[Skip to Content]
[Skip to Content Landing]
April 10, 1987

The Health Effects of Heptachlor

Author Affiliations

Harvard University Boston

Harvard University Boston

JAMA. 1987;257(14):1900. doi:10.1001/jama.1987.03390140070024

To the Editor.—  The investigation by Stehr-Green et al1 into the health effects of heptachlor exposure suffers from inadequacies in analysis and seriously overstated conclusions.As in many environmental health studies, direct exposure measurements of the variable of interest, heptachlor, did not exist. However, at least three proxy measures were available: serum levels of pesticide residues, self-reported levels of milk consumption, and membership in a "high-risk" farm family. Despite careful reading, we are unable to decipher whether the second or the third measure was used in the analysis; however, misclassification due to recall bias, variation in individual absorption, and other causes would be increased using either measure. Indeed, the data suggest that such misclassification occurred; only 23% to 31% of "exposed" subjects had elevated serum analytes, while 4% to 7% of participants in the second National Health and Nutrition Examination Survey had such elevations. It is possible that a