I haven’t read this study, but after years of treating patients, I’ve no doubt that discontinuation syndrome is a major issue, most often depending on which antidepressant was used. It can be a serious problem for the patient. Often can be mitigated by slow taper or adding one with long half life for 2-3 weeks, then discontinuing both. If this study lumped all antidepressants together, the effect may be lost in the weeds.
It would be interesting to study (perhaps it's already been done)....the difference in male/female hormones as an influence on depression. I also think it would be interesting to study the age at which depression actually begins....to see if depression is the result of early childhood experiences...from birth....familial as well as hormonal...in girls vs. boys. And...of course....the difference in all early experiences in those who do suffer from depression vs. those who do not.
Calling any of this "science" is a serious misnomer. There are many obvious problems with attempting to quantify and compare subjective symptoms in a group of patients placed on medications for largely subjective symptoms. Doing a meta-analysis on all this just compounds the problems.
The meta analysis showed what it showed. Following generally short duration of use of these various SSRI’s, there was relatively little withdrawal symptoms during very short term (1-2 week) follow up.
It does very little to address issues with long term users, because few of the included studies had such a cohort. And it doesn’t examine symptom prevalence farther removed from med discontinuation.
So this may be part of the puzzle but does not dispel the “withdrawal” phenomenon in anywhere close to its entirety.
If this paper were legitimate, sice most trials were short term, then the fact relapse allegedly did not occur would indicate that long-term drug treatment should not be a thing. Either way, the field has some 'splaining to do.
I have read that often the withdrawal symptoms are mistaken for the return of the original pathology and so the patient is rapidly returned on medication. If (a big if) it is true it is dramatic. I also know many people who have tried many times to taper off but since this must be done really really gradually (possible to cut the pills but it seems that they are very sensible to oxydation) it is quite difficult. I am not impressed by the study protocol.
I am a fan of John's work in general. But he has been badly misled by this review. The main conclusions from this study were based on 8-12 week studies, nothing like the years or decades people use these drugs for. There is a clear association between duration of use and risk of withdrawal. This was an attempt to mislead and minimise a serious public health issue causing embarrassment to establishment academics who have been reassuring the public and doctors for decades that there is no issue. It is reminiscent of a similar tactic for opioids where Purdue highlighted 'low rates of addiction' based on poorly conducted studies.
I'm a psychiatrist seeing 200+ patients a week, doing pharmacology only, and I can tell you that study doesnt bring anything new or reasuring to the field
Let me know if you're interested in a more extensive write up
It's just a way of disclosing my own bias in favor of psychopharmacology - but even with that bias, I can see there are a few things that went wrong with this study
The clinical literature is of very low quality.Very few RCTs , the last in the 1980s. Almost all “ studies” had no follow up beyond one month
The rest of the publications are case series - with only the psychiatrist report of “ success” . No patient follow up
ECT machines have never been through FDA approval - they were grandfathered in decades ago on the say so of the psychiatrist who owned the company that manufactures the machines!!!
ECT in practice today in America has no monitoring and no reporting
There is no follow up of the patients , regardless of where the ECT is done( teaching hospital vs small for profit free standing psych facilities ( UHS-United Hospital Systems owns hundreds of these throughout the country)
When ppl who are severely harmed by ECT try to report the problems( severe permanent retrograde amnesia and cognitive dysfunction and more) they are told” ECT doesn’t do that it’s your underlying diagnosis “ denigrated and abandoned .
First, the meta-analysis in question involved 50 trials, 90% studying people on antidepressants for <6 months (mostly at 8-12 weeks). In practice that's rare. People aiming to discontinue have usually been on them for months to years, making 90% of the studies functionally irrelevant. Second, the authors acknowledge withdrawal symptoms are statistically much more common with antidepressants than placebo— but then dismiss it based on a double misunderstanding: a) they use an unvalidated, made-up threshold for the 'minimal clinically important difference' (based on two references) on which they base their conclusions, AND b) they fail to realize the distinction between relapse and withdrawal 'syndromes' was the true goal when earlier authors used the threshold. Indeed, without knowing it they demonstrate that withdrawal is not only real, it is common and clinically quite painful (often as bad or worse than MDD, by their calculations)—even in the largely irrelevant short term studies they include. Unfortunately, they misinterpret their own data because of the above mistakes, and conclude the difference was not 'clinically significant'. Unfortunate.
Hope you have an open mind, and perhaps will look for yourself. If so, let me know how you think I have it wrong.
The recent publication in JAMA Psychiatry: Incidence and Nature of Antidepressant Discontinuation Symptoms: A Systematic Review and Meta-Analysis made me feel that I had been here before.
Once again, a group of influential doctors and academics have put their names to a paper that claims to have clarified the incidence of “discontinuation symptoms” yet at the same time admits to including predominantly short term studies. Efforts have also been made to disseminate the JAMA Psychiatry paper to worldwide media. This has given some of the authors opportunities to suggest that clinical guidelines should be modified and to ignore the experience of huge numbers of people on antidepressants.
