D-cycloserine with exposure therapy

At a glance

Compared with other treatments

Against a placebo pill, both with group CBT: no clear difference in how many people responded (79% vs 73%) or recovered (35% vs 24%, a gap that could be chance).

What it takes

Only used in research. A pill taken shortly before some therapy sessions.

What it is

D-cycloserine (DCS) is an old antibiotic, used for tuberculosis. In animal studies, low doses helped the brain learn that something once feared is now safe. Researchers hoped it could do the same in people.

The plan was never to use it on its own. You would take a single low dose an hour before an exposure session (facing a feared situation on purpose, in small steps). The pill was meant to make the lesson of that session stick better.

Why people were excited

Exposure therapy works, but it takes time and effort, and some people drop out. A pill that made each session count for more would mean shorter therapy and better results. Early small trials in social anxiety looked promising, and the idea got a lot of attention.

What the evidence shows

• The largest social anxiety trial found no difference at the end. In a US trial of 169 adults with social anxiety, everyone had 12 sessions of group cognitive behavioural therapy (CBT). Before 5 of the sessions, some took 50 mg of DCS and some a placebo pill. People on DCS improved 24% to 33% faster. But by the end, about the same share had responded: 79% with DCS and 73% with placebo. Recovery was 35% with DCS and 24% with placebo, a gap the trial could not tell apart from chance.
• Across many trials, the boost was small. A study that pooled the data of 1,047 people from 21 trials, in several anxiety problems, OCD and PTSD, found a small benefit from DCS at the end of therapy. At follow-up the benefit was even smaller and no longer clear. The authors could not find a group of people for whom it clearly worked better.

What the trials have not answered

• Would a different way of using it work? The analysis of the same trial suggests DCS might help if given only after a session goes well. That idea has not been tested properly in social anxiety.
• Does it help people who struggle most? The pooled study could not find any group of people who clearly benefited more.

Why it may have faded

An analysis of the same trial found that DCS seemed to help only after sessions that went well, when fear was low by the end. After sessions where fear stayed high, people on DCS did slightly worse at the next session. These are possible reasons for the fading, not proven ones.

What this means for you

DCS is not a treatment you can ask for. It is not used for social anxiety outside research.

The useful lesson is about exposure itself. The group CBT that both groups received in the largest trial helped a large share of people. Most of the change came from the therapy, not the pill.

Why this story is worth knowing

DCS is a good example of how science corrects itself. Early small trials looked exciting. The largest trial found no difference in the end result, and pooled data showed only a small effect. The same pattern has happened with attention training, oxytocin and, so far, the nasal spray fasedienol. When you read about a new pill that “boosts therapy”, remember this story and look for the large trials.

Side effects

DCS is a prescription antibiotic. At the higher daily doses used for tuberculosis, it can affect mood and the nervous system. It should never be taken without a doctor.