"And if this is true, the authors’ meta-regression was flawed". AND, it would appear at least possible if not highly likely that these flaws were understood and yet this study made no mention of them.
It would seem near impossible to conduct a scientific study of the effects of drug withdrawal that carry a heavy psychological component when the people are receiving the drugs to ameliorate psychological problems to begin with. Sorting out the placebo/nocebo effects from those of the drug itself and then its discontinuation would seem to require a lot of speculation. Add to that a meta-analysis and the chance of coming up with anything remotely scientific is extremely unlikely.
Two additional critiques that seem important to highlight:
1) The conflict of interest statement is a mile long. It’s in the pharmaceutical industry’s interest to downplay any withdrawal syndrome, so it’s not a difficult logical leap to presume an ideological bias among the authors, especially since they so strongly advocate for this flawed meta-analysis to influence clinical guidelines. As Dr. Horowitz notes, one of the authors (Allan Young) was part of the original push to downplay this in the 90s (https://mhorowitz.substack.com/p/no-the-kalfas-meta-analysis-doesnt)
2) I don’t think it’s mentioned here or in Horowitz’s analysis that a large majority of the included studies had an incredibly short follow-up period; of the 51 studies included in eAppendix 2, 39(!!) of them studied patients for two weeks or less. Only four formally tracked patient outcomes beyond 12 weeks. Providers who see patients with withdrawal syndrome will point out that a majority of patients who experience withdrawal do so after 2 weeks, so the studies will miss a substantial number of individuals who could develop issues.
The article now has a correction linked that says " ...one of the coauthor’s disclosures was inadvertently omitted. The Disclosure section has been corrected to include: “Dr Young reported personal fees from Flow Neuroscience, Novartis, Roche, Janssen, Takeda, Noema Pharma, Compass, AstraZeneca, Boehringer Ingelheim, Eli Lilly and Co, LivaNova, Lundbeck, Sunovion, Servier, Allergan, Bionomics, Sumitomo Dainippon Pharma, Sage,......" that sure is a lot of pharma fees.. 🧐🤔
I went and read the PREVENT study and found this statement:
Discontinuation due to adverse events was more common in the group receiving placebo, which may represent adverse events related to the discontinuation of Venlafaxine ER.
The authors were not unaware that discontinuation syndrome exists for these meds, evidently.
I have had many patients tell me they will not consider any antidepressant now or in the future because it was so hard for them to come off venlafaxine. Whenever I begin a discontinuation process of an antidepressant with a patient, I give them a whole talk about discontinuation syndrome, partly because it can be so unpleasant that people just go back on the med and also sometimes peole think there's something wrong with them because they're having troubles getting off it.
Psychiatric researchers are aware of discontinuation effects. Some cynics would say that these effects juice the results in favor of the intervention arm, in effect creating a favorable disparity between withdrawing patients in the placebo and treated patients. Such are the scurrilous(?) claims of anti-psychiatry.
This is not my area, but this is a fantastic commentary on medical research that is agnostic to the specifics of the clinical question: even when you have a patient who looks like the enrolled subjects in a clinical trial (ie meets all inclusion criteria), and the intervention/treatment in question was shown to be “beneficial”, in real world practice, how long do you keep a pt on therapy?
The reality is that ALL therapy trials are of limited duration…since you have to conclude at some point. In my field (cardiology), 3-5 years would be considered long follow up. The issue with very long (and very large) trials (beyond the logistics and cost of conducting such a study) is that it becomes easier for small treatment effects to reach statistical significance. There are results that we not infrequently ignore which are “statistically significant” (ie considered “real”) but of no clinical significance (due to very small treatment effect size). But even in cases of “real” effects that are “worth pursuing”, once a pt has been on therapy beyond the duration of the foundational study, we are in an evidence-free zone.
I see many pts on SSRIs in passing. So it is a bit disconcerting to find out that the evidence basis for their use is grounded on studies that are generally only 2-3 months in duration. In my med school days (last century), I recall terminology like “major depressive episode”. It seems like these trials were designed to test for benefit in treating such episodes, rather than as used for “chronic” conditions.
This seems like a corollary to the LAAO saga that Dr. JMM speaks of, but writ much larger. If millions of people have been given these drugs for decades, why didn’t somebody somewhere deem it necessary for purveyors to first prove (or at least concurrently prove) efficacy and safety?!?