I wrote a post [full text below] in January 2024 which explored the relationship between opinion leaders and the pharmaceutical industry. It is interesting to find that a number of the authors of the recently published JAMA Psychiatry paper feature in the timeline that I compiled at that time [see below]. Perhaps it is the recent publicity generated by Reports from inquiries into both the Post Office and infected blood scandals that has led me to reflect on the reasons why those in positions of power continue to do whatever they can to protect the reputations of their institutions despite growing evidence that something is wrong.
Personal anecdotes - 2 family members and 1 close friends have all in the last 3 years suffered greatly.
1. Two on discontinuation, one
merely tapered, all got suicidal thoughts
2. One poured boiling water over their arm etc
3. One was in and out of the A&E multiple times with suspected heart issues
4. All 3 ended up back on original doses in fact 2 of them ended up on more drugs than what they started with to “get stable”
5. For one it took over 6 months to get back some level of “normality” and she had only been on the pills as migraine prevention to start with (for many many years)
Personally I’ve seen too much in real life causing distrust in majority of doctors when it comes to withdrawals.
Even pretty “benign” cardiac pharmaceuticals have caused significant withdrawals for my husband although we are aware and always try to taper slowly.
I haven’t read this study, but after years of treating patients, I’ve no doubt that discontinuation syndrome is a major issue, most often depending on which antidepressant was used. It can be a serious problem for the patient. Often can be mitigated by slow taper or adding one with long half life for 2-3 weeks, then discontinuing both. If this study lumped all antidepressants together, the effect may be lost in the weeds.
It would be interesting to study (perhaps it's already been done)....the difference in male/female hormones as an influence on depression. I also think it would be interesting to study the age at which depression actually begins....to see if depression is the result of early childhood experiences...from birth....familial as well as hormonal...in girls vs. boys. And...of course....the difference in all early experiences in those who do suffer from depression vs. those who do not.
Calling any of this "science" is a serious misnomer. There are many obvious problems with attempting to quantify and compare subjective symptoms in a group of patients placed on medications for largely subjective symptoms. Doing a meta-analysis on all this just compounds the problems.
Were there any conflicts of interest from any of the authors of the studies used in the meta-analysis? Also the study sponsors would be of interest.
The meta analysis showed what it showed. Following generally short duration of use of these various SSRI’s, there was relatively little withdrawal symptoms during very short term (1-2 week) follow up.
It does very little to address issues with long term users, because few of the included studies had such a cohort. And it doesn’t examine symptom prevalence farther removed from med discontinuation.
So this may be part of the puzzle but does not dispel the “withdrawal” phenomenon in anywhere close to its entirety.
If this paper were legitimate, sice most trials were short term, then the fact relapse allegedly did not occur would indicate that long-term drug treatment should not be a thing. Either way, the field has some 'splaining to do.
I have read that often the withdrawal symptoms are mistaken for the return of the original pathology and so the patient is rapidly returned on medication. If (a big if) it is true it is dramatic. I also know many people who have tried many times to taper off but since this must be done really really gradually (possible to cut the pills but it seems that they are very sensible to oxydation) it is quite difficult. I am not impressed by the study protocol.
Millions of prescriptions for SSRIs are given to ppl for non psychiatric reasons( migraines, insomnia, neuropathy, fibromyalgia and others)
Many of these ppl experience the same debilitating withdrawal problems
Thieve drugs were marketed under the false “ chemical imbalance “ assertion
They hid the PSSD problems
They were largely studied for short
There is mounting evidence that SSRIs are deleterious to a developing fetus
One could easily make the case that these drugs should be much more tightly regulated or even pulled off the market for deceptive practices
I am a fan of John's work in general. But he has been badly misled by this review. The main conclusions from this study were based on 8-12 week studies, nothing like the years or decades people use these drugs for. There is a clear association between duration of use and risk of withdrawal. This was an attempt to mislead and minimise a serious public health issue causing embarrassment to establishment academics who have been reassuring the public and doctors for decades that there is no issue. It is reminiscent of a similar tactic for opioids where Purdue highlighted 'low rates of addiction' based on poorly conducted studies.
https://theconversation.com/antidepressant-withdrawal-new-review-downplays-symptoms-but-misses-the-mark-for-long-term-use-260708
Well said! The people who've been on SSRIs for years should not be 'reassured' that agomelatine has no withdrawal syndrome...
I'm a psychiatrist seeing 200+ patients a week, doing pharmacology only, and I can tell you that study doesnt bring anything new or reasuring to the field
Let me know if you're interested in a more extensive write up
France must have shorter appointments than the US.
That is not scientific evidence
Of course it isnt
It's just a way of disclosing my own bias in favor of psychopharmacology - but even with that bias, I can see there are a few things that went wrong with this study
“Good news Mr. Johnson! Those debilitating brain zaps you are experiencing are not clinically significant!”
