Research Review: Transcranial Magnetic Stimulation (TMS) for OCD

Approximately 10% of individuals diagnosed with OCD don’t get better with traditional treatment. Researchers are always identifying new strategies for these “treatment resistant” cases, one of which is called transcranial magnetic stimulation (TMS).

The following article reviews the literature:

Blom, R. M., Figge, M., Vulink, N., & Denys (2011). Update on Repetitive Transcranial Magnetic Stimulation in Obsessive-Compulsive Disorder: Different Targets. Current Psychiatry Reports, 13,  289–294. doi:  10.1007/s11920-011-0205-3

TMS was developed in the early 80s. It is a technique of using a machine to send a magnetic signal to the brain. In TMS, the signal is strong enough that it actually depolarizes the neurons – that is, it causes them to “fire” and send messages to other parts of the brain. Once TMS causes a neuron to fire there is a small time period where it won’t fire again (it is “hyper-polarized”). The trick with TMS treatment is to fire the neuron many times (called repetitive TMS) to cause that hyper-polarization time period to last longer. Do this every day and rTMS stops that part of the brain from working. Why would this help OCD? The theory is that OCD is related to overactive parts of the brain, especially the prefrontal subcortical areas. As such, stopping this part of the brain being as active may reduce both obsessions and compulsions. Cool. But does it work?

The authors review 10 different studies, of varying quality. In short, rTMS and placebo treatments work as well as each other, but both do better than nothing at all.

Those individuals receiving rTMS actually show improvement – typically up to 1 month after finishing treatment. In particular, the compulsions seem to be most reduced, and this effect lasts around 8 hours following the treatment – at which time the brain neurons are presumably acting fully again. However, in the more well-designed studies with randomization, placebo, and double-blinding, the placebo groups show equal improvement to the rTMS groups. Placebo works as well as both “high frequency” and “low frequency” rTMS. As such, rTMS actually doesn’t do better than a placebo effect.  In both groups improvement in symptoms is generally around the 20-30% mark up to 1 month after treatment – What happens beyond this time period is a question for future research. These findings are also supported by the Cochrane 2015 report.

The review article was really interesting, and their methodology and results seem clear and well presented. However, their discussion leaves a lot of questions to be answered. I was disappointed that they didn’t identify weaknesses in the literature. First, the pre-treatment Y-BOCS scores (their OCD severity measure) varied wildly across studies. It’s possible that severity of OCD may affect results. To this point, it is really unclear how the research defines “treatment resistant” OCD-They often say that patients have failed a course of treatment before doing rTMS, but who knows what treatment they failed?!  Second, I’m very curious what happens longer-term, and I’m surprised that most of these studies don’t have longer follow-ups – Or they did but it wasn’t reported. I think they should have highlighted this flaw. Third, I wish they had proposed why they think the placebo effect is equal to the treatment, as this repeated finding makes me question the entire mechanism of action – basically, we have no idea what we are doing! Not discussing this in more detail is a major problem.

The biggest omission though, was that the authors did not specify or discuss whether rTMS occurs alongside ongoing psychological treatment, so I assume that it doesn’t. Most studies allow ongoing psychiatric medication use. As such, I’m not really that surprised that rTMS doesn’t work. Surely you want to have ongoing treatment while undergoing rTMS?! From a research perspective, having participants also undergo therapy would not be ideal, but at the same time this is waaaaaay more likely to help patients. In particular, if it therapy can occur after a rTMS session and within the 8 hour window post-rTMS, a patient may be able to engage in treatment with significantly fewer obsessions and compulsions.  Specifically, they should theoretically be able to engage in exposures and have an easier time with response prevention (i.e. not doing their compulsions). To think of this differently, imagine giving a weight lifter a powder one hour before training that allows them to lift more weights while training – so they can be stronger come competition. Makes sense. However, if you just gave the athlete the powder but they didn’t workout, you wouldn’t expect them to be any stronger afterwards, right? To me, doing studies of rTMS without treatment is like giving a weight lifter a muscle building powder without exercise. You’re missing the whole point. I suspect there is promise for rTMS if done concurrently with exposure treatment but I don’t know because it’s not specified in the research and the authors don’t discuss this. By leaving this out, it further reinforces the idea that rTMS is a treatment in of itself, which I don’t think the research supports.

