[Skip to Content]
[Skip to Content Landing]
April 28, 1989

Estimates of Gestational Age-Reply

Author Affiliations

McGill University Faculty of Medicine Montreal, Canada

McGill University Faculty of Medicine Montreal, Canada

JAMA. 1989;261(16):2329-2330. doi:10.1001/jama.1989.03420160053018

In Reply.—  Drs Knoke and Sokol provide a clear and heuristic explanation of how regression to the mean can produce bias in estimates of gestational age when those estimates are based on the mother's recollection of the LNMP and are associated with appreciable random error. We in fact mentioned the phenomenon (on page 3308 of our article) in discussing possible reasons for our results, but dismissed it as insufficient to account for either the magnitude of the biases we observed or the differences between apparent (ie, LNMP-based) preterm and postterm births.As we stated in our article, "Regression toward the mean due to random errors in recall at the two extremes of menstrual dates can explain only a small part of the phenomenon, since random errors in menstrual dating have been shown to be much smaller in magnitude than those demonstrated herein. Nor does it explain the larger errors postterm