SSRIs do not "zap" brains, and ECT is an essential treatment for severe depression
ECT is wholly unscientific
There is zero basic science to support it
The clinical literature is of very low quality.Very few RCTs , the last in the 1980s. Almost all “ studies” had no follow up beyond one month
The rest of the publications are case series - with only the psychiatrist report of “ success” . No patient follow up
ECT machines have never been through FDA approval - they were grandfathered in decades ago on the say so of the psychiatrist who owned the company that manufactures the machines!!!
ECT in practice today in America has no monitoring and no reporting
There is no follow up of the patients , regardless of where the ECT is done( teaching hospital vs small for profit free standing psych facilities ( UHS-United Hospital Systems owns hundreds of these throughout the country)
When ppl who are severely harmed by ECT try to report the problems( severe permanent retrograde amnesia and cognitive dysfunction and more) they are told” ECT doesn’t do that it’s your underlying diagnosis “ denigrated and abandoned .
ECT is barbaric it should be banned
Read the Castleman report 2018
https://www.madinamerica.com/wp-content/uploads/2021/01/3Castleman_2018_Expert-Report.pdf
Brain zaps have nothing to do with ECT Freddie. What on EARTH.
Exactly! Electrical shocks was the exact words my child described to me trying to go of his SSRI's.
I expected better from Vinay Prasad...
Highly recommend: https://blog.maryannedemasi.com/p/antidepressant-withdrawalwhy-do-researchers. Nicely summarizes why the SRMA you refer to is neither patient-centered nor externally valid for an overwhelming majority of patients who take antidepressants.
Not remotely convincing, and you only think it is because that was your preferred conclusion to begin with.
Sorry to hear you're not convinced. I am.
First, the meta-analysis in question involved 50 trials, 90% studying people on antidepressants for <6 months (mostly at 8-12 weeks). In practice that's rare. People aiming to discontinue have usually been on them for months to years, making 90% of the studies functionally irrelevant. Second, the authors acknowledge withdrawal symptoms are statistically much more common with antidepressants than placebo— but then dismiss it based on a double misunderstanding: a) they use an unvalidated, made-up threshold for the 'minimal clinically important difference' (based on two references) on which they base their conclusions, AND b) they fail to realize the distinction between relapse and withdrawal 'syndromes' was the true goal when earlier authors used the threshold. Indeed, without knowing it they demonstrate that withdrawal is not only real, it is common and clinically quite painful (often as bad or worse than MDD, by their calculations)—even in the largely irrelevant short term studies they include. Unfortunately, they misinterpret their own data because of the above mistakes, and conclude the difference was not 'clinically significant'. Unfortunate.
Hope you have an open mind, and perhaps will look for yourself. If so, let me know how you think I have it wrong.
You may be interested in this post by myself:
https://holeousia.com/2025/07/14/incidence-and-nature-of-antidepressant-discontinuation-symptoms/
My post begins:
The recent publication in JAMA Psychiatry: Incidence and Nature of Antidepressant Discontinuation Symptoms: A Systematic Review and Meta-Analysis made me feel that I had been here before.
Once again, a group of influential doctors and academics have put their names to a paper that claims to have clarified the incidence of “discontinuation symptoms” yet at the same time admits to including predominantly short term studies. Efforts have also been made to disseminate the JAMA Psychiatry paper to worldwide media. This has given some of the authors opportunities to suggest that clinical guidelines should be modified and to ignore the experience of huge numbers of people on antidepressants.
I wrote a post [full text below] in January 2024 which explored the relationship between opinion leaders and the pharmaceutical industry. It is interesting to find that a number of the authors of the recently published JAMA Psychiatry paper feature in the timeline that I compiled at that time [see below]. Perhaps it is the recent publicity generated by Reports from inquiries into both the Post Office and infected blood scandals that has led me to reflect on the reasons why those in positions of power continue to do whatever they can to protect the reputations of their institutions despite growing evidence that something is wrong.
Hey the anti-psychiatry cultists have arrived.
Please critique what Dr Gordon said, rather than throw labels around. Do you think the study is any good?
Thanks Peter: a great write up of the disgraceful conduct of psychiatrists.
Yep, it is concerning. We need a renaissance of ethics.
Hey the anti-psychiatry cultists have arrived.
Et tu Dr Mandorla?
Personal anecdotes - 2 family members and 1 close friends have all in the last 3 years suffered greatly.
1. Two on discontinuation, one
merely tapered, all got suicidal thoughts
2. One poured boiling water over their arm etc
3. One was in and out of the A&E multiple times with suspected heart issues
4. All 3 ended up back on original doses in fact 2 of them ended up on more drugs than what they started with to “get stable”
5. For one it took over 6 months to get back some level of “normality” and she had only been on the pills as migraine prevention to start with (for many many years)
Personally I’ve seen too much in real life causing distrust in majority of doctors when it comes to withdrawals.
Even pretty “benign” cardiac pharmaceuticals have caused significant withdrawals for my husband although we are aware and always try to taper slowly.
Anecdotes are not scientific evidence
The psych world uses anecdotal stories to justify so much of what they do
ECT is another great example