There are several wise comments here, I think. As a parent/caregiver who participates in some private social media groups where other parent/caregivers gather, I can unequivocally say the experiences that parents report of their children's responses to the same medication varies considerably within diagnoses. For that reason, as already pointed out by others, the mean response is a poor indicator of any individual patient's response. For this reason, parent/caregivers in the mentioned social media groups tend to avoid recommending a particular medication that is working well for their child to other parents. I have also noticed clinicians more often using phrases like "bipolar spectrum disorders" and "schizophrenia spectrum disorders" in what appears to be a nod in the direction of recognizing what we lump under a particular diagnosis is probably an assortment of different illnesses with similar symptoms. Likewise, perhaps depression is a spectrum of disorders with similar symptoms with the variability in the effectiveness of a medication among patients being partly related to those patients actually having different illnesses. There are so many unanswered questions in the complicated field of psychiatry that the safest position may be we simply don't know.
1. I was very surprised that there was not an in-depth discussion of antidepressant-induced mania and/or psychosis. Essentially any antidepressant can induce mania (unless it’s also an anti-manic or antipsychotic).
2. The majority of antidepressant suicides are people in their late teens and early 20s with a personal and/or family history of bipolarity or psychotic disorders. The state where a person is at highest risk of suicide is by far the bipolar mixed state especially if it is accompanied by psychotic features.
3. Psychiatry has been very bad at admitting to patients that despite our best judgment we sometimes make treatment choices that leave them worse off. That’s a fact of medicine to some extent. But our response is not.
4. When antidepressants induce mania the patient rarely gets the immediate attention from a clinician who recognizes the signs and more importantly the severity of this complication and how fragile it is going to make the patient for weeks or months. Often it is written off as initiation jitters and the dose is increased or a benzo script is written for a week or two. We aren’t rigorously honest about the dependency and the misery of discontinuation from long term use. We don’t adequately value the cost of sexual dysfunction on human wellbeing and drive to live or of weight gain in a person whose worst fear is feeling like they are fat and unattractive.
5. As a clinician I do get a bit tired of the endless ‘do SSRIs work’ slugfest because to me it is a) delusional but more importantly b) misses the whole point. First, from working in psych, I don’t see how a person could honestly say they don’t think they work. They obviously work. I simply don’t see how a person working with actual patients could say this with a straight face.
I’m open to the argument that they don’t work amazingly well, response rates aren’t great, relief is very partial, their side effects can be pretty bad for quality of life, the initiation period is uncomfortable and sometimes volatile, use in children and teens is not well studied at all, they often induce mania in latent bipolar or bipolar 2 presenting as chronic depression, it’s unclear (though many studies do find a prophylactic effect against future episodes) whether it leads to less of the lifespan being spent in states of depression. They’re overprescribed and used for things that obviously either require therapy or are normal human grief and transition processes that would probably resolve on their own. In fact, I think all these things are true.
Maybe you don’t see patients do well in the long-term. You might feel like it is a temporary aid with a long tail of dependence and side effects. You might view it as a dangerously understudied use of drugs that work most of the time but make things *way* worse frequently enough that they’re simply not worth messing with. All of these are valid arguments and thoughts I’ve had myself.
6. But to deny that they work would be to deny that the sky appears blue on a sunny summer’s day. It is interesting in that he does have actual clinical experience - I have noted that those making this argument and publishing the big popular books about how antidepressants don’t do anything are almost as a rule not written by medical professionals.
They’re pretty much always people from the social sciences or economics or physics world trying to come in, generally with a pretty obvious pre-existing bias, who insist that scientific rigor requires the exclusion of all clinical knowledge and experience we train in for years. I suppose David Healy is an exception, but I don’t think he says antidepressants are so dangerous as to be worthless. Certainly he doesn’t have a high opinion of them and is appalled by how widely they’re used but he doesn’t try to argue that they don’t help individual patients feel less depressed.
This argument seems much subtler, but I wish he would come out and state it directly rather than go full Emperor’s New Drugs. Because he’s clearly brilliant, so he knows the data and knows the drugs do work. After all, if they were inactive brand name sugar pills, then it would be bizarre to blame them for something as horrible as a suicide attempt. It seems highly unlikely to me that drugs with the potential to induce manic psychosis just aren’t doing much in the brain. Clearly they’re doing something!
If he wants to argue that the harm they’re doing is so severe, random, and greater in magnitude compared to the good, that’s a genuinely worthwhile discussion to have.
His argument seems to be that they don’t work nearly as much as we think and that they aren’t the risk-free drugs they can be portrayed as - and that this volatility and unpredictability happens widely and unpredictably enough that we should be much more cautious in our use of these drugs and analyzing assumptions about the long-term and population effects much more than we currently do.
7. I guess my litmus test for critical/anti-psychiatric value is does the argument still follow logical principles of non self-contradiction. I don’t expect somebody to articulate a full theory or anything, and I don’t expect a level of biological science knowledge equal to that of a medical professional.
But for me if you’re going to criticize antidepressants either argue:
A) they don’t really work
OR
B) they have horrible side effects and cause people to kill themselves and be addicted forever
Like....it just can’t be both ways. It seems like this critique refuses to pick a side and whenever it gets solidly refuted it transmutes or waveform collapses into whichever argument is left standing.
I refuse to believe that they’re placebos that also happen to cause otherwise non-suicidal people to kill themselves. I don’t know the secrets of the brain, but I do know that they can’t be *both* of those things.
