Thirty double-blind RCTs, containing approximately 4,800 patients, were included in the review.
Antidepressant efficacy versus placebo (11 RCTs):
Three large multicentre trials comparing citalopram (10 to 80 mg) with placebo found significantly greater improvement in HAM-D, MADRS and CGI scores for depressed patients receiving citalopram. Three meta-analyses of placebo-controlled trials generated results consistent with those seen in the multicentre trials. A 6-week placebo-controlled trial of citalopram (20 to 30 mg) in depressed elderly patients with or without dementia found similar improvements, plus improvements on several individual items on the Gotfries-Brane-Steen scale of cognitive functioning.
Efficacy of citalopram versus placebo in the prevention of relapse (n=2):
When using citalopram as a continuation therapy for the prevention of relapse, patients who responded to citalopram (after 6 or 8 weeks) were randomly assigned to citalopram (20 to 60mg) or placebo for a further 24 weeks. In both trials, the rates of relapse (defined by a threshold MADRS score of 22 in one trial, 25 in the other) were significantly lower for citalopram than placebo.
Efficacy of citalopram in preventing recurrence of depression (n=2):
In one study, patients with a history of at least 2 prior depressive episodes, who responded to treatment with citalopram, received 16 weeks of continuous treatment at their established effective dose. Patients who continued to respond were randomised to at least 48 weeks of double-blind treatment with either continued citalopram or placebo. Eighty percent of citalopram-treated patients remained free of recurrence (MADRS score 22) throughout the maintenance phase of the study, compared with 50% of patients assigned to placebo (p<0.0001). In the second study with a similar design, citalopram was shown to be effective in the prevention of depression recurrence in patients aged over 65 years.
Comparative efficacy of citalopram versus tricyclic and tetracyclic antidepressants:
In 3 double-blind studies, citalopram (20 to 60 mg/day) and amitriptyline (50 to 225 mg/day) were found to be similarly efficacious among patients with major depression and elderly depressed patients.
A 5-week study of 102 depressed patients found citalopram (40 mg/day) to be less effective than clomipramine (150 mg/day). The rate of complete recovery (HAM-D score of 7 or less) was 60% in the clomipramine group and 30% in the citalopram group (p<0.005).
A large multicentre study in 400 depressed patients found patients receiving citalopram (10 to 30 mg/day; 20 to 60 mg/day) or imipramine (50 to 150 mg/day) experienced similar reductions in HAM-D scores at the end of both 6 and 22 weeks of treatment.
Two double-blind comparative trials concluded that citalopram was as effective as maprotiline in reducing HAM-D, MADRS and CGI scores from baseline.
Two small studies compared the efficacy of citalopram with mianserin in patients with depression. In one of these, citalopram appeared to produce a more rapid onset of effect as measured by both the MADRS and the CGI score. In the other study, citalopram (40 to 60 mg/day) and mianserin (60 to 90 mg/day) were equally effective in patients with endogenous depression, but mianserin was more effective among nonendogenously depressed patients.
A retrospective meta-analysis of 5 trials comparing citalopram (30 to 60 mg) with tricyclic antidepressants (amitriptyline, clomipramine, nortriptyline and amitriptyline) showed that, on the basis of a 50% reduction in HAM-D total scores, the efficacy of citalopram was similar to that of tricyclic antidepressants.
Comparative efficacy of citalopram versus other SSRIs:
The efficacy of citalopram was compared with that of fluoxetine in two 8-week trials among psychiatric patients and among patients from general medical practice. In the psychiatrist-based trial, citalopram (40 mg) and fluoxetine (20 mg) were equally efficacious, as measured by reductions in MADRS, HAM-D and CGI-S assessments. Citalopram was found to be significantly more efficacious than fluoxetine for severely depressed patients with HAM-D scores of at least 25 (p=0.0003). Similar outcomes were found in the general medical practice setting, where the two treatments were similarly efficacious; except for a subgroup of patients not receiving benzodiazepines, the mean reduction in total MADRS scores favoured citalopram at 2, 4, 6 and 8 weeks.
Two multicentre trials comparing citalopram with sertraline found both to be similarly effective in producing a meaningful reduction in total MADRS score, although one study found citalopram to be associated with anti-anxiety effects.
A 6-week, double-blind multicentre study found citalopram (20 to 40 mg/day) and fluvoxamine to be similarly effective. Complete plus partial response rates (based on HAM-D scores) were 42 and 39% in the citalopram and fluvoxamine groups, respectively. CGI and Zung self-rating depression scores provided similar results with regard to efficacy.
Safety and tolerability:
In comparison with placebo, citalopram was shown to be safe and well- tolerated in both the short- and long-term. The side-effects observed in patients treated with citalopram included somnolence, nausea, dry mouth, increased sweating and short-term anorgasmia. One study also found that asthenia, tiredness, lassitude and emotional indifference occurred significantly more frequently in elderly patients treated with citalopram, than in those receiving placebo.
As with other SSRIs, citalopram is generally better tolerated than the tricyclic antidepressants. In particular, citalopram appears to produce fewer and less severe central nervous system and anticholinergic effects. The available data suggest that the tolerability profile of citalopram is similar to that of other SSRIs.
Evidence included in the review suggested that citalopram does not have any adverse effects in terms of psychomotor function or cardiovascular safety, and is safe in overdose.