In my non-professional estimation, instead of trying to push SSRIs on everyone with even the slightest mental health issue (for which there is no test), doctors should be trying to find ways to keep patients off of these drugs.
Taking these drugs does NOTHING to alleviate the true cause of mental anguish or trepidation. It is NEVER a chemical imblance as they want us to believe without question. Drugs are the trap to keep you drugged forever. Does the doctor ever tell you you can get off your meds? That they will cure you so that you won't need them any longer? Very seldom.
Imagine you wanted to create a fake drug with spurious benefit and intentionally dupe the public. Basically you would create the trials, narrative, and theater of antidepressants except call them something else. Antidepressant research is a tome of how to violate basic tenets of good science. Not even the short terms trials say what they claim they say.
Great article and I hope many read it and heed it! I can assure you I know the different kinds of withdrawal I have experienced over the years…opiates, check. Alcohol, check. Nicotine, check. Venlafaxine withdrawal was by far the most unpleasant for the longest time. I’d been on it >10 years. I am very glad I got sober - especially for the opiates - when I did prior to buperenorphine. Many “experts” in OUD stated that there is no withdrawal from it. Nearly every person I treated who was on it and wanted off had withdrawal worse than heroin. I wonder if there’s been an analysis on that?
Dr. Mandrola, I'm a frequent reader of Sensible Medicine, but I still find it fascinating that you are ready to publish articles that dispute your own writing. I know, it should be standard, but we are far away from standard recently. Disagreements are mostly published as a fierce rebuttal, with personal attacks and language that guarantees very strong language back. Or not published at all. Disagreements published here make me believe that science can still work as it's supposed to. Through politely formulated objections, questioning, and proposals for amendments. Thank you for that.
Wandering into this discussion as a lay person with observation and opinion. Antidepressants are prescribed without much thought, and there is a mix-and-match cavalier attitude (one not working? oh, let’s try this one). Had a young family member, three tours in Afghanistan, taken off a particularly nasty antidepressant and awaiting the next - committed suicide. Another friend, older retired teacher, same problem with same result. Young annd old, different life experiences, same bad outcome. Only thankful they did not ‘take others with them’. Sorry for interrupting the scholarly responses. Take care.
So true. My mom was diagnosed with cancer and I was obviously under distress. I saw a psychiatrist (referral from my uncle who is a psychiatrist, so go figure). First visit ended with a script for Celexa. My boyfriend at the time begged me not to do it. So I didn't. Reading this article, I am glad I never took them. It makes me sad to think of how hard this is for people.
You have reminded me of a personal close call with Valium. That was a long time ago. Had been thrown into a job where I had to sink or swim, responsibilities waaaay above my ‘pay grade’, but made it happen. Very stressful. Had never had that ‘pit’ in the bottom of my stomach that wouldn’t go away, like twisting a knife, so went to my wonderful old-time DO/GP. He could find nothing wrong. Decided to cover all the bases and went to my OB/GYN (DES exposure in mother’s womb). Doc couldn’t find anything wrong, either, but gave me a script for Valium. It was as if he had slapped me in the face. I was shocked he thought I needed that. Thanked him, but on the way home gave myself a talking to, haha, admonishing myself for putting myself in that position. Never filled the prescription. Glad your boyfriend talked you out of your close call. Take care.
I have been on venlafaxine for almost 30 years. I tapered off the drug about 10 years ago using a taper that my psychiatrist thought would be very conservative. Things went fine until a couple of weeks after complete discontinuation. The symptoms were totally debilitating. I was a partner in a law firm and had to take a leave of absence because my anxiety was out of control. My doctor tried putting me on a different SSRI, but to no avail. I went back on venlafaxine and eventually recovered after several weeks. I now feel trapped. A tapered down to 75 mg., but I am afraid to attempt discontinuation, even using a multi year taper. The symptoms from discontinuation were unbearable.
