Does TMS work for depression? What the pooled trial evidence shows

Several forms of transcranial magnetic stimulation roughly triple the odds of responding to treatment for a major depressive episode compared with a sham procedure, according to the largest network meta-analysis of brain stimulation to date [s1]. But those techniques sit below electroconvulsive therapy on efficacy, and most of the trials behind them are small and at unclear risk of bias [s1].

That is the honest shape of the evidence for a treatment that is now widely marketed, and it is more qualified than most clinics selling it will say.

What was pooled

The analysis brought together 113 randomised trials — 262 treatment arms in all — covering 6,750 patients with major depressive disorder or bipolar depression, mean age 47.9 years, 59% of them women [s1]. It compared electroconvulsive therapy, several kinds of transcranial magnetic stimulation (repetitive, accelerated, priming, deep and synchronised), theta-burst stimulation, magnetic seizure therapy and transcranial direct-current stimulation, each against sham [s1]. The primary outcomes were response and all-cause discontinuation — a measure of how tolerable a treatment is — reported as odds ratios [s1].

Ten of the eighteen treatment strategies were associated with higher response than sham [s1]. The form most people mean by "TMS" — high-frequency left repetitive TMS, the standard clinical protocol — had an odds ratio of 3.17 (95% confidence interval 2.29 to 4.37) [s1]. Bilateral rTMS came out higher, at 4.92 (2.93 to 8.25), as did low-frequency right rTMS at 3.65 (2.13 to 6.24) and intermittent theta-burst stimulation at 3.20 (1.45 to 7.08) [s1]. Transcranial direct-current stimulation, a different and gentler technology, was 2.65 (1.55 to 4.55) [s1].

Where TMS sits against ECT

The ceiling belonged to electroconvulsive therapy. Bitemporal ECT had the highest response odds of any strategy, 8.91 (2.57 to 30.91), and high-dose right unilateral ECT was 7.27 (1.90 to 27.78) [s1]. In the head-to-head contrasts within the network, those two ECT forms were the treatments associated with increased response over other active interventions [s1]. So the pooled evidence does not make TMS the equal of ECT; it makes TMS an effective option that is easier to deliver and does not require anaesthesia or induce a seizure.

On tolerability the picture is reassuring: all the treatment strategies were at least as acceptable as sham, meaning people did not drop out of the active treatments more than out of the placebo one [s1].

A faster protocol that held up head to head

One question the network could only hint at has since been tested directly. Standard high-frequency rTMS takes about 37.5 minutes per session; a newer variant, intermittent theta-burst stimulation, can be delivered in 3 minutes [s2]. The THREE-D trial randomised adults with treatment-resistant depression — 205 to 10 Hz rTMS and 209 to theta-burst — and found the two equally effective: Hamilton depression scores fell from 23.5 to 13.4 in the rTMS group and from 23.6 to 13.4 in the theta-burst group, meeting the trial's non-inferiority bar (adjusted difference 0.103; p=0.0011) [s2]. Side-effects and dropout were similar, though self-rated pain during treatment was slightly higher with theta-burst [s2]. The practical upshot is that a clinic can treat many more patients a day without a measured loss of benefit — the kind of finding that reshapes access more than efficacy.

The caveat that runs through all of it

The quality of the underlying trials was modest. Ninety-four of the 113 trials — 83% — were rated at low or unclear risk of bias, and the precision of the summary estimates varied considerably, as the wide confidence intervals on the newer techniques show [s1]. The most-studied comparisons were high-frequency left rTMS and direct-current stimulation against sham; the newer treatments, the authors note, remain understudied [s1]. That is why an odds ratio can look impressive and still rest on thin ground.

This matters for how a reader should read a clinic's claims. The evidence supports considering non-surgical brain stimulation as an alternative or add-on treatment for a major depressive episode [s1]. It does not support a specific device as clearly best, and it flags the shortage of well-designed trials comparing the newer methods directly as a research priority rather than a settled question [s1].

What it does and does not settle

It settles that TMS, in its established forms, beats a convincing sham on response and is tolerated about as well [s1]. It does not settle which TMS protocol is best, how durable the benefit is beyond the acute phase, or how the technique compares with medication or psychotherapy — questions the network, built on short-term response data, was not designed to answer.

A separate strand of research argues the remaining gains lie less in the choice of machine than in precision — where the pulse is aimed — and in gentler alternatives to ECT. And for many people the first comparison that matters is not between stimulation devices but between stimulation and the treatments they would otherwise try first. This article describes evidence and is not medical advice.

Sources

• Comparative efficacy and acceptability of non-surgical brain stimulation for the acute treatment of major depressive episodes in adults: systematic review and network meta-analysis — BMJ, 2019-03-27
• Effectiveness of theta burst versus high-frequency repetitive transcranial magnetic stimulation in patients with depression (THREE-D): a randomised non-inferiority trial — The Lancet, 2018-04-26