I was diagnosed with major depressive disorder over 40 years ago, suffered from depression starting in my early teens. I started taking Prozac after the birth of my third child when I was suffering from severe postpartum depression. I remember the experience when it finally started working, it was like a gray cloud was lifted from my brain. I tried tapering off with the help of a medical doctor about 20 years ago, but relapsed into a major depressive incident. So now I’ve been on Prozac for probably 40+ years. I still have ups and downs. My psychiatrist and I have talked about tapering off, but he does not think that in my situation it’s a good idea, and I agree. It still works for me. The only thing I get worried about is the long-term effects on my brain, as I am 75 years old. But I think every patient is different and it’s really important to listen to what their experience is. My biggest fear in life has been my depression. At one time prior to Prozac I knew that that’s how my life would end, with a depressive incident. I appreciate all the studies that are being done and I love reading articles like this. Thank you.
I personally think that all the hubbub about antidepressant withdrawal is pretty pointless. First of all, I am not "confused". When patients complain of withdrawal symptoms (meaning they complain of various problems after decreasing or stopping an antidepressant), I appreciate as much variability as I do with patients' starting antidepressants.
Second of all, since the dosing and side effect information in the inserts is completely inadequate to encompass all the reactions my patients have when they start antidepressants, why would I expect less variability when they decrease or stop antidepressants?
Third of all, every single discussion I have seen (both about starting an antidepressant and stopping an antidepressant), presupposes that a doctor makes clinical decisions for a passive patient based on his total mastery of total knowledge. That is not anything like the reality. Each of my patients has different values related to stability, uncertainty, risk, and taking medication itself. All I can or should do is tell them what I know, which isn't much because patients are so unique, and promise close follow-up and the use of my clinical intincts.
Fourth, we speak about patients as if they form some collective unity that can be treated as one individual about whom guidelines shift depending on the prevailing wind of online chatter and literature. They can't. I have found that I solved every every clinical problem with intense follow-up and close collaboration with the patient, unless I failed to form an alliance (and that can happen, for example I didn't give the patient the Xanax they wanted or the patient didn't feel heard beacuse of my own anxiety and high volume practice). Since we know most antidepressants are prescribed by non-psychiatrists, this is where the challenges lie, most patients are on their own.
Fifth, after that most antidepressants are prescribed by non-psychiatrists, the worst problem is that most information traded between both non-clinicians and clinicians for the past quarter century has been heavily contaminated by pharma marketing efforts. Refusing pens and lunches has done nothing to eliminate or even reduce this issue. Prescribers' knowledge about the meds they prescribe, or even reasonable decision making, is useless because they are based on a marketing hype that has nothing to do with reality or even real knowledge or evidence.
None of this is anti-psychiatry. This is anti-fake psychiatry.
Very true, Dr. Bresch. Intense follow-up and close collaboration with the patient is the secret sauce to tapering. Trial and error, correct as you go along. But many practitioners seem leery of this.
Do you think there is some kind of generalized withdrawal syndrome independent of the kind of drug involved? My guess from first principles would be no and that different drugs and different receptors might behave very differently. I mean why should discontinuing two different drugs be any more similar than the drugs are in the first place?
For instance, with opiates many of the effects of withdrawal seem to be traceable pretty directly to the loss of many of the g-protein coupled receptors. I believe that with opiates there is some evidence that there is incomplete cross-tolerance (perhaps relating to differential affinity for receptor subtypes). That might suggest in that context there could be benefits from rotating the drugs supplied during tapering.
OTOH it seems like the mechanisms producing SSRI withdrawal are generally a bit more indirect and complex.
Across psychotropics, withdrawal symptoms and especially the protracted conditions are remarkably similar (Lerner & Klein, 2019 https://doi.org/10.1093/braincomms/fcz025). This suggests a central cause that is not specific to the drug. After all, psychotropics are categorized as central nervous system drugs. Our thinking about them is far too brain- and receptor-specific.
The CNS and autonomic nervous system are extremely complex and poorly understood. Rather than puzzle them out in order to completely understand withdrawal syndromes, it makes more sense to minimize actions that might cause their disruption (including swapping psychotropics) and thereby minimize these undesirable iatrogenic outcomes.
As someone who's stopped literally dozens of psychotropics cold turkey with no issues... Including after a year of 30mg diazepam/day, and even drugs with acknowledged rebound illnesses... I suspect that there are plenty of patients out there like me who can stop cold turkey with minimal issues. The nocebo effect is real. So are withdrawals, but forcing everyone into a long drawn out taper ignores the fact that it will often be unnecessary.