Since this article and other research, the field has moved on a little to studying the effects of “deep” rTMS. In regular rTMS the magnetic field affects neurons around 3-4cm below the skin whereas in deep TMS the pulse can go around 5-6cm below the scalp. People theorized that rTMS will be more effective if it can penetrate deeper into the brain and so deep rTMS was developed. However, the deeper the TMS signal, the less precise it becomes, meaning there is potential for less specificity in the brain areas it targets. There is initial research supporting that deep rTMS works for OCD, but I am very skeptical for all the same reasons as discussed above, and because there are no studies yet to include control groups. Let’s take a recent article:

Modirrousta, M., Shams, E., Katz, C., Mansouri, B., Moussavi, Z., Sareen, J., & Enns, M. (2015). The Efficacy of Deep Repetitive Transcranial Magnetic Stimulation Over the Medial Prefrontal Cortex in Obsessive Compulsive Disorder: Results from an Open-Label Study. Depression and Anxiety, 32, 445-450.

What They Did:

The authors recruited 10 participants for the study through posters and referrals. All were diagnosed with OCD and had “failed to respond” to at least one treatment. They all received 10 sessions of deep rTMS (1200 pulses per session) over 2 weeks, and their OCD symptoms were measured 1 week before treatment, on the same day as the last treatment (immediately after the last trial), and 1 month following the last treatment. The average decrease in OCD symptoms after 1 month was ~39%, which was statistically significant.

The Strengths of the Study:

The finding that symptoms remain low for one month post-treatment is promising. The authors also accounted for depression symptoms in their results, which is great because some rTMS research shows that OCD symptoms only improve as a result of depressive symptoms improving. The study is also quite a neat and simple design.

Weaknesses of the Study:

The researchers correctly identify that the lack of a control group is the biggest weakness. It means that they have no idea if the patients got any better than people who did not do the treatment, did a fake treatment (placebo), or did some other treatment. As such, there is no way to know if this deep rTMS works better than anything else. However, they argue that OCD patients are “largely placebo non responsive”, quoting a study from 2004. To me, this seems like pick and choosing the literature (or lack of research on the topic) – It is quite clear from the rTMS OCD research that placebo works the same as rTMS.

Another major issue to me is that the lead author was the one who tested the patients after treatment and gave scores on the severity of OCD symptoms – the measure of OCD was the gold standard, but it’s also a measure that the clinician administers and scores. Doing the study this way means that the results are likely influenced by the lead author’s potential bias. Since the author knew they were doing the study and what they wanted to find, so I suspect the results were influenced – thus the improvement shown is probably an exaggeration.


Overall the research basically shows that rTMS doesn’t work any better than a fake treatment. That is very concerning, and to me suggests that we have no real understanding of what we are doing. Since the literature has settled on this conclusion, the field has now moved to testing “deep” rTMS. While it’s possible that this “better” technology could make a difference, the skeptic in me doubts it. It also seems like unprovable science; that is, companies develop something that, by the time it has been tested in the research, they have already modified and developed a newer version. This makes the procedures essentially untestable because it’s always evolving, which is very concerning – especially since it’s clear from the research and older technology that we don’t really understand the mechanism of action! I’m not against further research on this topic, but rather than trying to research rTMS as a treatment in and of itself, I think the field would be better suited by looking at the interaction of rTMS and psychological treatment – namely exposure and response prevention. To test this, research might compare rTMS, to rTMS + therapy, to placebo rTMS + therapy, to placebo, to nothing at all. You can see how this might take a lot of people!

Thanks for reading, please leave any comments of thoughts.


Leave a Reply