Schulze-Rauschenbach et al (Reference Schulze-Rauschenbach, Harms and Schlaepfer2005) report an interesting study comparing electroconvulsive therapy (ECT) with repetitive transcranial magnetic stimulation (rTMS) for the treatment of major depression. They conclude that both are equally efficacious, but rTMS is associated with fewer cognitive side-effects.

Ultimately, a therapeutic plan that optimally helps a given individual cope with disability and recover can only be devised when the physician transcends the traditionally proven, generic insights about pathological anatomy, physiopathology and aetiology of disease, and identifies the unique characteristics of a specific patient. Therefore, therapeutic decisions should always consider the risk–benefit ratio of each treatment option for the specific patient and circumstances. The focus on differences in side-effect profile of rTMS and ECT in major depression by Schulze-Rauschenbach et al is thus most valuable.

In 2003 we conducted a similar study of which the authors appear to be unaware (Reference O'Connor, Brenninkmeyer and MorganO'Connor et al, 2003). We compared 14 patients with medication-refractory depression who underwent ECT with 14 who underwent rTMS. ECT had a significantly greater positive effect on mood than a 2-week trial of rTMS. The reason for this difference between our results and other published trials comparing rTMS with ECT (Reference Janicak, Dowd and MartisJanicak et al, 2002; Reference Grunhaus, Schreiber and DolbergGrunhaus et al, 2003; Reference Schulze-Rauschenbach, Harms and SchlaepferSchulze-Rauschenbach et al, 2005) is unclear. In our study, as in that of Schulze-Rauschenbach et al, ECT was applied unilaterally approximately three times per week for 2–4 weeks. We applied rTMS in sessions of 1600 stimuli at 10 Hz and 90% of motor threshold intensity to the left dorsolateral prefrontal cortex daily (Monday through Friday) for 2 consecutive weeks. Thus we employed ‘stronger’ parameters of rTMS than Schulze-Rauschenbach et al, who applied shorter trains and also limited stimulation to only 2 weeks. Nevertheless, in both studies ECT and rTMS may have been used at insufficient doses, since progression to bilateral ECT (UK ECT Review Group, 2003) or extension of daily rTMS to 3–4 weeks (Reference Gershon, Dannon and GrunhausGershon et al, 2003; Reference Rumi, Gattaz and RigonattiRumi et al, 2005) were not considered.

In our study ECT exerted a deleterious but transient effect on various components of memory that was no longer detected 2 weeks after the end of treatment. However, there was evidence of persistent retrograde amnesia after ECT. Patients undergoing rTMS experienced only a modest reduction in the severity of depression but there was no evidence of anterograde or retrograde memory deficits and there was a remarkable suggestion of cognitive improvement even in those patients with no antidepressant benefits. These findings, as those of Schulze-Rauschenbach et al, suggest that the cognitive effects of rTMS might not be the consequence of the mood effects. The suggestion of independent effects of rTMS on mood and cognition also seems to be supported by a previous study of rTMS in major depression (Reference Moser, Jorge and MonesMoser et al, 2002) and studies in patients with cerebrovascular (Reference Rektorova, Megova and BaresRektorova et al, 2005) and Parkinson's disease (Reference Boggio, Fregni and BermpohlBoggio et al, 2005). Boggio et al (Reference Boggio, Fregni and Bermpohl2005) showed that 10 days of rTMS treatment (15 Hz, left dorsolateral prefrontal cortex) improved cognition and depression in patients with Parkinson's disease, but this cognitive improvement was not correlated with mood change. Furthermore, there was no correlation between cognitive and motor function improvement. Thus, it appears that left prefrontal rTMS exerts differential effects on cognition, mood and motor function. Even in individuals without psychiatric illness, we have recently shown that suppression of the right hemisphere by slow rTMS can enhance verbal memory, while left-sided slow rTMS disrupts it (Reference Kahn, Pascual-Leone and TheoretKahn et al, 2005). Therefore, cognitive and antidepressant effects of rTMS may be the consequence of modulation of dissociable neural networks.