1. Thanks for bringing up antidepressant-induced mania. I suspect that a substantial proportion of antidepressant-induced suicidality may indeed be related to unrecognized bipolarity. It’s seems like a plausible hypothesis (although research evidence on this kinda mixed too)
2. Yes
3. Agreed
4. Agreed
5. Well, we aren’t talking about antidepressant efficacy here, so we haven’t directly addressed the relevant consideration. I do think antidepressants work reasonably well for many depressed individuals, although a big chunk don’t respond very well as we all know. I made the case for efficacy in this post: https://awaisaftab.substack.com/p/the-case-for-antidepressants-in-2022 (Martin has a more skeptical assessment of their efficacy based on the average difference from placebo & uncertainty around treatment effect heterogeneity in clinical trials, but I don’t think he’ll say that they work for no one.)
6. Psychoactive substances don’t have to be inert placebos to be ineffective for a condition & cause adverse effects. I mean, take Haldol. It does a lot in the brain and produces many adverse effects. But it’s likely not gonna beat placebo for depression (might even be worse than placebo) unless it’s psychotic depression. As I mentioned in 5, I do think antidepressants work, but “clearly they are doing something” is not by itself a persuasive argument for efficacy.
1. As a person who had mania induced by an antidepressant twice as a college student misdiagnosed with MDD and not bipolar, it’s something I’m always on the watch for in my patients.
Working with young adults it happens enough that even in a short career I’ve seen it happen 5-10x and I’m always grateful for my own experience as I do feel like it helps me recognize both how different induced mania is from a rough initiation syndrome and the severity.
It’s often written off as something exceedingly rare that is an adverse side effect but not a psych emergency; IME it’s not that rare if you’re working with a younger population and it is an *incredibly* severe psych emergency where every hour carries a significant risk of suicide.
The person doesn’t know what’s happening to them, they can’t sleep or eat, the doctor is telling them it’s just side effects, you start acting strangely and push friends and family away, fail classes/drop-out of school...and yet often the treatment plan given is ‘let’s increase the dose but add a benzo for a week or two’.
Got this as a patient and seen it done as a clinician. Should be malpractice imo. If you have to put somebody on benzos bc the AD you prescribed them has literally made it impossible to sleep a single hour for 3 days...you likely put them on the wrong drug to begin with. I took the higher dose for one day and it felt like what I can only imagine something like a gram of methamphetamine would cause one to feel.
That’s the most manic I’ve ever been in my life from 3 days of mirtazapine. I stopped the med immediately after that, if I had taken it one more day or not had supportive friends or the benzos I would have been hospitalized. I took the benzos until the mania went away but that 2 week period was almost as harrowing as my major non-drug related, ~6 month long manic psychosis.
So yeah I would never argue latent/undiagnosed/subclinical bipolar with manic induction is responsible for 100% of AD-related suicides - there are a lot of factors involved in suicide and there are absolutely other feelings or side effects or aspects of ADs that could conceivably lead to suicidality.
I do think that at least half of the cases are manic induction with very poor follow-up treatment, that we should screen for bipolar before prescribing ADs to any young person, and if a person does have a manic episode from an AD they should get an incredibly thorough work up to determine whether they need treatment for bipolar going forward after the drug-induced episode is resolved. So while it’s really not monocausal to me it does seem by far the most plausible primary cause and so I did find it odd that it wasn’t mentioned whatsoever.
5. Yeah, a bit off topic, and I did read your post about that (and really liked it). In fact I was curious a bit why you didn’t push back at some points about the data.
I totally understand that efficacy isn’t the topic or thrust of the discussion, but low/zero efficacy was one of his main supporting points for creating a new black box warning. Since you’re both an AD skeptic but also somebody who has dug deeply into the data and knows that on the whole the data on their efficacy consistently finds that they have a small but significant benefit patients on the whole and a larger benefit the more severe the depression, I was a little confused that when he was (I’ll be frank in my criticism here) cherry-picking meta analyses for ones that support his argument that you didn’t push back and say yes look the data is messy, the effect size is small, but we do know that these things work and we’d have to pry them out of the hands of a lot of people who feel they saved their lives. You also noted - something I’ve always felt but really, really appreciated reading from somebody more experienced - that the RCTs are of questionable value bc of the deeply flawed (and just bizarro) instruments that we’ve decided are the best/most objective rating scales for. depression. The HAM-D is okay (I guess) and the BDI is horrible. The BDI is a terrible scale that is *extremely* focused on sex and sleep, so if the drug doesn’t improve those, the BDI score isn’t going to say it’s effective as an antidepressant . It also means the sexual side effects statistically cancel out the antidepressant benefits inside the scale, which is just a really poorly designed instrument.
Also my bad - I switched the HAM-D and BDI around. was just looking through HAM-D again - so many things that many ppl with depression don’t suffer from and that we know antidepressants don’t help! So weird man. I really wonder how this thing got so popular.
Oh also not to quibble, but Haldol *does* work. Again, it’s the kind of thing where I just don’t think that’s the point. The problem with Haldol isn’t that it doesn’t ever work - it’s just as good as the other 1st gene. The problem with Haldol that it makes you feel terrible, shortens your lifespan, and gives you horrible irreversible side effects. Similarly, the “are antidepressants the best way to treat moderate depression given their side effects and risks?” is a conversation worth having while “shouldn’t we consider that antidepressants could be both placebos and dangerous neurotoxins since it could theoretically be possible?” isn’t (at least for me).
Haldol works indeed, for psychosis, and I use it often enough.