I succeeded in tapering off venlafaxine extended release and have not had a problem, thank goodness. It sounds as though you went about it slowly, but maybe not slowly enough? Maybe I was just lucky. I went through each reduced stage for at least 6 weeks, probably more - don't recall exactly. It took me about two years. Not a medical professional here, but I did set the course myself; my doctor would have done it much faster and more drastically, and I thought that was dangerous. I was starting out taking 225mg. Then I would take 187.5 one out of every 4 nights, with 225 the other nights; then I would increase the frequency of the lower dose to once every 3 nights and so on till I could drop totally to the lower dose. Next stage would be the same. Until finally I was off it completely. Maybe it helped that I was also taking trazodone at night; I maintained the dosing for that during the venlafaxine tapering. Eventually I did the same thing to get off the trazodone. I still have trouble sleeping and use extended release melatonin. I wish you the very best.
I would take 37.5 for several nights* and then nothing one night. I would stay on that routine for maybe a month. And then I would take 37.5 for one fewer night. *I don't remember but let's just say 6 nights on and one off; did that for a while. Then 5 nights on, one night off. And so forth. Write your dates down. I just thought the step-downs were too drastic. Did the same thing years ago using the nicotine patch to quit smoking; "they" said go from whole patch to half a patch then to quarter patch; I cut very small sections off to taper far more conservatively. I tapered off omeprazole the same way as I did the venlafaxine. (And not simulataneously.)
Purely anecdotally, I finally saw success by tapering off over a 12 month period. I stayed on each lower dosage for a minimum of four weeks. It sounds excruciatingly slow, but it’s better than feeling trapped indefinitely. Best of luck!
"I now feel trapped"...that is one of the major tenets of drug use. To keep you hooked for life. The next doctor that tells me I need to take a drug for life is going to rue that day.
Thanks for sharing. I’ve been on bupropion for over 10 years and have tried to discontinue twice. Couldn’t do it. You have to be able to sleep to function. May try again next year after I retire.
A suggestion based on anecdotal experience: switch to the liquid formulation and use pediatric dosing syringes to do a painstaking taper over at least 6 months. My understanding is that there are websites that walk people through this in more detail but I doubt many physicians are trained in this unless you find someone with a niche interest.
I see the incredible and knowledgeable @markhorowitz1 has commented on here - I am sure he can help point in a good direction.
There is also a new book out by Anders Sorensen - Crossing Zero - he as well a Mark Horowitz have substacks which might be of interest and further aid.
Best wishes, Eva
Edited - apologies Mark I got your Substack handle wrong.
I appreciate the rigorous analysis suggesting that a more careful study might demonstrate that antidepressant withdrawal is a demonstrable finding.
But suppose the data from a perfect study don't show it. Does that mean it doesn't exist? No.
We all know that we can have type II errors. That **absence of evidence** for an effect is not **evidence of absence** of an effect.
Let's not be fooled by the often-unexamined assumption of biochemical homogeneity. If there are selected but few individuals who get significant withdrawal, the signal is likely to be lost in a population-based study. That doesn't mean the issue isn't there.
Of course, we need population-based studies to guide decision-making.
But our responsibility is to our individual patients as well.
I had forgotten about this SM piece also by Wil Ward: Universal Depression Screening Leads to Unnecessary Harm. I'm glad he's covering everything from screening to withdrawal. Writing about the lack of informed consent for anti-depressants would be a nice trifecta (nudge).
Vinay at FDA or NIH could initiate a few RCT's of antidepressants for a longer duration of 2-3 years! These studies will be never funded from pharma for sure.
These studies are vital to conduct so we understand the long term safety and efficacy of these drugs, which are now extrapolated from short term studies.
I “discovered” Gene Glass and meta-analysis (see Wiki) in the ‘80’s, when I was doing a dissertation on Agent Orange for my Dr. P. H. and needed to mix apples and oranges, so to speak. I was hoping to makes sense of the disparate studies to distinguish correlation from causality in attributing harm from dioxins. Meta Analysis helped me informally as a framework but hadn’t been obfuscated to the the pantheon it occupies now.
Bravo thank you for pointing out why so many have a very difficult time coming off these drugs. Most folks have been on these drugs for decades
"And if this is true, the authors’ meta-regression was flawed". AND, it would appear at least possible if not highly likely that these flaws were understood and yet this study made no mention of them.
SHAMEFUL. This is where we are at.
