Trends in preschool lead exposure, mental retardation, and scholastic achievement: Association or causation?
February 11, 2009
Rick Nevin, National Center for Healthy Housing
Journal of Environmental Research
Could the incidence of mental retardation and lagging SAT scores be related to juvenile lead exposure? Quite possibly, suggests Virginia economist Rick Nevin, in a study published this month. There appears to be a strong statistical relationship between mental retardation, SAT trends, and blood lead levels in children between 1936 and 1990. As lead levels went up, SAT scores went down, and vice versa. A rise in lead levels was associated with an increase in mental retardation cases. If a causal relationship were true, this would be a shocking discovery. Unfortunately, these findings are somewhat compromised by less-than-sturdy analytic methods. First, the "unit of analysis" is the entire United States--not an individual, much less a state; such a large scope means it's much easier to mask other factors that may have impacted the findings. Second, Nevin uses the percent of students with mental retardation in special education as his mental retardation variable, but we know that definitions of special education categories vary widely across states and have changed substantially over the last 50 years. And finally, when treating SAT takers, Nevin controlled for the number of students speaking a foreign language at home and for the number of students taking SAT test prep courses--both factors that could impact scores--but he didn't control for economic status, which is more relevant. A more credible approach would emerge from a true longitudinal study that followed lead-exposed children and correlated individual-level data on mental retardation and achievement as they grew older. We should by all means protect kids from consuming lead, but the verdict is still out on whether doing so will send SAT scores through the roof--and the incidence of mental retardation into the cellar. The study is available for a fee here.