I also agree that “antidepressants are placebos but also dangerous neurotoxins” isn’t a conversation worth having, which is why that is not the conversation I went for.
PS. I have a Q&A with Peter Kramer planned for this Sunday. I think that will be more to your liking 🙃
I appreciate you sharing your own personal experience — and very instructive example! — as well as your experience working with patients. You are right that there should’ve been a more explicit discussion of antidepressant-induced mania. It’s a very important cause of AD suicidality for sure, esp in children and young adults. I’m not a child & adolescent psychiatrist. The few cases of AD-related suicidality I’ve seen weren’t quite explainable by mania, although with psychological agitation it can be hard to tell. One person who had experienced suicidality twice on two different trials on Wellbutrin ended up doing quite well on an SSRI and a stimulant (for ADHD).
Regarding not pushing back, I didn’t quite see Dr Ploederl as making strong statements that they are merely placebo or that they don’t work or that we shouldn’t use them, so I didn’t feel the need. One point where he says something about minimal efficacy, he’s talking about efficacy in children and adolescents, and I didn’t push back there because the RCT data for children & adolescents is even weaker than it is for adults. Sure, they can still be clinically useful, but it would’ve required a detour that I didn’t wanna take. Plus both of us nodded towards the fact that we have disagreements on the issue of efficacy and I didn’t think I needed to litigate that 😊
For sure. Appreciate all the detailed responses. I still enjoyed the post and Dr. Ploederl’s work. My experience with the under 18 population is very, very limited and I don’t know the study data very well for children and younger teens (except to know that for most drugs it barely/doesn’t really exist).
I apologize if I got a bit carried away in my criticism. Overall I am so glad that providers are getting involved in and studying the risks of and insisting on honesty with patients and the public about psychiatric medication. There is a great need for this kind of work and skeptical attitude from inside the profession.
As a young person & patient I tried to give my doctors the benefit of the doubt figuring that they just didn’t know about things because the drugs were too new. anyways I had a terrible home life as a teen so I didn’t know whether I was depressed bc ADs weren’t working or because no drug was going to make it tolerable to be living in a house with a parent who also had bipolar but is in total denial.
She spent about half her time on unplanned impulsive trips to Europe & Florida that forced me & my brother to learn (or rather, not learn). The other half she would come back and be go into a narcissistic rage about anything not perfect about the house, us not getting perfect grades, how much our friends sucked, that we were drug addicts, my girlfriend was a gold digger, we were aimless and dependent on my dad to take care of ourselves, and just generally being a narcissistic tyrant at home and acting know she was a reliable authority figure and totally isn’t going to just go to Germany again with just a post-it note after getting home from school to inform me that she would be gone for an unknown period of time. (happened like 3x total)
But anyways, SSRIs were marketed deceptively and we did go along with it to a large extent. Prescribers do deserve some discretion in how we present medication but there are certain basic properties and risks that need to *always* be included even if it might cause the person to reject a med that would help them a lot.
There are company documents about SSRI problems from the early 90s - triggering mania/psychosis, suicide case studies, the development of strong physical dependence with a severe withdrawal syndrome, interdose withdrawal in shorter half-life drugs like Paxil and Effexor, agitation, migraines - and it seems impossible to me that by the early 2000s any psychiatrist could still attempt to maintain that physical dependence didn’t exist.
If they didn’t know, that’s an almost negligent level of disinterest in continuing education and patient feedback and lack of curiosity about the most basic properties, common adverse effects, and serious risks posed by their most commonly prescribed drugs. Yet my doctors withheld all of this knowledge from me.
Did they consider it unimportant that I know? Were they in a rush? Did they worry that I would refuse medication if I was properly informed about its risks?
As a clinician, to all but the last question, I would say yes. It’s all of these things and rarely just one at a time.
How much was in my true interest and how much was just coercion and deception to make me compliant without actually involving me in my care because that will be more time-consuming and frustrating?
Most of it was done in my interest, but it could have been done in ways that didn’t leave me with memories of terror and humiliation. In the hospitals there just aren’t enough staff for you to be treated like a human being. And the decision to exclude me from even the basics of the treatment plan (such as my daily meds) led to me being placed on many drugs with pretty unbearable side effects.
Coercion was warranted - I was extremely manic & psychotic and highly paranoid, and in retrospect I made life miserable for the staff at every hospital I went through. But dishonesty doesn’t sit well with me. Even if I’m going to have to take a pill or get an injection, I’d still rather know what it is and be able to look it up on a cell phone or a computer.
I honestly have no recollection of what meds I was on for the first weeks because they kept changing the antipsychotic and didn’t bother to keep me in the loop. Tbf I was pretty out of it bc of all the sedation and side effects but I wasn’t unconscious or like totally incapable of communication and decision making.
___
After a very lengthy series of hospitalizations I had such mixed feelings about psychiatry. On one hand I was grateful and knew I to acknowledge it had saved my life and restored me to sanity, but the experience of being hospitalized and put on a series of meds without my knowledge or consent was so hellish and my care was managed so poorly that the cost I paid for having my manic episode resolved was literally years.
Years of depression, self-hatred, rage and terror at the thought of being hospitalized, distrust of medical professionals, and learning how to deal with the flashbacks from experiences in the hospital that are so deeply repressed I literally can’t access them voluntarily.
Yet at the same time, coming from a family of serious mental illness and denial and abuse and the agony of the disease when it’s untreated, being hospitalized changed the trajectory of my life for the better. It took 5-10 years for me to see that, but I realized that with medication and healthy behavior, my life could be relatively normal - but without it I would likely follow my mom down the path to self-destruction and ruining my career and relationship.