It would seem near impossible to conduct a scientific study of the effects of drug withdrawal that carry a heavy psychological component when the people are receiving the drugs to ameliorate psychological problems to begin with. Sorting out the placebo/nocebo effects from those of the drug itself and then its discontinuation would seem to require a lot of speculation. Add to that a meta-analysis and the chance of coming up with anything remotely scientific is extremely unlikely.
Two additional critiques that seem important to highlight:
1) The conflict of interest statement is a mile long. It’s in the pharmaceutical industry’s interest to downplay any withdrawal syndrome, so it’s not a difficult logical leap to presume an ideological bias among the authors, especially since they so strongly advocate for this flawed meta-analysis to influence clinical guidelines. As Dr. Horowitz notes, one of the authors (Allan Young) was part of the original push to downplay this in the 90s (https://mhorowitz.substack.com/p/no-the-kalfas-meta-analysis-doesnt)
2) I don’t think it’s mentioned here or in Horowitz’s analysis that a large majority of the included studies had an incredibly short follow-up period; of the 51 studies included in eAppendix 2, 39(!!) of them studied patients for two weeks or less. Only four formally tracked patient outcomes beyond 12 weeks. Providers who see patients with withdrawal syndrome will point out that a majority of patients who experience withdrawal do so after 2 weeks, so the studies will miss a substantial number of individuals who could develop issues.
The article now has a correction linked that says " ...one of the coauthor’s disclosures was inadvertently omitted. The Disclosure section has been corrected to include: “Dr Young reported personal fees from Flow Neuroscience, Novartis, Roche, Janssen, Takeda, Noema Pharma, Compass, AstraZeneca, Boehringer Ingelheim, Eli Lilly and Co, LivaNova, Lundbeck, Sunovion, Servier, Allergan, Bionomics, Sumitomo Dainippon Pharma, Sage,......" that sure is a lot of pharma fees.. 🧐🤔
prostitutes see fewer clients
That's funny
I went and read the PREVENT study and found this statement:
Discontinuation due to adverse events was more common in the group receiving placebo, which may represent adverse events related to the discontinuation of Venlafaxine ER.
The authors were not unaware that discontinuation syndrome exists for these meds, evidently.
I have had many patients tell me they will not consider any antidepressant now or in the future because it was so hard for them to come off venlafaxine. Whenever I begin a discontinuation process of an antidepressant with a patient, I give them a whole talk about discontinuation syndrome, partly because it can be so unpleasant that people just go back on the med and also sometimes peole think there's something wrong with them because they're having troubles getting off it.
My mom was one. She was so angry
Psychiatric researchers are aware of discontinuation effects. Some cynics would say that these effects juice the results in favor of the intervention arm, in effect creating a favorable disparity between withdrawing patients in the placebo and treated patients. Such are the scurrilous(?) claims of anti-psychiatry.
This is not my area, but this is a fantastic commentary on medical research that is agnostic to the specifics of the clinical question: even when you have a patient who looks like the enrolled subjects in a clinical trial (ie meets all inclusion criteria), and the intervention/treatment in question was shown to be “beneficial”, in real world practice, how long do you keep a pt on therapy?
The reality is that ALL therapy trials are of limited duration…since you have to conclude at some point. In my field (cardiology), 3-5 years would be considered long follow up. The issue with very long (and very large) trials (beyond the logistics and cost of conducting such a study) is that it becomes easier for small treatment effects to reach statistical significance. There are results that we not infrequently ignore which are “statistically significant” (ie considered “real”) but of no clinical significance (due to very small treatment effect size). But even in cases of “real” effects that are “worth pursuing”, once a pt has been on therapy beyond the duration of the foundational study, we are in an evidence-free zone.
I see many pts on SSRIs in passing. So it is a bit disconcerting to find out that the evidence basis for their use is grounded on studies that are generally only 2-3 months in duration. In my med school days (last century), I recall terminology like “major depressive episode”. It seems like these trials were designed to test for benefit in treating such episodes, rather than as used for “chronic” conditions.
This seems like a corollary to the LAAO saga that Dr. JMM speaks of, but writ much larger. If millions of people have been given these drugs for decades, why didn’t somebody somewhere deem it necessary for purveyors to first prove (or at least concurrently prove) efficacy and safety?!?
