Interet de l'EEG de sommeil en tant que marqueur biologique des etats depressifs. Comparaison avec trois tests neuro-endocriniens*

Summary In a sample of 12 endogenous depressive inpatients (8 primary and 4 secondary depressives), we compared the diagnostic usefulness of REM latency (recorded during at least 4 consecutive nights) with 3 neuroendocrine tests: dexamethasone suppression test and GH response after cionidine (a α-adrenergic agonist) and apomorphine (a dopaminergic agonist) challenges. Shortened REM latency (less than 50 min during at least 1 night) was present in 67% of depressives. However, REM latency presented a clear night to night intra-patient variability that makes it necessary to record at least 3 consecutive nights for the best sensitivity. Non-suppression after dexamethasone was present in 50% of depressives, blunted GH response after clonidine, in 75% and blunted response after apomorphine, in 42%. A total of 92% of patients exhibited at least one abnormal biological parameter (100% of primary and 75% of secondary depressives); 67% of patients exhibited at least two disturbed parameters and these patients constituted the whole primary depressive group (100%). These results show that these 4 potential biological markers of depression are not necessarily distributed in the same population. This suggests the potential usefulness of their concurrent use for improved accuracy of diagnosis.