This is why I get very nervous when psychiatrists publicly question the value of broad classes of medication, but it’s more of a personal thing that I have to get over. People like my family absolutely love and weaponize studies and prominent figures that oppose medication as confirmation that they were just keeping me safe & being prudent and imply that my life is out of control and needs an intervention because I’m taking multiple psych meds.
But getting the truth out to patients is more important than giving more ammunition to the crazies. In the long run I hope greater honesty will take some of the steam out of the fiercer opposition and suspicion psychiatry faces in the US.
I probably will enjoy your talk with Dr. Kramer but I just wanted to say that I did appreciate this post and guest
I would like to thank Dr. Plöderl and Dr. Aftab for an enlightening discussion of this complex and controversial topic. As a (now retired) specialist in adult mood disorders, I acknowledge the uncertainties inherent in studies of SSRIs/SNRIs and suicidality; however, I believe the preponderance of the evidence does not support the claim that these agents increase suicidality (a vague term, to be sure) or completed suicide in adult patients 25 years and older. And, there are at least some data pointing to a reduction of suicidality (suicide attempts) in antidepressant-treated patients age 65 and older, as the discussion with Dr. Plöderl notes (Stone et al, 2009)
A recent, carefully-done, observational study by Lagerberg et al generally supports the above conclusions. The authors write:
"We found an increased risk of suicidal behaviour among individuals aged below 25 years who were treated with an SSRI after a depression diagnosis as compared to those who were not. We found no evidence of an effect among older age categories. The results are similar to those from RCTs, lending validity to this observational study. However, the issue of unmeasured confounding remains, and studies from data sources with more detailed information on confounders—notably depression severity—are called for. Our results confirm that individuals with a history of suicidal behaviour are a high-risk group, but we found no evidence that SSRI initiation conferred elevated risks of suicidal behaviour in this subgroup." [1]
Like all observational studies, this one has both strengths and limitations, as the authors acknowledge.
Finally, I would like to comment on the statement that,
"Put differently, the claim that “antidepressants are live-saving” is not supported by the evidence, at least not for the average patient."
Clinicians, of course, do not treat "the average" patient or an aggregate of patients. Each patient we treat is unique. Having treated many hundreds of severely depressed patients over the course of more than 25 years--many of whom I followed closely for 2, 3 or more years--I have certainly seen many whose suicidal ideation or impulses diminished greatly in the course of antidepressant treatment.
I am well aware that such observations are often dismissed as "anecdotal" or the result of an unacknowledged "placebo effect." Yet when this observation is made dozens and dozens of times, in patients whose depression did not respond to two or three adequate antidepressant trials, but did respond to the fourth agent, the clinician is naturally reluctant to accept the, "it was just a placebo effect" explanation. In such cases, I believe it is entirely appropriate to say that the totality of treatment--antidepressant plus the therapeutic alliance, probably acting synergistically--was indeed "life-saving."
Thanks again to you both for this useful discussion!
Respectfully,
Ronald W. Pies, MD
Professor Emeritus of Psychiatry
1. Lagerberg, T., Matthews, A.A., Zhu, N. et al. Effect of selective serotonin reuptake inhibitor treatment following diagnosis of depression on suicidal behaviour risk: a target trial emulation. Neuropsychopharmacol. 48, 1760–1768 (2023). https://doi.org/10.1038/s41386-023-01676-3
Something that I really would love to see an article or interview with you about Dr. Aftab: the scientific validity (or rather invalidity) of psychiatric nosology.
As others touched on - MDD is heterogenous, and as others have argued elsewhere - we have subsumed many bipolar spectra and cyclothymic temperaments under MDD.
It certainly dovetails with the observation of antidepressant induced suicidality, specifically that many individuals who are becoming activated, anxious, suicidal with antidepressants. As many will argue, the index episode of bipolar is depression, and the onset of illness is a median of 19, a full decade before MDD typically. It would be no wonder then that the signal for suicidality would strengthen with younger age.
Our ability to see not only adverse effects, but to understand genuine efficacy (antidepressants, as you have touched on, significantly help a certain proportion of patients, and it is often obscured by the “average” effect sizes we see) has been hampered by the rigidity of the DSM in many ways, and specifically the FDA’s reliance on broad, often invalid DSM diagnoses in order to seek approval.
For me the overlap between the period of suicidality and the period where bipolar symptoms begin to emerge is far too coincidental. Induced manic episodes frequently lead to suicide attempts and usually aren’t noticed by clinicians unless you see the person a few days later or they have the insight to recognize that something has gone *very* wrong and reach out to you.
But many psychs in that situation are just like “the first week can be hard, just hang in there, here’s some trazodone/benzos to help you sleep, in a week or two you’ll be feeling better”. More likely in a week or two you’ll be dead if you keep taking the pills. Fortunately most ppl realize that they’re making things way worse, unfortunately not all.
Always important to remember time from first bipolar episode to diagnosis average is like 8 years. The chance of a person who is 20 years old having an accurate diagnosis for their emerging bipolar is basically zero. Many bipolar 1 and most bipolar 2 find out what mania is when they take ADs for the first time.
No hating on ADs here. Once I got on mood stabilizers I had one added back on and it helps immensely. But bipolar is underdiagnosed until you get to like late 20s/early 30s and should always be considered whenever there is any kind of activating/agitating crisis induced by an AD.
