You know I've spent the better part of the last two years deep in a withdrawal group trying to figure this all out because I think it's much more interesting than people give it credit for.
I agree with you about the hyperbolic thing, it's vanishingly unlikely to beat just going slower.
You've underestimated the homogeneity though. It could be the population I'm involved with, perhaps more severe cases go to the effort of joining support groups but far from a menagerie, its really quite astounding how few and narrow the phenotypes are.
The CFS/post viral thing is the most interesting part. At least among the population I'm involved in most cases seem to fall under an acquired/organic neurasthenic rubric. So you would think it would follow that a great many are neurotic but this doesn't seem to be the case at all. And certainly I'm sufficiently insecure to have considered it. The thing is, I'm quite satisfied that the onset of this "CFS" syndrome is de novo in the vast majority of cases. Temporally it's very closely tied either to going on the drug or coming off it and despite many having mood disorder that put them on the drug, for most that mood disorder was not neurasthenia, nor anxiety neurosis.
Look, if you'd have asked me to guess before I got chin deep in it I would have absolutely predicted neurosis but it doesn't reconcile once you're deep enough in the qualitative data.
The other interesting angle is that closer questioning has revealed a supervising number of descriptions that shade very close to temporal lobe epilepsy.
There also seems to be some strange relationship to PMDD and possibly confusional psychosis.
Also, lupus. That's another weird wrinkle.
It's all very strange and there is great temptation to conclude that the problem was pre-existing but some cases are well into their middle years having suffered an all together very different problem.
Thank you for the excellent deconstruction of the nebulous and protean term, "withdrawal," Awais, as well as for your informed skepticism regarding various claims about the nature, frequency, and severity of antidepressant-related "withdrawal" syndromes [ARWS]. This is a topic that easily devolves into what I call, the 3 p's: polarization, personalization, and politicization. For example, suggesting that the vast majority of ARWS are neither severe nor prolonged often gets excoriated in social media as "Psychiatry's denial and minimization" of the problem.
Having reviewed this literature for a planned article in Psychiatric News, I am convinced that the kind of rigorous, controlled studies of ARWS that we need--including use of the standardized DESS scale--have simply not been done. Thus, our position as a profession ought to be one of cautious humility, not polemical certainty. Much more can and should be said, and I hope your readers will find the piece by Dr. Jonathan Henssler and me of some merit.
Our main conclusion: "To be clear, we want to acknowledge that whatever the usual frequency and duration of AWDS, some patients may experience severe and prolonged withdrawal symptoms that greatly interfere with activities of daily living, vocational function, and quality of life. These facts were not sufficiently emphasized in early clinical guidelines. Nevertheless, in our experience, most problems with antidepressant discontinuation can be averted by slow, careful, individualized tapering, often over several months, and—all other things being equal—by avoiding short half-life agents (eg, paroxetine and venlafaxine) in the first place."
Outstandingly clear and helpful - thank you! It’s only anecdotal, but I’ve never observed craving for the discontinued antidepressant, whereas craving is a hallmark - just about pathognomonic - of discontinuation, ‘true’ withdrawal, of alcohol, tobacco or opiates. For what it’s worth, I think this is a meaningful difference between discontinuation phenomena and withdrawal, which describe distinct symptoms or states. Rapid reinstatement after abstinence is also something I haven’t met with antidepressants. And tolerance isn’t the same either. All of this argues that discontinuation/withdrawal is more than a semantic discussion.
(1)I found this summary helpful, exhaustive and somewhat exhausting. Hope that some of these categories will coalesce and others disappear, and not fall into some dreaded NOS category we love so much about DSM.
(2) It seems so hard not to avoid the problem that many words that mean different things to different people, thereby making agreements of what's testable and how to do so painfully difficult.
(3) I was struck by your heading of "what do we REALLY (my caps ) know?" - having to add "really", I guess distinguishes what follows from what we know - but not really. It seems you're wanting to make a distinction from, perhaps, what others may think but not truly know. A real tightrope.
(4) I'd just add that there is a constituency out there that thinks that using meds is almost always the wrong way to go, which makes it easy for them to add "withdrawal" to their list of psychiatric sins. If we were wrong to medicate in the first place, we've set the table for the "withdrawal," which that group REALLY knows is inevitable.
Yet *another* useful essay, bravo. Three thoughts leading to one strong recommendation:
1. I've seen very credible severe withdrawal with 1 mg steps at the end of a citalopram taper where larger steps earlier were not so bad. So regardless of receptor occupancy (but thanks for pointing out complexities there), tiny steps at the end are definitely needed by some.
2. But probably not many? Our small study found 90% of people in primary care did not even receive a prescription for the smallest available dose before stopping their antidepressant. They may have suffered but didn't go back on (or disappear; Phelps et al, JABFM, 2023). But who will be among those with horrible withdrawal -- whatever the basis --so as to know who really needs that many-months quasi-linear approach? Simple...
3. Start *everyone* with the smallest decrement possible. Many need it at the end anyway and this step protects everyone. It addresses concerns about worsening, lowering nocebo risks. Those who don't fear withdrawal can take faster or bigger steps on the way down; those who do can go by that same small decrement all the way. Anyone who gets symptoms on the way down should slow down. If each step is still rough, get a liquid (hardly ever necessary in my experience). I never had -- but know they're out there -- a patient go through what Adele Framer has said is needed for some, but for quite a few it was many, many months. Not years.
Oh and ever notice how serotonergic withdrawal symptoms look like bipolar mixed states? Withdrawal more common in people with bipolarity, I've always wondered? So that lamotrigine idea...
Severe withdrawal with tiny steps at the end… I don’t reject the possibility and I defer to your report; my main point is that whatever the mechanism for this phenomenon, receptor occupancy curve doesn’t explain it. Something other mechanism needs to be invoked, a super-sensitivity of some sort or whatever.
Your recommendations are quite sensible. And there is indeed a curious similarity between serotonergic withdrawal and mixed states!
Yes, they do look like mixed states. In fact, I would even go so far as to say many are mixed states by definition, but are they "bipolar mixed states"? I used to think so but now I'm not nearly as sure. The closest thing I've found is Michael Alan Taylor's description of "limbic sensitisation" syndrome in his little book on clinical rounds.
To be honest, having spent much of the ladt two years talking to people in these states, I'm struck by the level of lability. I think it shades closely towards temporal lobe and limbic abnormalities but not necessarily classic bipolar mania. More like the brief paroxysmal mania sometimes seen in cases of TLE or where the EEG is abnormal. Some might recall Teicher and Cole's original case series included one case with TLE and I think a further two with abnormal EEG.
One interesting side thread, there is also a great similarity in some cases to classical descriptions of severe premenstrual tension. In fact, there is one excellent early description of PMT that includes features some would say are pathognomonic of akathisia.
There might be too many cooks in the kitchen right now with "withdrawal studies", but if the skeptics push hard enough against the concept, they might get more than what they bargained for.
It can happen qualitatively, amongst new categories, or quantitatively, by way of severity. In the short term I'm generally in favor of qualitative creep, because multiple realizability remains a fact in cognitive science. It is the quantitative kind that I get suspicious of. I agree both are taking place here.
However, when the charge of concept creep was applied to the trauma field over the last decade, it actually resulted in that field refining and adopting some rather strict, theory-driven / mechanistic definitions. Those definitions are now stricter than both DSM-4 & 5. Contrary to pop-psychology, they are not equivalent to "I feel wronged by X-event, therefore, Criterion A is fulfilled". They instead require simultaneous triadic processes, loading on phenomenological, neurological, and interpersonal domains ontop of event-causation. Hence, multiple realizability. (E.g., see complex trauma operationalizations by APA Div. 56 after 2019).
Who remembers the late 90s & early 2000s, when it was held that the function of pharmacology was to suppress signs of child abuse? A generation of pre-adolescents were misdiagnosed with bipolar disorder in the early 2000s, given, child abuse & neglect is widely known to cause dissociative processes that, to a psychiatrist unfamiliar with Munchausen syndrome by proxy, would likely label it "bipolar disorder". Today, textbooks in history of psychology are terrified of publishing it, but it is widely known in the public record.
It's tricky to say if lamotrigine would help. It seems that the lamotrigine withdrawal syndrome is a variant of the same syndrome seen in antidepressants. Also we say "withdrawal", but whatever this syndrome is, some case are triggered at titration, others at taper but the syndrome remains the same so far as I can tell.
So far the most consistently beneficial treatment to have emerged seems to be benzodiazapines. Propranolol doesn't seem to do any harm but there doesn’t seem to be a clear-cut benefit. Gabapentin seems to make things worse or cause again the same syndrome in some people, which is very curious if this were to have anything to do with metabolism. I need more on the gabapentin though, not enough people try it. Lithium orotate seems to help but only in the standard 5mg.
Whatever the syndrome is, once set in motion the people suffering from it really do seem to develop an exquisite sensitivity to a wide variety of seemingly unrelated drugs.
I think the attrition to neurosis or a somatic disorder, is confabulation on the part of the psychiatrist. I don't blame them, the limbic dysregulation aspect of the syndrome mimics BPD and there is often a lot neuropathic problems and diffuse neurological signs. In short, it mimics exactly the sort of thing people would be sceptical of. The issue is once you take sufficient history and speak to family members (which I'd like to do more of) it becomes clear that the withdrawal syndrome is markedly different than the original mood disorder. Also the course doesn't fit, for a start some people recover entirely and what appeared to be personality traits vanish. Others follow a paroxysmal course that has a strong whiff of the biological about it. Some recover between taper drops. In short, its not the steady course of an enduring lifelong illness or personality trait. I think there is confounder here that makes it even harder for psychiatrists to spot, which is that among a population of people with mood disorders, you will find a lot of symptoms from before the drug that makes it look uncertain. Rates of neuroticism amomg such a population probably are higher for example. Its only when you really get amongst it that you realise how distinctive the syndrome is.
You know I've spent the better part of the last two years deep in a withdrawal group trying to figure this all out because I think it's much more interesting than people give it credit for.
I agree with you about the hyperbolic thing, it's vanishingly unlikely to beat just going slower.
You've underestimated the homogeneity though. It could be the population I'm involved with, perhaps more severe cases go to the effort of joining support groups but far from a menagerie, its really quite astounding how few and narrow the phenotypes are.
The CFS/post viral thing is the most interesting part. At least among the population I'm involved in most cases seem to fall under an acquired/organic neurasthenic rubric. So you would think it would follow that a great many are neurotic but this doesn't seem to be the case at all. And certainly I'm sufficiently insecure to have considered it. The thing is, I'm quite satisfied that the onset of this "CFS" syndrome is de novo in the vast majority of cases. Temporally it's very closely tied either to going on the drug or coming off it and despite many having mood disorder that put them on the drug, for most that mood disorder was not neurasthenia, nor anxiety neurosis.
Look, if you'd have asked me to guess before I got chin deep in it I would have absolutely predicted neurosis but it doesn't reconcile once you're deep enough in the qualitative data.
The other interesting angle is that closer questioning has revealed a supervising number of descriptions that shade very close to temporal lobe epilepsy.
There also seems to be some strange relationship to PMDD and possibly confusional psychosis.
Also, lupus. That's another weird wrinkle.
It's all very strange and there is great temptation to conclude that the problem was pre-existing but some cases are well into their middle years having suffered an all together very different problem.
Thank you for the excellent deconstruction of the nebulous and protean term, "withdrawal," Awais, as well as for your informed skepticism regarding various claims about the nature, frequency, and severity of antidepressant-related "withdrawal" syndromes [ARWS]. This is a topic that easily devolves into what I call, the 3 p's: polarization, personalization, and politicization. For example, suggesting that the vast majority of ARWS are neither severe nor prolonged often gets excoriated in social media as "Psychiatry's denial and minimization" of the problem.
Having reviewed this literature for a planned article in Psychiatric News, I am convinced that the kind of rigorous, controlled studies of ARWS that we need--including use of the standardized DESS scale--have simply not been done. Thus, our position as a profession ought to be one of cautious humility, not polemical certainty. Much more can and should be said, and I hope your readers will find the piece by Dr. Jonathan Henssler and me of some merit.
https://www.psychiatrictimes.com/view/antidepressant-withdrawal-syndromes-listening-to-the-patient-and-taking-it-slow
Our main conclusion: "To be clear, we want to acknowledge that whatever the usual frequency and duration of AWDS, some patients may experience severe and prolonged withdrawal symptoms that greatly interfere with activities of daily living, vocational function, and quality of life. These facts were not sufficiently emphasized in early clinical guidelines. Nevertheless, in our experience, most problems with antidepressant discontinuation can be averted by slow, careful, individualized tapering, often over several months, and—all other things being equal—by avoiding short half-life agents (eg, paroxetine and venlafaxine) in the first place."
Best regards,
Ron
Ronald W. Pies, MD
Outstandingly clear and helpful - thank you! It’s only anecdotal, but I’ve never observed craving for the discontinued antidepressant, whereas craving is a hallmark - just about pathognomonic - of discontinuation, ‘true’ withdrawal, of alcohol, tobacco or opiates. For what it’s worth, I think this is a meaningful difference between discontinuation phenomena and withdrawal, which describe distinct symptoms or states. Rapid reinstatement after abstinence is also something I haven’t met with antidepressants. And tolerance isn’t the same either. All of this argues that discontinuation/withdrawal is more than a semantic discussion.
(1)I found this summary helpful, exhaustive and somewhat exhausting. Hope that some of these categories will coalesce and others disappear, and not fall into some dreaded NOS category we love so much about DSM.
(2) It seems so hard not to avoid the problem that many words that mean different things to different people, thereby making agreements of what's testable and how to do so painfully difficult.
(3) I was struck by your heading of "what do we REALLY (my caps ) know?" - having to add "really", I guess distinguishes what follows from what we know - but not really. It seems you're wanting to make a distinction from, perhaps, what others may think but not truly know. A real tightrope.
(4) I'd just add that there is a constituency out there that thinks that using meds is almost always the wrong way to go, which makes it easy for them to add "withdrawal" to their list of psychiatric sins. If we were wrong to medicate in the first place, we've set the table for the "withdrawal," which that group REALLY knows is inevitable.
Yet *another* useful essay, bravo. Three thoughts leading to one strong recommendation:
1. I've seen very credible severe withdrawal with 1 mg steps at the end of a citalopram taper where larger steps earlier were not so bad. So regardless of receptor occupancy (but thanks for pointing out complexities there), tiny steps at the end are definitely needed by some.
2. But probably not many? Our small study found 90% of people in primary care did not even receive a prescription for the smallest available dose before stopping their antidepressant. They may have suffered but didn't go back on (or disappear; Phelps et al, JABFM, 2023). But who will be among those with horrible withdrawal -- whatever the basis --so as to know who really needs that many-months quasi-linear approach? Simple...
3. Start *everyone* with the smallest decrement possible. Many need it at the end anyway and this step protects everyone. It addresses concerns about worsening, lowering nocebo risks. Those who don't fear withdrawal can take faster or bigger steps on the way down; those who do can go by that same small decrement all the way. Anyone who gets symptoms on the way down should slow down. If each step is still rough, get a liquid (hardly ever necessary in my experience). I never had -- but know they're out there -- a patient go through what Adele Framer has said is needed for some, but for quite a few it was many, many months. Not years.
Oh and ever notice how serotonergic withdrawal symptoms look like bipolar mixed states? Withdrawal more common in people with bipolarity, I've always wondered? So that lamotrigine idea...
Thank you again as always, Awais!
Thank you Jim!
Severe withdrawal with tiny steps at the end… I don’t reject the possibility and I defer to your report; my main point is that whatever the mechanism for this phenomenon, receptor occupancy curve doesn’t explain it. Something other mechanism needs to be invoked, a super-sensitivity of some sort or whatever.
Your recommendations are quite sensible. And there is indeed a curious similarity between serotonergic withdrawal and mixed states!
Yes, they do look like mixed states. In fact, I would even go so far as to say many are mixed states by definition, but are they "bipolar mixed states"? I used to think so but now I'm not nearly as sure. The closest thing I've found is Michael Alan Taylor's description of "limbic sensitisation" syndrome in his little book on clinical rounds.
To be honest, having spent much of the ladt two years talking to people in these states, I'm struck by the level of lability. I think it shades closely towards temporal lobe and limbic abnormalities but not necessarily classic bipolar mania. More like the brief paroxysmal mania sometimes seen in cases of TLE or where the EEG is abnormal. Some might recall Teicher and Cole's original case series included one case with TLE and I think a further two with abnormal EEG.
One interesting side thread, there is also a great similarity in some cases to classical descriptions of severe premenstrual tension. In fact, there is one excellent early description of PMT that includes features some would say are pathognomonic of akathisia.
Had a blast reading, thank you
I agree with most of your predictions, but I do expect neuroticism to be a huge moderating factor
I wish we spent more time with the severe mental illness population but hey, what can you do
There might be too many cooks in the kitchen right now with "withdrawal studies", but if the skeptics push hard enough against the concept, they might get more than what they bargained for.
It can happen qualitatively, amongst new categories, or quantitatively, by way of severity. In the short term I'm generally in favor of qualitative creep, because multiple realizability remains a fact in cognitive science. It is the quantitative kind that I get suspicious of. I agree both are taking place here.
However, when the charge of concept creep was applied to the trauma field over the last decade, it actually resulted in that field refining and adopting some rather strict, theory-driven / mechanistic definitions. Those definitions are now stricter than both DSM-4 & 5. Contrary to pop-psychology, they are not equivalent to "I feel wronged by X-event, therefore, Criterion A is fulfilled". They instead require simultaneous triadic processes, loading on phenomenological, neurological, and interpersonal domains ontop of event-causation. Hence, multiple realizability. (E.g., see complex trauma operationalizations by APA Div. 56 after 2019).
Who remembers the late 90s & early 2000s, when it was held that the function of pharmacology was to suppress signs of child abuse? A generation of pre-adolescents were misdiagnosed with bipolar disorder in the early 2000s, given, child abuse & neglect is widely known to cause dissociative processes that, to a psychiatrist unfamiliar with Munchausen syndrome by proxy, would likely label it "bipolar disorder". Today, textbooks in history of psychology are terrified of publishing it, but it is widely known in the public record.
It's tricky to say if lamotrigine would help. It seems that the lamotrigine withdrawal syndrome is a variant of the same syndrome seen in antidepressants. Also we say "withdrawal", but whatever this syndrome is, some case are triggered at titration, others at taper but the syndrome remains the same so far as I can tell.
So far the most consistently beneficial treatment to have emerged seems to be benzodiazapines. Propranolol doesn't seem to do any harm but there doesn’t seem to be a clear-cut benefit. Gabapentin seems to make things worse or cause again the same syndrome in some people, which is very curious if this were to have anything to do with metabolism. I need more on the gabapentin though, not enough people try it. Lithium orotate seems to help but only in the standard 5mg.
Whatever the syndrome is, once set in motion the people suffering from it really do seem to develop an exquisite sensitivity to a wide variety of seemingly unrelated drugs.
I think the attrition to neurosis or a somatic disorder, is confabulation on the part of the psychiatrist. I don't blame them, the limbic dysregulation aspect of the syndrome mimics BPD and there is often a lot neuropathic problems and diffuse neurological signs. In short, it mimics exactly the sort of thing people would be sceptical of. The issue is once you take sufficient history and speak to family members (which I'd like to do more of) it becomes clear that the withdrawal syndrome is markedly different than the original mood disorder. Also the course doesn't fit, for a start some people recover entirely and what appeared to be personality traits vanish. Others follow a paroxysmal course that has a strong whiff of the biological about it. Some recover between taper drops. In short, its not the steady course of an enduring lifelong illness or personality trait. I think there is confounder here that makes it even harder for psychiatrists to spot, which is that among a population of people with mood disorders, you will find a lot of symptoms from before the drug that makes it look uncertain. Rates of neuroticism amomg such a population probably are higher for example. Its only when you really get amongst it that you realise how distinctive the syndrome is.