Seventeen RCTs (n=2,252) were included in the review. At least 4 studies were of a crossover design, with washout periods ranging from 4 to 14 days. A further 8 studies reported using a placebo lead-in period of one week.
TCAs (11 RCTs).
Five studies of people over 60 years of age reported that imipramine showed a statistically significant superiority to placebo using the HAM-D (P<0.05), but not to the other antidepressant agents, i.e. trazodone (1 study), fluvoxamine (1 study), buspirone (1 study) and nomifensine (2 studies).
One study of people over 55 years of age reported that imipramine was superior to placebo. Two studies compared imipramine with placebo and bupropion. One study reported imipramine as superior to placebo, whereas the other reported no difference in HAM-D scores; neither reported a difference between imipramine and bupropion. Three studies reported a statistically significant reduction in HAM-D with nortriptyline compared with placebo (P<0.05). The NNT to gain at least a 50% reduction in HAM-D with TCAs (1 study) was 5 (95% CI: 3, 9).
MAO inhibitors (1 RCT).
Phenelzine was reported as being as effective as nortriptyline, and superior to placebo in treating depression.
SSRIs (3 RCTs).
One study comparing fluoxetine with placebo reported no difference in HAM-D when a 10 to 50% reduction was sought, whereas another reported a statistically significant reduction in HAM-D with fluoxetine when a greater than 50% reduction was sought. One study reported a statistically significant improvement in HAM-D with sertraline compared with placebo. The combined NNT with SSRIs to gain at least a 50% reduction in HAM-D was 8 (95% CI: 5, 11).
Other antidepressants.
Bupropion was found to be superior to placebo in one study, but not in another. Mirtazapine (1 study) and nomifensine (1 study) significantly lowered HAM-D scores compared with placebo.
Further results for the additional secondary outcome scales were also presented.