I don't think using one's own personal experience, is a good guide for clinical decision making, mostly because you are only one person. Your experience is valid. But your patients shouldn't have to compare themselves to you explictily or implicitly. I also think that talking about placebo effects and nocebo effects is pejorative, disheartening and inaccurate, since the placebo effect is something we discern statistically, not in an individual case (unless we had the bad judgment to use a sugar pill, or accidentally adminsitered water (for example). Even then, all illnesses have variability and that is inherent to illness, not suggestibility.
I don’t think that was the intention. I think it was just using the exact same methods Ms Framer used for her own initiative. We have to consider that most people have no issues and prolonged tapering regimens are useless in those cases
Indications are that a substantial minority of antidepressant users have serious issues with current haphazard tapering methods. The assumption that more gradual tapering is useless is completely unfounded. The trend is towards ever-more gradual tapering, as pain medicine and addiction medicine are investigating for their psychotropics. Would you consider harming 1 out of 10 of your psychiatric patients to be a minor issue? Over the entire population of those taking psychiatric drugs, that is still millions of cases of avoidable harm.
I actually don't know is this is true. Mindful that 80% of antidepressants are prescribed by non-psychiatrists, patients might assort such that the ones with problems all go to the specialist like me. And I have to say that in my practice, most of my patients had problems with antidepressants, starting thekm, stopping them, taking them. It's a lot. I have more complaints about antidepressants than I do benzodiazepines for example or stimulants, thats for sure!
In my observation, the boundary between patients treated with psychiatric drugs by non-psychiatrists and those by psychiatrists is porous, with most traffic going from non-psychiatrists to psychiatrists when drug treatment goes wrong. Then, what was originally a mild-moderate or undiagnosed condition might get an escalated diagnosis, such as "treatement resistant", transforming pseudo-resistance or an otherwise non-condition ripe for deprescribing into a more severe candidate for serial polypharmacy. See Howes, O. D., Thase, M. E., & Pillinger, T. (2021). Treatment resistance in psychiatry: State of the art and new directions. Molecular Psychiatry. https://doi.org/10.1038/s41380-021-01200-3
I think that highlights the issue. To me, Framer suggests fitting tapering regimens around the folks who experience these withdrawal symptoms based off her research. With an n=23k since 2011, that is a small sample size given the millions of people who do take antidepressants. These methods are insufficient to guide clinical practice and it suffers from selection bias since the edge cases naturally seek each other out on the internet. A 2024 Lancet study pegs the incidence rate around 15% with severe discontinuation around 3%. More interestingly, 17% of patients on a placebo reported an effect.
In 2025, JAMA Psychiatry reported patients reported 1 additional symptom after stopping an antidepressant compared to placebo using the DESS scale. Ms. Framer did not include any of these perspectives in her article or address these studies. I agree we went way too hard on discontinuing with ludicrous suggestions like breaking the dose in half for 3 days (looking at you, Effexor) and stopping. But as so often happens in American medicine, we swing way too hard the other way like we are seeing with opioids in pain management.
In my experience, benzos have been much more difficult to stop than antidepressants. Anyway, again, my aim here is not to discredit Framer but to push back against the withdrawal conversation and ground this in peer-reviewed studies. Studies aren't perfect but neither are selection effect experiences.
Remarkable that N=23K (only a portion of the population in online withdrawal groups) is considered insufficient evidence that something is not right in psychiatric drug land. There is also quite a lot of literature going back decades about other psychotropic withdrawal syndromes, but oh well....
Nobody has swung in any way. We are only at the dawn of figuring out tapering. Yet even this glimmer is a swing too far?
Please feel free to deprescribe psychiatric drugs any way that you feel is appropriate, but don't do this: 1) cold turkey; 2) “cut in half, then half again, then off”; 3) skip doses to taper. If your own method goes wrong, take responsibility for it and apologize to your patient.
Ms. Framer -- I did not find this threatening and I am not sure where you got this. I did not suggest we taper people off in 3 days. This is my direct quote:
"I agree we went way too hard on discontinuing with ludicrous suggestions like breaking the dose in half for 3 days (looking at you, Effexor) and stopping."
Could you direct me to where you saw 3 days as my recommended approach?
I feel like words are being put into my mouth here. I am coming here in good faith with some peer reviewed studies that complicate the picture but as the internet often does, we turned it into mud slinging.
Maybe. but if there are other patients who can't that's important.
Also, 30mg/day diazepam isn't all that much in the grand scheme of things. Come back and share your experiences after you've gone cold turkey from 120mg/day of methadone.
Diazepam has the benefit of a relatively long half-life and 30mg may be large for a dose with current prescribing guidelines (we learned in the 70s about the memory harms of higher dosing regimes) it really isn't that much higher than starting doses. Also it matters if you have been using it for a few months or 20 years.
What an ignorant comment. So your experience is the only true experience? I’m sure if you were sitting on the other side you’d be speaking much differently.