In my non-professional estimation, instead of trying to push SSRIs on everyone with even the slightest mental health issue (for which there is no test), doctors should be trying to find ways to keep patients off of these drugs.
Taking these drugs does NOTHING to alleviate the true cause of mental anguish or trepidation. It is NEVER a chemical imblance as they want us to believe without question. Drugs are the trap to keep you drugged forever. Does the doctor ever tell you you can get off your meds? That they will cure you so that you won't need them any longer? Very seldom.
Imagine you wanted to create a fake drug with spurious benefit and intentionally dupe the public. Basically you would create the trials, narrative, and theater of antidepressants except call them something else. Antidepressant research is a tome of how to violate basic tenets of good science. Not even the short terms trials say what they claim they say.
Great article and I hope many read it and heed it! I can assure you I know the different kinds of withdrawal I have experienced over the years…opiates, check. Alcohol, check. Nicotine, check. Venlafaxine withdrawal was by far the most unpleasant for the longest time. I’d been on it >10 years. I am very glad I got sober - especially for the opiates - when I did prior to buperenorphine. Many “experts” in OUD stated that there is no withdrawal from it. Nearly every person I treated who was on it and wanted off had withdrawal worse than heroin. I wonder if there’s been an analysis on that?
Dr. Mandrola, I'm a frequent reader of Sensible Medicine, but I still find it fascinating that you are ready to publish articles that dispute your own writing. I know, it should be standard, but we are far away from standard recently. Disagreements are mostly published as a fierce rebuttal, with personal attacks and language that guarantees very strong language back. Or not published at all. Disagreements published here make me believe that science can still work as it's supposed to. Through politely formulated objections, questioning, and proposals for amendments. Thank you for that.
Wandering into this discussion as a lay person with observation and opinion. Antidepressants are prescribed without much thought, and there is a mix-and-match cavalier attitude (one not working? oh, let’s try this one). Had a young family member, three tours in Afghanistan, taken off a particularly nasty antidepressant and awaiting the next - committed suicide. Another friend, older retired teacher, same problem with same result. Young annd old, different life experiences, same bad outcome. Only thankful they did not ‘take others with them’. Sorry for interrupting the scholarly responses. Take care.
So true. My mom was diagnosed with cancer and I was obviously under distress. I saw a psychiatrist (referral from my uncle who is a psychiatrist, so go figure). First visit ended with a script for Celexa. My boyfriend at the time begged me not to do it. So I didn't. Reading this article, I am glad I never took them. It makes me sad to think of how hard this is for people.
You have reminded me of a personal close call with Valium. That was a long time ago. Had been thrown into a job where I had to sink or swim, responsibilities waaaay above my ‘pay grade’, but made it happen. Very stressful. Had never had that ‘pit’ in the bottom of my stomach that wouldn’t go away, like twisting a knife, so went to my wonderful old-time DO/GP. He could find nothing wrong. Decided to cover all the bases and went to my OB/GYN (DES exposure in mother’s womb). Doc couldn’t find anything wrong, either, but gave me a script for Valium. It was as if he had slapped me in the face. I was shocked he thought I needed that. Thanked him, but on the way home gave myself a talking to, haha, admonishing myself for putting myself in that position. Never filled the prescription. Glad your boyfriend talked you out of your close call. Take care.
I have been on venlafaxine for almost 30 years. I tapered off the drug about 10 years ago using a taper that my psychiatrist thought would be very conservative. Things went fine until a couple of weeks after complete discontinuation. The symptoms were totally debilitating. I was a partner in a law firm and had to take a leave of absence because my anxiety was out of control. My doctor tried putting me on a different SSRI, but to no avail. I went back on venlafaxine and eventually recovered after several weeks. I now feel trapped. A tapered down to 75 mg., but I am afraid to attempt discontinuation, even using a multi year taper. The symptoms from discontinuation were unbearable.
I succeeded in tapering off venlafaxine extended release and have not had a problem, thank goodness. It sounds as though you went about it slowly, but maybe not slowly enough? Maybe I was just lucky. I went through each reduced stage for at least 6 weeks, probably more - don't recall exactly. It took me about two years. Not a medical professional here, but I did set the course myself; my doctor would have done it much faster and more drastically, and I thought that was dangerous. I was starting out taking 225mg. Then I would take 187.5 one out of every 4 nights, with 225 the other nights; then I would increase the frequency of the lower dose to once every 3 nights and so on till I could drop totally to the lower dose. Next stage would be the same. Until finally I was off it completely. Maybe it helped that I was also taking trazodone at night; I maintained the dosing for that during the venlafaxine tapering. Eventually I did the same thing to get off the trazodone. I still have trouble sleeping and use extended release melatonin. I wish you the very best.
