The effect of in or outpatient status on EEG sleep data and postdex controls levels in depressed patients

Abstract

In a sample of inpatients with major depressive disorder (MDD), we reported that patients with short REM latency (RL) had higher postdexamethasone (postdex) plasma cortisol levels. We now report on the general association of EEG sleep measures and postdex cortisol levels in the above sample, with comparison to a group of 36 outpatients. EEG sleep data averaged over two nights were compared to data for the 1 .O mg DST. After a two-week drug wash-out period, all patients had a SADS/RDC diagnosis of definite MDD and had a HRS > 14. Clinical and demographic data showed the 52 inpatients did not differ from the 36 outpatients with respect to mean age or HRS. Concerning a comparison of EEG sleep measures, however, the inpatient group showed significantly more impaired sleep efficiency (77.4 ? 14.0% vs 83.0 & 11.77%, p < .05) and other measures of sleep continuity, and had greater abnormalities of REM sleep, including a shorter RL (35.0 ? 25.3 minutes vs 56.3 + 27.5 minutes, p < .OOl), and greater REM time (25.1 ? 13.8 vs 18.5 ? 9.3, p < .05) and REM activity (55.2 + 47.5 vs 29.9 + 24.0, p < .Ol) in the first REM period. A significant correlation between sleep measures and cortisol existed only within the inpatient group, in which the maximum postdex cortisol level was correlated positively with%awake(+0.40,p<.01),%stage1(+0.36,pC.01)andnumberof~ousals (+0.32, p < .05), and negatively with time spent asleep (-0.28, p < .05) and sleep efficiency (-0.44, p < .Ol). Cortisol was negatively correlated with REM latency minus awake (-0.44, p C .OOl) but not with REM activity or REM density. These data suggest that the relationship between sleep and postdex cortisol secretion may hold only in a more severely ill (i.e. inpatient) group.