I was diagnosed with major depressive disorder over 40 years ago, suffered from depression starting in my early teens. I started taking Prozac after the birth of my third child when I was suffering from severe postpartum depression. I remember the experience when it finally started working, it was like a gray cloud was lifted from my brain. I tried tapering off with the help of a medical doctor about 20 years ago, but relapsed into a major depressive incident. So now I’ve been on Prozac for probably 40+ years. I still have ups and downs. My psychiatrist and I have talked about tapering off, but he does not think that in my situation it’s a good idea, and I agree. It still works for me. The only thing I get worried about is the long-term effects on my brain, as I am 75 years old. But I think every patient is different and it’s really important to listen to what their experience is. My biggest fear in life has been my depression. At one time prior to Prozac I knew that that’s how my life would end, with a depressive incident. I appreciate all the studies that are being done and I love reading articles like this. Thank you.
Brilliantly written Adele!
I personally think that all the hubbub about antidepressant withdrawal is pretty pointless. First of all, I am not "confused". When patients complain of withdrawal symptoms (meaning they complain of various problems after decreasing or stopping an antidepressant), I appreciate as much variability as I do with patients' starting antidepressants.
Second of all, since the dosing and side effect information in the inserts is completely inadequate to encompass all the reactions my patients have when they start antidepressants, why would I expect less variability when they decrease or stop antidepressants?
Third of all, every single discussion I have seen (both about starting an antidepressant and stopping an antidepressant), presupposes that a doctor makes clinical decisions for a passive patient based on his total mastery of total knowledge. That is not anything like the reality. Each of my patients has different values related to stability, uncertainty, risk, and taking medication itself. All I can or should do is tell them what I know, which isn't much because patients are so unique, and promise close follow-up and the use of my clinical intincts.
Fourth, we speak about patients as if they form some collective unity that can be treated as one individual about whom guidelines shift depending on the prevailing wind of online chatter and literature. They can't. I have found that I solved every every clinical problem with intense follow-up and close collaboration with the patient, unless I failed to form an alliance (and that can happen, for example I didn't give the patient the Xanax they wanted or the patient didn't feel heard beacuse of my own anxiety and high volume practice). Since we know most antidepressants are prescribed by non-psychiatrists, this is where the challenges lie, most patients are on their own.
Fifth, after that most antidepressants are prescribed by non-psychiatrists, the worst problem is that most information traded between both non-clinicians and clinicians for the past quarter century has been heavily contaminated by pharma marketing efforts. Refusing pens and lunches has done nothing to eliminate or even reduce this issue. Prescribers' knowledge about the meds they prescribe, or even reasonable decision making, is useless because they are based on a marketing hype that has nothing to do with reality or even real knowledge or evidence.
None of this is anti-psychiatry. This is anti-fake psychiatry.
Very true, Dr. Bresch. Intense follow-up and close collaboration with the patient is the secret sauce to tapering. Trial and error, correct as you go along. But many practitioners seem leery of this.
Do you think there is some kind of generalized withdrawal syndrome independent of the kind of drug involved? My guess from first principles would be no and that different drugs and different receptors might behave very differently. I mean why should discontinuing two different drugs be any more similar than the drugs are in the first place?
For instance, with opiates many of the effects of withdrawal seem to be traceable pretty directly to the loss of many of the g-protein coupled receptors. I believe that with opiates there is some evidence that there is incomplete cross-tolerance (perhaps relating to differential affinity for receptor subtypes). That might suggest in that context there could be benefits from rotating the drugs supplied during tapering.
OTOH it seems like the mechanisms producing SSRI withdrawal are generally a bit more indirect and complex.
Across psychotropics, withdrawal symptoms and especially the protracted conditions are remarkably similar (Lerner & Klein, 2019 https://doi.org/10.1093/braincomms/fcz025). This suggests a central cause that is not specific to the drug. After all, psychotropics are categorized as central nervous system drugs. Our thinking about them is far too brain- and receptor-specific.
The CNS and autonomic nervous system are extremely complex and poorly understood. Rather than puzzle them out in order to completely understand withdrawal syndromes, it makes more sense to minimize actions that might cause their disruption (including swapping psychotropics) and thereby minimize these undesirable iatrogenic outcomes.
As someone who's stopped literally dozens of psychotropics cold turkey with no issues... Including after a year of 30mg diazepam/day, and even drugs with acknowledged rebound illnesses... I suspect that there are plenty of patients out there like me who can stop cold turkey with minimal issues. The nocebo effect is real. So are withdrawals, but forcing everyone into a long drawn out taper ignores the fact that it will often be unnecessary.
I don't think using one's own personal experience, is a good guide for clinical decision making, mostly because you are only one person. Your experience is valid. But your patients shouldn't have to compare themselves to you explictily or implicitly. I also think that talking about placebo effects and nocebo effects is pejorative, disheartening and inaccurate, since the placebo effect is something we discern statistically, not in an individual case (unless we had the bad judgment to use a sugar pill, or accidentally adminsitered water (for example). Even then, all illnesses have variability and that is inherent to illness, not suggestibility.
