This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
—Dr Grossman's assertion that cross-sectional correlation studies (like ours) relating thyroid indexes to blood lead level convey no information regarding the temporal sequence of events is, of course, true in the mathematical sense. Provided only the quantitative results, one could not discern the "chicken from the egg."However, given the actual methodology employed, considerable interpretation is possible. We went to a factory and drew thyroid function studies and lead levels on all 47 apparently healthy employees. Twelve (26%) had low-thyroid indexes. What is the likehood of such a finding under the null hypothesis that lead has no influence on thyroid function? Given the approximately normal distribution of thyroxine (T4) values in our hospital laboratory, the expected number with low T4 among 47 men would be slightly greater than one with less than.00001 probability of finding 12. Even granting that the true distribution of T4 values may
Robins JM, Cullen MR. Thyroid Indexes-Reply. Arch Intern Med. 1983;143(9):1837. doi:10.1001/archinte.1983.00350090215052