Double-blind, placebo-controlled assessment of combined clonazepam with paroxetine compared with paroxetine monotherapy for generalized social anxiety disorder.

BACKGROUND: Generalized social anxiety disorder (GSAD) is a pervasive form of social anxiety that affects approximately 5% of persons in the community. Among evidence-based pharmacologic treatments for the disorder, selective serotonin reuptake inhibitors (SSRIs) have become widely used and are known to be efficacious. Monotherapy with the benzodiazepine clonazepam is also efficacious for GSAD, but the adjunctive use of clonazepam with an SSRI to potentially improve outcomes has not been studied to date.

METHOD: Twenty-eight patients (22 men and 6 women) with DSM-IV-defined GSAD were randomly assigned to receive double-blind clonazepam (or placebo), 1.0 to 2.0 mg/day (divided b.i.d.) along with open-label paroxetine, 20 to 40 mg/day, for 10 weeks. A 2-week taper of double-blind medication was followed by an additional 8 weeks of open-label paroxetine treatment (during which the dose of paroxetine could be increased to a maximum of 50 mg/day). Twenty-three patients (82%) met DSM-IV criteria for avoidant personality disorder. The patients' mean +/- SD age was 31.2 +/- 7.7 years, and their mean duration of illness was 12.1 +/- 5.8 years. Data were gathered from August 2001 to April 2002.

RESULTS: Nineteen (68%) of 28 patients completed treatment. At the end of the 10-week double-blind treatment, there was a trend (p <.06) favoring the paroxetine/clonazepam group, who had a 79% response rate (Clinical Global Impressions-Global Improvement scale [CGI-I] score of 1 or 2) compared with a 43% response rate for the paroxetine/placebo group. However, no significant differences on other outcome measures were noted between the 2 groups in an intent-to-treat analysis, in terms of either very early (2-4 weeks) or not as early (5-10 weeks) responses during treatment. Dropout rates due to adverse events were rare (1 patient in each group), indicating that the paroxetine/clonazepam combination was well tolerated.

CONCLUSION: Coadministration of clonazepam with an SSRI, in contrast to findings in panic disorder, did not lead to more rapid resolution of symptoms in GSAD. On the other hand, there is some evidence that the clonazepam-added group had superior global outcomes (e.g., as measured on the CGI-I), although power to detect such differences in this study was small. These observations suggest that a role for adjunctive benzodiazepines in patients with GSAD (e.g., for augmenting SSRI partial response or nonresponse) is deserving of further controlled investigation.