Successful Pharmacologic Treatment of Major Depressive Disorder Attenuates Amygdala Activation to Negative Facial Expressions: A Functional Magnetic Resonance Imaging Study

Henricus G. Ruhé, Jan Booij, Dick J. Veltman, Martin C. Michel, Aart H. Schene

Research output: Contribution to journalArticleAcademicpeer-review

61 Citations (Scopus)


Objective: Studies of the effects of pharmacotherapy for major depressive disorder (MDD) on limbic-subcortical-prefrontal brain networks show variable results. We quantified functional changes in the amygdala and the related limbic-subcortical-prefrontal structures after paroxetine treatment with functional magnetic resonance imaging relative to clinical responder status. Method: We scanned 22 patients with unipolar, DSM-lV-defined MDD (men and women aged 25-55 years; 17-item Hamilton Depression Rating Scale [HDRS17] score > 18) at study entry and after 6 (TO) and 12 (T1) weeks of paroxetine treatment. Our paradigm contrasted negative (fearful, angry), happy, and neutral faces relative to scrambled faces. Twenty-one age-matched (+/-2.5 y) and sex-matched controls were scanned once. Patients received open-label paroxetine 20 mg/d for 6 weeks (TO). Nonresponders at TO were randomly assigned to receive double-blind true dose escalation (paroxetine 30-50 mg/d) or placebo dose escalation for another 6 weeks (T1). The study was conducted from July 2005 to February 2007. Results: At study entry, MDD patients showed increased ventral/limbic and decreased dorsal prefrontal activations to negative faces. At T0 and T1, respectively, 5/20 and 13/20 patients responded to paroxetine. After 12 weeks (at T1), overall amygdala activations remained unchanged relative to study entry. However, amygdala activations were significantly lower in treatment responders versus nonresponders (P=.001). Amygdala activations correlated with HORS17 scores (P <.04). Left amygdala activation correlated inversely with pregenual anterior cingulate cortex activation (P=.001). Dorsal cingulate gyrus and dorsolateral prefrontal activations increased after 6 and 12 weeks of treatment, regardless of clinical response. Conclusions: Successful paroxetine treatment decreases amygdala activation, presumably by improved frontolimbic control, in line with selective serotonin reuptake inhibitor-induced increased functional connectivity between the pregenual anterior cingulated cortex, prefrontal cortex, and amygdala. Changes in amygdala activation when processing negative faces might serve as an indicator for improved frontolimbic control, which is required for clinical response. Trial Registration: ISRCTN identifier: ISRCTN44111488 J Clin Psychiatry 2012;73(4):451-459 (C) Copyright 2011 Physicians Postgraduate Press, Inc
Original languageEnglish
Pages (from-to)451-459
JournalJournal of clinical psychiatry
Issue number4
Publication statusPublished - 2012

Cite this