There are several wise comments here, I think. As a parent/caregiver who participates in some private social media groups where other parent/caregivers gather, I can unequivocally say the experiences that parents report of their children's responses to the same medication varies considerably within diagnoses. For that reason, as already pointed out by others, the mean response is a poor indicator of any individual patient's response. For this reason, parent/caregivers in the mentioned social media groups tend to avoid recommending a particular medication that is working well for their child to other parents. I have also noticed clinicians more often using phrases like "bipolar spectrum disorders" and "schizophrenia spectrum disorders" in what appears to be a nod in the direction of recognizing what we lump under a particular diagnosis is probably an assortment of different illnesses with similar symptoms. Likewise, perhaps depression is a spectrum of disorders with similar symptoms with the variability in the effectiveness of a medication among patients being partly related to those patients actually having different illnesses. There are so many unanswered questions in the complicated field of psychiatry that the safest position may be we simply don't know.
Thank you Joseph!
1. I was very surprised that there was not an in-depth discussion of antidepressant-induced mania and/or psychosis. Essentially any antidepressant can induce mania (unless it’s also an anti-manic or antipsychotic).
2. The majority of antidepressant suicides are people in their late teens and early 20s with a personal and/or family history of bipolarity or psychotic disorders. The state where a person is at highest risk of suicide is by far the bipolar mixed state especially if it is accompanied by psychotic features.
3. Psychiatry has been very bad at admitting to patients that despite our best judgment we sometimes make treatment choices that leave them worse off. That’s a fact of medicine to some extent. But our response is not.
4. When antidepressants induce mania the patient rarely gets the immediate attention from a clinician who recognizes the signs and more importantly the severity of this complication and how fragile it is going to make the patient for weeks or months. Often it is written off as initiation jitters and the dose is increased or a benzo script is written for a week or two. We aren’t rigorously honest about the dependency and the misery of discontinuation from long term use. We don’t adequately value the cost of sexual dysfunction on human wellbeing and drive to live or of weight gain in a person whose worst fear is feeling like they are fat and unattractive.
5. As a clinician I do get a bit tired of the endless ‘do SSRIs work’ slugfest because to me it is a) delusional but more importantly b) misses the whole point. First, from working in psych, I don’t see how a person could honestly say they don’t think they work. They obviously work. I simply don’t see how a person working with actual patients could say this with a straight face.
I’m open to the argument that they don’t work amazingly well, response rates aren’t great, relief is very partial, their side effects can be pretty bad for quality of life, the initiation period is uncomfortable and sometimes volatile, use in children and teens is not well studied at all, they often induce mania in latent bipolar or bipolar 2 presenting as chronic depression, it’s unclear (though many studies do find a prophylactic effect against future episodes) whether it leads to less of the lifespan being spent in states of depression. They’re overprescribed and used for things that obviously either require therapy or are normal human grief and transition processes that would probably resolve on their own. In fact, I think all these things are true.
Maybe you don’t see patients do well in the long-term. You might feel like it is a temporary aid with a long tail of dependence and side effects. You might view it as a dangerously understudied use of drugs that work most of the time but make things *way* worse frequently enough that they’re simply not worth messing with. All of these are valid arguments and thoughts I’ve had myself.
6. But to deny that they work would be to deny that the sky appears blue on a sunny summer’s day. It is interesting in that he does have actual clinical experience - I have noted that those making this argument and publishing the big popular books about how antidepressants don’t do anything are almost as a rule not written by medical professionals.
They’re pretty much always people from the social sciences or economics or physics world trying to come in, generally with a pretty obvious pre-existing bias, who insist that scientific rigor requires the exclusion of all clinical knowledge and experience we train in for years. I suppose David Healy is an exception, but I don’t think he says antidepressants are so dangerous as to be worthless. Certainly he doesn’t have a high opinion of them and is appalled by how widely they’re used but he doesn’t try to argue that they don’t help individual patients feel less depressed.
This argument seems much subtler, but I wish he would come out and state it directly rather than go full Emperor’s New Drugs. Because he’s clearly brilliant, so he knows the data and knows the drugs do work. After all, if they were inactive brand name sugar pills, then it would be bizarre to blame them for something as horrible as a suicide attempt. It seems highly unlikely to me that drugs with the potential to induce manic psychosis just aren’t doing much in the brain. Clearly they’re doing something!
If he wants to argue that the harm they’re doing is so severe, random, and greater in magnitude compared to the good, that’s a genuinely worthwhile discussion to have.
His argument seems to be that they don’t work nearly as much as we think and that they aren’t the risk-free drugs they can be portrayed as - and that this volatility and unpredictability happens widely and unpredictably enough that we should be much more cautious in our use of these drugs and analyzing assumptions about the long-term and population effects much more than we currently do.
7. I guess my litmus test for critical/anti-psychiatric value is does the argument still follow logical principles of non self-contradiction. I don’t expect somebody to articulate a full theory or anything, and I don’t expect a level of biological science knowledge equal to that of a medical professional.
But for me if you’re going to criticize antidepressants either argue:
A) they don’t really work
OR
B) they have horrible side effects and cause people to kill themselves and be addicted forever
Like....it just can’t be both ways. It seems like this critique refuses to pick a side and whenever it gets solidly refuted it transmutes or waveform collapses into whichever argument is left standing.
I refuse to believe that they’re placebos that also happen to cause otherwise non-suicidal people to kill themselves. I don’t know the secrets of the brain, but I do know that they can’t be *both* of those things.
1. Thanks for bringing up antidepressant-induced mania. I suspect that a substantial proportion of antidepressant-induced suicidality may indeed be related to unrecognized bipolarity. It’s seems like a plausible hypothesis (although research evidence on this kinda mixed too)
2. Yes
3. Agreed
4. Agreed
5. Well, we aren’t talking about antidepressant efficacy here, so we haven’t directly addressed the relevant consideration. I do think antidepressants work reasonably well for many depressed individuals, although a big chunk don’t respond very well as we all know. I made the case for efficacy in this post: https://awaisaftab.substack.com/p/the-case-for-antidepressants-in-2022 (Martin has a more skeptical assessment of their efficacy based on the average difference from placebo & uncertainty around treatment effect heterogeneity in clinical trials, but I don’t think he’ll say that they work for no one.)
6. Psychoactive substances don’t have to be inert placebos to be ineffective for a condition & cause adverse effects. I mean, take Haldol. It does a lot in the brain and produces many adverse effects. But it’s likely not gonna beat placebo for depression (might even be worse than placebo) unless it’s psychotic depression. As I mentioned in 5, I do think antidepressants work, but “clearly they are doing something” is not by itself a persuasive argument for efficacy.
Thanks for the detailed reply!
1. As a person who had mania induced by an antidepressant twice as a college student misdiagnosed with MDD and not bipolar, it’s something I’m always on the watch for in my patients.
Working with young adults it happens enough that even in a short career I’ve seen it happen 5-10x and I’m always grateful for my own experience as I do feel like it helps me recognize both how different induced mania is from a rough initiation syndrome and the severity.
It’s often written off as something exceedingly rare that is an adverse side effect but not a psych emergency; IME it’s not that rare if you’re working with a younger population and it is an *incredibly* severe psych emergency where every hour carries a significant risk of suicide.
The person doesn’t know what’s happening to them, they can’t sleep or eat, the doctor is telling them it’s just side effects, you start acting strangely and push friends and family away, fail classes/drop-out of school...and yet often the treatment plan given is ‘let’s increase the dose but add a benzo for a week or two’.
Got this as a patient and seen it done as a clinician. Should be malpractice imo. If you have to put somebody on benzos bc the AD you prescribed them has literally made it impossible to sleep a single hour for 3 days...you likely put them on the wrong drug to begin with. I took the higher dose for one day and it felt like what I can only imagine something like a gram of methamphetamine would cause one to feel.
That’s the most manic I’ve ever been in my life from 3 days of mirtazapine. I stopped the med immediately after that, if I had taken it one more day or not had supportive friends or the benzos I would have been hospitalized. I took the benzos until the mania went away but that 2 week period was almost as harrowing as my major non-drug related, ~6 month long manic psychosis.
So yeah I would never argue latent/undiagnosed/subclinical bipolar with manic induction is responsible for 100% of AD-related suicides - there are a lot of factors involved in suicide and there are absolutely other feelings or side effects or aspects of ADs that could conceivably lead to suicidality.
I do think that at least half of the cases are manic induction with very poor follow-up treatment, that we should screen for bipolar before prescribing ADs to any young person, and if a person does have a manic episode from an AD they should get an incredibly thorough work up to determine whether they need treatment for bipolar going forward after the drug-induced episode is resolved. So while it’s really not monocausal to me it does seem by far the most plausible primary cause and so I did find it odd that it wasn’t mentioned whatsoever.
5. Yeah, a bit off topic, and I did read your post about that (and really liked it). In fact I was curious a bit why you didn’t push back at some points about the data.
I totally understand that efficacy isn’t the topic or thrust of the discussion, but low/zero efficacy was one of his main supporting points for creating a new black box warning. Since you’re both an AD skeptic but also somebody who has dug deeply into the data and knows that on the whole the data on their efficacy consistently finds that they have a small but significant benefit patients on the whole and a larger benefit the more severe the depression, I was a little confused that when he was (I’ll be frank in my criticism here) cherry-picking meta analyses for ones that support his argument that you didn’t push back and say yes look the data is messy, the effect size is small, but we do know that these things work and we’d have to pry them out of the hands of a lot of people who feel they saved their lives. You also noted - something I’ve always felt but really, really appreciated reading from somebody more experienced - that the RCTs are of questionable value bc of the deeply flawed (and just bizarro) instruments that we’ve decided are the best/most objective rating scales for. depression. The HAM-D is okay (I guess) and the BDI is horrible. The BDI is a terrible scale that is *extremely* focused on sex and sleep, so if the drug doesn’t improve those, the BDI score isn’t going to say it’s effective as an antidepressant . It also means the sexual side effects statistically cancel out the antidepressant benefits inside the scale, which is just a really poorly designed instrument.
Also my bad - I switched the HAM-D and BDI around. was just looking through HAM-D again - so many things that many ppl with depression don’t suffer from and that we know antidepressants don’t help! So weird man. I really wonder how this thing got so popular.
Oh also not to quibble, but Haldol *does* work. Again, it’s the kind of thing where I just don’t think that’s the point. The problem with Haldol isn’t that it doesn’t ever work - it’s just as good as the other 1st gene. The problem with Haldol that it makes you feel terrible, shortens your lifespan, and gives you horrible irreversible side effects. Similarly, the “are antidepressants the best way to treat moderate depression given their side effects and risks?” is a conversation worth having while “shouldn’t we consider that antidepressants could be both placebos and dangerous neurotoxins since it could theoretically be possible?” isn’t (at least for me).
Haldol works indeed, for psychosis, and I use it often enough.
I also agree that “antidepressants are placebos but also dangerous neurotoxins” isn’t a conversation worth having, which is why that is not the conversation I went for.
PS. I have a Q&A with Peter Kramer planned for this Sunday. I think that will be more to your liking 🙃
I appreciate you sharing your own personal experience — and very instructive example! — as well as your experience working with patients. You are right that there should’ve been a more explicit discussion of antidepressant-induced mania. It’s a very important cause of AD suicidality for sure, esp in children and young adults. I’m not a child & adolescent psychiatrist. The few cases of AD-related suicidality I’ve seen weren’t quite explainable by mania, although with psychological agitation it can be hard to tell. One person who had experienced suicidality twice on two different trials on Wellbutrin ended up doing quite well on an SSRI and a stimulant (for ADHD).
Regarding not pushing back, I didn’t quite see Dr Ploederl as making strong statements that they are merely placebo or that they don’t work or that we shouldn’t use them, so I didn’t feel the need. One point where he says something about minimal efficacy, he’s talking about efficacy in children and adolescents, and I didn’t push back there because the RCT data for children & adolescents is even weaker than it is for adults. Sure, they can still be clinically useful, but it would’ve required a detour that I didn’t wanna take. Plus both of us nodded towards the fact that we have disagreements on the issue of efficacy and I didn’t think I needed to litigate that 😊
For sure. Appreciate all the detailed responses. I still enjoyed the post and Dr. Ploederl’s work. My experience with the under 18 population is very, very limited and I don’t know the study data very well for children and younger teens (except to know that for most drugs it barely/doesn’t really exist).
I apologize if I got a bit carried away in my criticism. Overall I am so glad that providers are getting involved in and studying the risks of and insisting on honesty with patients and the public about psychiatric medication. There is a great need for this kind of work and skeptical attitude from inside the profession.
As a young person & patient I tried to give my doctors the benefit of the doubt figuring that they just didn’t know about things because the drugs were too new. anyways I had a terrible home life as a teen so I didn’t know whether I was depressed bc ADs weren’t working or because no drug was going to make it tolerable to be living in a house with a parent who also had bipolar but is in total denial.
She spent about half her time on unplanned impulsive trips to Europe & Florida that forced me & my brother to learn (or rather, not learn). The other half she would come back and be go into a narcissistic rage about anything not perfect about the house, us not getting perfect grades, how much our friends sucked, that we were drug addicts, my girlfriend was a gold digger, we were aimless and dependent on my dad to take care of ourselves, and just generally being a narcissistic tyrant at home and acting know she was a reliable authority figure and totally isn’t going to just go to Germany again with just a post-it note after getting home from school to inform me that she would be gone for an unknown period of time. (happened like 3x total)
But anyways, SSRIs were marketed deceptively and we did go along with it to a large extent. Prescribers do deserve some discretion in how we present medication but there are certain basic properties and risks that need to *always* be included even if it might cause the person to reject a med that would help them a lot.
There are company documents about SSRI problems from the early 90s - triggering mania/psychosis, suicide case studies, the development of strong physical dependence with a severe withdrawal syndrome, interdose withdrawal in shorter half-life drugs like Paxil and Effexor, agitation, migraines - and it seems impossible to me that by the early 2000s any psychiatrist could still attempt to maintain that physical dependence didn’t exist.
If they didn’t know, that’s an almost negligent level of disinterest in continuing education and patient feedback and lack of curiosity about the most basic properties, common adverse effects, and serious risks posed by their most commonly prescribed drugs. Yet my doctors withheld all of this knowledge from me.
Did they consider it unimportant that I know? Were they in a rush? Did they worry that I would refuse medication if I was properly informed about its risks?
As a clinician, to all but the last question, I would say yes. It’s all of these things and rarely just one at a time.
How much was in my true interest and how much was just coercion and deception to make me compliant without actually involving me in my care because that will be more time-consuming and frustrating?
Most of it was done in my interest, but it could have been done in ways that didn’t leave me with memories of terror and humiliation. In the hospitals there just aren’t enough staff for you to be treated like a human being. And the decision to exclude me from even the basics of the treatment plan (such as my daily meds) led to me being placed on many drugs with pretty unbearable side effects.
Coercion was warranted - I was extremely manic & psychotic and highly paranoid, and in retrospect I made life miserable for the staff at every hospital I went through. But dishonesty doesn’t sit well with me. Even if I’m going to have to take a pill or get an injection, I’d still rather know what it is and be able to look it up on a cell phone or a computer.
I honestly have no recollection of what meds I was on for the first weeks because they kept changing the antipsychotic and didn’t bother to keep me in the loop. Tbf I was pretty out of it bc of all the sedation and side effects but I wasn’t unconscious or like totally incapable of communication and decision making.
___
After a very lengthy series of hospitalizations I had such mixed feelings about psychiatry. On one hand I was grateful and knew I to acknowledge it had saved my life and restored me to sanity, but the experience of being hospitalized and put on a series of meds without my knowledge or consent was so hellish and my care was managed so poorly that the cost I paid for having my manic episode resolved was literally years.
Years of depression, self-hatred, rage and terror at the thought of being hospitalized, distrust of medical professionals, and learning how to deal with the flashbacks from experiences in the hospital that are so deeply repressed I literally can’t access them voluntarily.
Yet at the same time, coming from a family of serious mental illness and denial and abuse and the agony of the disease when it’s untreated, being hospitalized changed the trajectory of my life for the better. It took 5-10 years for me to see that, but I realized that with medication and healthy behavior, my life could be relatively normal - but without it I would likely follow my mom down the path to self-destruction and ruining my career and relationship.
