To the Editor.
—The attempt by Extein and associates (Archives 1982;39:77-81) to compare neuroendocrine responses in distinct diagnostic groups represents an important strategy that eventually will facilitate diagnosis in patients with ambiguous symptoms. The authors measured levels of thyroid-stimulating hormone (TSH) after infusion of protirelin (thyrotrophin-releasing hormone) and discovered that manics seem to have a blunted response compared with that of normal controls and schizophrenics. As is appropriate when measuring a multidetermined aspect such as level of TSH, they cautioned that "factors that may influence the TSH response to protirelin must be considered in evaluating the diagnostic difference in the results...." For example, by excluding patients taking lithium carbonate, they avoided a potential source of error found in an earlier attempt to characterize the response to protirelin in manics.1 However, by failing to report the sleep patterns of their sample, they seemingly failed to achieve their goal of showing that
Green HS, Kane JM. Neuroendocrine Responses in Diagnostic Groups. Arch Gen Psychiatry. 1982;39(10):1219-1220. doi:10.1001/archpsyc.1982.04290100075018