The International Journal of Neuropsychopharmacology

Discontinuation symptoms in depression and anxiety disorders

David S. Baldwin a1c1, Stuart A. Montgomery a2, Rico Nil a3 and Malcolm Lader a4
a1 Clinical Neurosciences Division, School of Medicine, University of Southampton, UK
a2 Imperial College School of Medicine, London, UK
a3 Lundbeck (Switzerland) Ltd, Glattbrugg, Switzerland
a4 Institute of Psychiatry, University of London, London, UK

Article author query
baldwin ds   [PubMed][Google Scholar] 
montgomery sa   [PubMed][Google Scholar] 
nil r   [PubMed][Google Scholar] 
lader m   [PubMed][Google Scholar] 


The present overview investigates whether different antidepressants have differing discontinuation symptoms upon treatment cessation, if these symptoms vary between depression and anxiety disorders, and with length of treatment. Data came from two comparative studies of escitalopram in major depressive disorder (MDD) (one vs. venlafaxine XR and one vs. paroxetine), two studies in social anxiety disorder (SAD) (one of which used paroxetine as the active reference), and one study in generalized anxiety disorder (GAD), using paroxetine as an active reference [total number of patients: escitalopram (n=1051); paroxetine (n=336); venlafaxine XR (n=124); placebo (n=239)]. All studies included a defined discontinuation period and used the Discontinuation Emergent Signs and Symptoms (DESS) checklist to record the number of discontinuation symptoms. All three antidepressants showed more discontinuation symptoms compared with placebo (p<0.001). Patients reported significantly fewer discontinuation symptoms with escitalopram than with paroxetine and venlafaxine XR in MDD (p<0.05). Escitalopram showed significantly fewer discontinuation symptoms than paroxetine in SAD (p<0.05) and GAD (p<0.001). For each antidepressant, no differences in discontinuation symptoms were observed between the three indications and there was no evidence for increased symptom incidence with increased length of treatment. Thus, discontinuation profiles differ between antidepressants of the same class and are broadly similar in different disorders. No evidence was seen for a higher discontinuation burden with longer treatment.

(Received September 12 2005)
(Reviewed October 19 2005)
(Revised October 26 2005)
(Accepted October 30 2005)
(Published Online December 19 2005)

Key Words: Affective disorders; discontinuation effects; escitalopram; paroxetine; venlafaxine.

c1 University Department of Mental Health, Royal South Hants Hospital, Graham Road, Southampton, SO14 0YG, UK. Tel.: +44 (0) 2380 825533 Fax: +44 (0) 2380 234243 E-mail: