> otherwise identical cohorts.

You could do this stratification/matching for any IQ study. What's special about lead-IQ?