exposure with a decrease in IQ, including:

• Whenever lead levels are shown to be a predictor of IQ, authors control for several variables and conclude that the lead alone accounts for the IQ loss. This ignores all other important variables not controlled in the study, e.g., smoking and drinking by the mothers while pregnant.
• In most studies, parental IQ is not measured. This is a crucial variable affecting children’s IQ that is rarely accounted for in studies. In the 2005 analysis, maternal IQs were well-matched to their children.
• Since the IQ tests are considered to have a margin of error of approximately 5 points on either side, the small increases seen in most lead studies are within the margin of error.

With all these shortcomings, there is reason to question the results of any study that equates lead exposure with IQ loss.

Conclusions

As my colleague pointed out, the extrapolation is only as strong as its weakest link. I’ve tried to point to the many weak links in this study. Nobody questions that getting lead out of the environment is a worthwhile goal, and the recent efforts to remove lead from drinking water pipes across the US are important. However, it is a significant stretch to allow baby boomers to blame our problems and failures on our IQ losses due to lead exposure.