We investigated the capacity of several thyroid-axis measures to disti
nguish between depressed and control subjects and determined whether t
hese variables were related to antidepressant treatment response. We s
tudied 105 subjects who fulfilled the DSM-III-R criteria for a current
major depressive episode and 41 volunteers with no current mental dis
order. The following thyroid-axis variables were measured: difference
between T-4 levels at 09.00 hours and 13.00 hours; baseline TSH; maxim
al TSH response to 400 mu g TRH (Delta(max) TSH); and presence of a bl
unted Delta(max) TSH. The T-4 difference variable alone distinguished
between depressed and control subjects. In multivariate analyses, T-4
difference and Delta(max) TSH were independently related to antidepres
sant-treatment outcome, and predicted a modest proportion (14%) of the
variance in outcome. The relationship between these two variables and
treatment outcome was particularly strong in depressed male subjects
who were receiving desipramine, for whom they accounted for 36% of the
variance in treatment outcome. The T-4 difference variable both disti
nguished between depressed and control subjects and was related to tre
atment outcome. Although this finding requires replication, it is cons
istent with other reports of the usefulness of thyroid-axis indices me
asured at different times of day in depressed patients.