A new review of antidepressant withdrawal effects – written by academics, many of whom have close ties to drug manufacturers – risks underestimating the potential harms to long-term antidepressant users by focusing on short-term, industry-funded studies.
There is growing recognition that stopping antidepressants – especially after long-term use – can cause severe and sometimes debilitating withdrawal symptoms, and it is now acknowledged by the UK government as a public health issue.
One of the main reasons this issue took decades to recognise after the release of modern antidepressants onto the market is because medical guidelines, such as those produced by Nice (England's National Institute for Health and Care Excellence), had for many years declared withdrawal effects to be "brief and mild".
This description was based on studies run by drug companies, where people had only taken the medication for eight to 12 weeks. As a result, when patients later showed up with severe, long-lasting symptoms, many doctors didn't take them seriously because these experiences contradicted what the guidelines led them to expect.
Related: Some Antidepressants Could Speed Decline in Dementia Patients
Our recent research helps explain this mismatch. We found a clear link between how long someone takes antidepressants and how likely they are to experience withdrawal symptoms – and how severe these symptoms are.
We surveyed NHS patients and found that people who had used antidepressants for more than two years were ten times more likely to have withdrawal effects, five times more likely for those effects to be severe, and 18 times more likely for them to be long lasting compared with those who had taken the drugs for six months or less.
For patients who used antidepressants for less than six months, withdrawal symptoms were mostly mild and brief. Three-quarters reported no or mild symptoms, most of which lasted less than four weeks.

Only one in four of these patients was unable to stop when they wanted to. However, for long-term users (more than two years), two-thirds reported moderate or severe withdrawal effects, with one-quarter reporting severe withdrawal effects.
Almost one-third of long-term users reported symptoms that lasted for more than three months. Four-fifths of these patients were unable to stop their antidepressants despite trying.
About 2 million people on antidepressants in England have been taking them for over five years, according to a BBC investigation. And in the US at least 25 million people have taken antidepressants for more than five years. What happens to people in eight-to-12-week studies is a far cry from what happens to millions of people when they stop.
Studying what happens to people after just eight to 12 weeks on antidepressants is like testing car safety by crashing a vehicle into a wall at 5km/h – ignoring the fact that real drivers are out on the roads doing 60km/h.
History repeating itself?
Against this backdrop, a review has just been published in JAMA Psychiatry. Several of the senior authors declare payments from drug companies.
In what looks like history repeating itself, the review draws on short-term trials – many funded by the pharmaceutical industry – that were similar to those used to shape early treatment guidelines. The authors conclude that antidepressants do not cause significant withdrawal effects.
Their main analysis is based on eleven trials that compared withdrawal symptoms in people who had stopped antidepressants with those who had continued them or stopped taking a placebo. Six of these trials had people on antidepressants for eight weeks, four for 12 weeks and just one for 26 weeks.
They reported a slightly higher number of withdrawal symptoms in people who had stopped antidepressants, which they say does not constitute a "clinically significant" withdrawal syndrome. They also suggest the symptoms could be explained by the "nocebo effect" – where negative expectations cause people to feel worse.
In our view, the results are likely to greatly underestimate the risk of withdrawal for the millions of people on these drugs for years. The review found no relationship between the duration of use of antidepressants and withdrawal symptoms, but there were too few long-term studies to test this association properly.
The review probably underestimates, in our view, short-term withdrawal effects too by assuming that the fact that people experience withdrawal-like symptoms when stopping a placebo or continuing an antidepressant cancels out withdrawal effects from antidepressants. But this is not a valid assumption.
We know that antidepressant withdrawal effects overlap with side-effects and with everyday symptoms, but this does not mean they are the same thing. People stopping a placebo report symptoms such as dizziness and headache, because these are common occurrences.
However, as was shown in another recent review, symptoms following discontinuation of a placebo tend to be milder than those experienced when stopping antidepressants, which can be intense enough to require emergency care.
So deducting the rate of symptoms after stopping a placebo or continuing an antidepressant from antidepressant withdrawal symptoms is likely to underestimate the true extent of withdrawal.

The review also doesn't include several well-designed drug company studies that found high rates of withdrawal symptoms. For example, an American study found that more than 60% of people who stopped antidepressants (after eleven months) experienced withdrawal symptoms.
The authors suggest that depression after stopping antidepressants is probably a return of the original condition, not withdrawal symptoms, because similar rates of depression were seen in people who stopped taking a placebo.
But this conclusion is based on limited and unreliable data (that is, relying on participants in studies to report such events without prompting, rather than assessing them systematically) from just five studies.
We hope uncritical reporting of a review based on the sort of short-term studies that led to under-recognition of withdrawal effects in the first place, does not disrupt the growing acceptance of the problem and slow efforts by the health system to help potentially millions of people who may be severely affected.
The authors and publisher of the new review have been approached for comment.
Mark Horowitz, Visiting Clinical Research Fellow in Psychiatry, UCL and Joanna Moncrieff, Professor of Critical and Social Psychiatry, UCL
This article is republished from The Conversation under a Creative Commons license. Read the original article.