Thanks. Two years is a long time, but I think that’s probably what it will take. Maybe more for me.
I’ve never heard of anyone doing it the way you did. What did you do when you got down to 37.5 mg? Going from that to zero is what got me.
I would take 37.5 for several nights* and then nothing one night. I would stay on that routine for maybe a month. And then I would take 37.5 for one fewer night. *I don't remember but let's just say 6 nights on and one off; did that for a while. Then 5 nights on, one night off. And so forth. Write your dates down. I just thought the step-downs were too drastic. Did the same thing years ago using the nicotine patch to quit smoking; "they" said go from whole patch to half a patch then to quarter patch; I cut very small sections off to taper far more conservatively. I tapered off omeprazole the same way as I did the venlafaxine. (And not simulataneously.)
Purely anecdotally, I finally saw success by tapering off over a 12 month period. I stayed on each lower dosage for a minimum of four weeks. It sounds excruciatingly slow, but it’s better than feeling trapped indefinitely. Best of luck!
Were you on a long acting capsule where you counted the individual balls?
"I now feel trapped"...that is one of the major tenets of drug use. To keep you hooked for life. The next doctor that tells me I need to take a drug for life is going to rue that day.
Thanks for sharing. I’ve been on bupropion for over 10 years and have tried to discontinue twice. Couldn’t do it. You have to be able to sleep to function. May try again next year after I retire.
I’m now retired but have no clue how to discontinue and don’t know a doctor who does either.
A suggestion based on anecdotal experience: switch to the liquid formulation and use pediatric dosing syringes to do a painstaking taper over at least 6 months. My understanding is that there are websites that walk people through this in more detail but I doubt many physicians are trained in this unless you find someone with a niche interest.
I see the incredible and knowledgeable @markhorowitz1 has commented on here - I am sure he can help point in a good direction.
There is also a new book out by Anders Sorensen - Crossing Zero - he as well a Mark Horowitz have substacks which might be of interest and further aid.
Best wishes, Eva
Edited - apologies Mark I got your Substack handle wrong.
I appreciate the rigorous analysis suggesting that a more careful study might demonstrate that antidepressant withdrawal is a demonstrable finding.
But suppose the data from a perfect study don't show it. Does that mean it doesn't exist? No.
We all know that we can have type II errors. That **absence of evidence** for an effect is not **evidence of absence** of an effect.
Let's not be fooled by the often-unexamined assumption of biochemical homogeneity. If there are selected but few individuals who get significant withdrawal, the signal is likely to be lost in a population-based study. That doesn't mean the issue isn't there.
Of course, we need population-based studies to guide decision-making.
But our responsibility is to our individual patients as well.
I had forgotten about this SM piece also by Wil Ward: Universal Depression Screening Leads to Unnecessary Harm. I'm glad he's covering everything from screening to withdrawal. Writing about the lack of informed consent for anti-depressants would be a nice trifecta (nudge).
https://www.sensible-med.com/p/universal-depression-screening-leads?lli=1&utm_source=profile&utm_medium=reader2
Vinay at FDA or NIH could initiate a few RCT's of antidepressants for a longer duration of 2-3 years! These studies will be never funded from pharma for sure.
These studies are vital to conduct so we understand the long term safety and efficacy of these drugs, which are now extrapolated from short term studies.
Next time when you speak with Vinay, Adam and Mandrola let him know about please :). And this is something Dr. Makary and everyone would endorse!
I “discovered” Gene Glass and meta-analysis (see Wiki) in the ‘80’s, when I was doing a dissertation on Agent Orange for my Dr. P. H. and needed to mix apples and oranges, so to speak. I was hoping to makes sense of the disparate studies to distinguish correlation from causality in attributing harm from dioxins. Meta Analysis helped me informally as a framework but hadn’t been obfuscated to the the pantheon it occupies now.
John Hopkins