This is why I get very nervous when psychiatrists publicly question the value of broad classes of medication, but it’s more of a personal thing that I have to get over. People like my family absolutely love and weaponize studies and prominent figures that oppose medication as confirmation that they were just keeping me safe & being prudent and imply that my life is out of control and needs an intervention because I’m taking multiple psych meds.
But getting the truth out to patients is more important than giving more ammunition to the crazies. In the long run I hope greater honesty will take some of the steam out of the fiercer opposition and suspicion psychiatry faces in the US.
I probably will enjoy your talk with Dr. Kramer but I just wanted to say that I did appreciate this post and guest
I would like to thank Dr. Plöderl and Dr. Aftab for an enlightening discussion of this complex and controversial topic. As a (now retired) specialist in adult mood disorders, I acknowledge the uncertainties inherent in studies of SSRIs/SNRIs and suicidality; however, I believe the preponderance of the evidence does not support the claim that these agents increase suicidality (a vague term, to be sure) or completed suicide in adult patients 25 years and older. And, there are at least some data pointing to a reduction of suicidality (suicide attempts) in antidepressant-treated patients age 65 and older, as the discussion with Dr. Plöderl notes (Stone et al, 2009)
A recent, carefully-done, observational study by Lagerberg et al generally supports the above conclusions. The authors write:
"We found an increased risk of suicidal behaviour among individuals aged below 25 years who were treated with an SSRI after a depression diagnosis as compared to those who were not. We found no evidence of an effect among older age categories. The results are similar to those from RCTs, lending validity to this observational study. However, the issue of unmeasured confounding remains, and studies from data sources with more detailed information on confounders—notably depression severity—are called for. Our results confirm that individuals with a history of suicidal behaviour are a high-risk group, but we found no evidence that SSRI initiation conferred elevated risks of suicidal behaviour in this subgroup." [1]
Like all observational studies, this one has both strengths and limitations, as the authors acknowledge.
Finally, I would like to comment on the statement that,
"Put differently, the claim that “antidepressants are live-saving” is not supported by the evidence, at least not for the average patient."
Clinicians, of course, do not treat "the average" patient or an aggregate of patients. Each patient we treat is unique. Having treated many hundreds of severely depressed patients over the course of more than 25 years--many of whom I followed closely for 2, 3 or more years--I have certainly seen many whose suicidal ideation or impulses diminished greatly in the course of antidepressant treatment.
I am well aware that such observations are often dismissed as "anecdotal" or the result of an unacknowledged "placebo effect." Yet when this observation is made dozens and dozens of times, in patients whose depression did not respond to two or three adequate antidepressant trials, but did respond to the fourth agent, the clinician is naturally reluctant to accept the, "it was just a placebo effect" explanation. In such cases, I believe it is entirely appropriate to say that the totality of treatment--antidepressant plus the therapeutic alliance, probably acting synergistically--was indeed "life-saving."
Thanks again to you both for this useful discussion!
Respectfully,
Ronald W. Pies, MD
Professor Emeritus of Psychiatry
1. Lagerberg, T., Matthews, A.A., Zhu, N. et al. Effect of selective serotonin reuptake inhibitor treatment following diagnosis of depression on suicidal behaviour risk: a target trial emulation. Neuropsychopharmacol. 48, 1760–1768 (2023). https://doi.org/10.1038/s41386-023-01676-3
Something that I really would love to see an article or interview with you about Dr. Aftab: the scientific validity (or rather invalidity) of psychiatric nosology.
As others touched on - MDD is heterogenous, and as others have argued elsewhere - we have subsumed many bipolar spectra and cyclothymic temperaments under MDD.
It certainly dovetails with the observation of antidepressant induced suicidality, specifically that many individuals who are becoming activated, anxious, suicidal with antidepressants. As many will argue, the index episode of bipolar is depression, and the onset of illness is a median of 19, a full decade before MDD typically. It would be no wonder then that the signal for suicidality would strengthen with younger age.
Our ability to see not only adverse effects, but to understand genuine efficacy (antidepressants, as you have touched on, significantly help a certain proportion of patients, and it is often obscured by the “average” effect sizes we see) has been hampered by the rigidity of the DSM in many ways, and specifically the FDA’s reliance on broad, often invalid DSM diagnoses in order to seek approval.
Agreed.
For me the overlap between the period of suicidality and the period where bipolar symptoms begin to emerge is far too coincidental. Induced manic episodes frequently lead to suicide attempts and usually aren’t noticed by clinicians unless you see the person a few days later or they have the insight to recognize that something has gone *very* wrong and reach out to you.
But many psychs in that situation are just like “the first week can be hard, just hang in there, here’s some trazodone/benzos to help you sleep, in a week or two you’ll be feeling better”. More likely in a week or two you’ll be dead if you keep taking the pills. Fortunately most ppl realize that they’re making things way worse, unfortunately not all.
Always important to remember time from first bipolar episode to diagnosis average is like 8 years. The chance of a person who is 20 years old having an accurate diagnosis for their emerging bipolar is basically zero. Many bipolar 1 and most bipolar 2 find out what mania is when they take ADs for the first time.
No hating on ADs here. Once I got on mood stabilizers I had one added back on and it helps immensely. But bipolar is underdiagnosed until you get to like late 20s/early 30s and should always be considered whenever there is any kind of activating/agitating crisis induced by an AD.
Excellent discussion. Thank you.