I don’t think that was the intention. I think it was just using the exact same methods Ms Framer used for her own initiative. We have to consider that most people have no issues and prolonged tapering regimens are useless in those cases
Indications are that a substantial minority of antidepressant users have serious issues with current haphazard tapering methods. The assumption that more gradual tapering is useless is completely unfounded. The trend is towards ever-more gradual tapering, as pain medicine and addiction medicine are investigating for their psychotropics. Would you consider harming 1 out of 10 of your psychiatric patients to be a minor issue? Over the entire population of those taking psychiatric drugs, that is still millions of cases of avoidable harm.
I actually don't know is this is true. Mindful that 80% of antidepressants are prescribed by non-psychiatrists, patients might assort such that the ones with problems all go to the specialist like me. And I have to say that in my practice, most of my patients had problems with antidepressants, starting thekm, stopping them, taking them. It's a lot. I have more complaints about antidepressants than I do benzodiazepines for example or stimulants, thats for sure!
In my observation, the boundary between patients treated with psychiatric drugs by non-psychiatrists and those by psychiatrists is porous, with most traffic going from non-psychiatrists to psychiatrists when drug treatment goes wrong. Then, what was originally a mild-moderate or undiagnosed condition might get an escalated diagnosis, such as "treatement resistant", transforming pseudo-resistance or an otherwise non-condition ripe for deprescribing into a more severe candidate for serial polypharmacy. See Howes, O. D., Thase, M. E., & Pillinger, T. (2021). Treatment resistance in psychiatry: State of the art and new directions. Molecular Psychiatry. https://doi.org/10.1038/s41380-021-01200-3
I think that highlights the issue. To me, Framer suggests fitting tapering regimens around the folks who experience these withdrawal symptoms based off her research. With an n=23k since 2011, that is a small sample size given the millions of people who do take antidepressants. These methods are insufficient to guide clinical practice and it suffers from selection bias since the edge cases naturally seek each other out on the internet. A 2024 Lancet study pegs the incidence rate around 15% with severe discontinuation around 3%. More interestingly, 17% of patients on a placebo reported an effect.
In 2025, JAMA Psychiatry reported patients reported 1 additional symptom after stopping an antidepressant compared to placebo using the DESS scale. Ms. Framer did not include any of these perspectives in her article or address these studies. I agree we went way too hard on discontinuing with ludicrous suggestions like breaking the dose in half for 3 days (looking at you, Effexor) and stopping. But as so often happens in American medicine, we swing way too hard the other way like we are seeing with opioids in pain management.
Lancet study: https://pubmed.ncbi.nlm.nih.gov/38851198/
JAMA: https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2836262?utm_source=openevidence&utm_medium=referral
In my experience, benzos have been much more difficult to stop than antidepressants. Anyway, again, my aim here is not to discredit Framer but to push back against the withdrawal conversation and ground this in peer-reviewed studies. Studies aren't perfect but neither are selection effect experiences.
Remarkable that N=23K (only a portion of the population in online withdrawal groups) is considered insufficient evidence that something is not right in psychiatric drug land. There is also quite a lot of literature going back decades about other psychotropic withdrawal syndromes, but oh well....
Nobody has swung in any way. We are only at the dawn of figuring out tapering. Yet even this glimmer is a swing too far?
Please feel free to deprescribe psychiatric drugs any way that you feel is appropriate, but don't do this: 1) cold turkey; 2) “cut in half, then half again, then off”; 3) skip doses to taper. If your own method goes wrong, take responsibility for it and apologize to your patient.
Ms. Framer -- I did not find this threatening and I am not sure where you got this. I did not suggest we taper people off in 3 days. This is my direct quote:
"I agree we went way too hard on discontinuing with ludicrous suggestions like breaking the dose in half for 3 days (looking at you, Effexor) and stopping."
Could you direct me to where you saw 3 days as my recommended approach?
I feel like words are being put into my mouth here. I am coming here in good faith with some peer reviewed studies that complicate the picture but as the internet often does, we turned it into mud slinging.
Maybe. but if there are other patients who can't that's important.
Also, 30mg/day diazepam isn't all that much in the grand scheme of things. Come back and share your experiences after you've gone cold turkey from 120mg/day of methadone.
Diazepam has the benefit of a relatively long half-life and 30mg may be large for a dose with current prescribing guidelines (we learned in the 70s about the memory harms of higher dosing regimes) it really isn't that much higher than starting doses. Also it matters if you have been using it for a few months or 20 years.
What an ignorant comment. So your experience is the only true experience? I’m sure if you were sitting on the other side you’d be speaking much differently.