No need to assign ill intent to the field of psychiatry in regard to brain scan studies. The sorts of research errors they likely are making are actually common throughout various fields. Lack of preregistered studies / publication bias, lack of control of confounding variables, mistaking correlation for causation, various forms of statistical fudging—mostly these things can be explained by sloppiness, the academic publish-or-perish mentality, and the genuine difficulty of performing quality research that leads to robust causal conclusions.
I’m not saying there aren’t bad actors. And yes, the incentive structure for researchers is largely messed up, rewarding quantity over quality. But I think mostly it’s genuine mistakes and lack of understanding that lead to these sorts of problems, and not just in psychiatry.
I will once again recommend PubPeer and RetractionWatch as places to get insight into the kinds of problems afflicting scientific research today. Deliberate fraud exists, but so do many other problems.
This is the first I’d heard of a new edition of Anatomy of an Epidemic. Has this already been released, or is that still upcoming?
Great article. Reminds me of the book “Crazy Like Us” which was mind-opening.
“the increase in serotonin tells the body that it’s ‘fully satisfied.'”
This reasoning relies on the problematic assumption that antidepressants actually increase serotonin levels in the brain.
Kudos for bringing together all of these related claims in one place. What I’d really like to see in addition is an explanation of what justifies these claims being considered “graveyard” material. Quotes from corresponding refutations would be especially helpful.
The correlation with risk of joining a gang, and using a weapon after having joined a gang, made me laugh. I guess it’s possible but that feels like such a stretch to me!
I also generally admire Dr. Breggin’s work but feel he goes to far with this latest. There’s certainly a need for guilt, shame, and anxiety to put into context, and it’s important to learn that they are emotional responses to things but aren’t inherently representative of truth. However, I don’t think a complete disregard for these feelings—treating them as evolutionary baggage that needs to be left behind—is the right approach. None of these “negative legacy emotions” is wholly negative. In fact, all can be markedly positive when part of an integrated emotional life that includes moderators such as reason and love (which Breggin mentions).
To sum it up, the old folks were annoying nursing home staff, so the staff fried their brains and wrote up this nice paper to tell the world how helpful they found it.
As police violence is to the general public, nursing home practices are to the elderly: the interests of the authority figures (police, nursing home staff) are seen as paramount and pursued at the expense of the interests of the individuals they are paid to protect.
In case you’re not familiar, PubPeer is a site for critical discussion of scientific publications. Commenters on PubPeer have already uncovered numerous instances of fraudulent research, resulting in retractions.
The PubPeer page for the Archarya, et al. article has already gotten a few views and will likely receive more in the future, so if you place a comment there it will become visible to researchers interested in critical discussion of the article in question. Be sure to read the howto first: https://pubpeer.com/howto
As many Mad in America posts involve criticism of published research, I believe it would be good practice as a community for authors to extract the purely factual elements of their criticisms and post them as PubPeer comments. This would be one way to expand the reach of the work being done here beyond the existing readership of MIA.
@Bossco, none of what you have described actually proves the Zoloft is effective due to its chemical properties. It’s just as likely that the Zoloft is effective because you expect that it will be effective (placebo response.) Obviously I can’t prove that, either, but I think it’s a possibility you should consider. I mean, how great would that be to discover you didn’t even need to be taking the Zoloft?
For about a decade I was just as convinced as you that “antidepressants” saved my life. I was equally convinced that I needed to take a benzodiazepine to cure me of anxiety. But eventually I learned that wasn’t true, and now it seems very much like it was my belief that the drugs would work that made them seem to work. Getting off the benzo actually reduced my anxiety level dramatically, and now that I’m halfway off my “antidepressants” and still doing great I am persuaded that I don’t need those either.
Now, exercise and talking openly with people about what’s going on in my life have become my real antidepressants. Maybe that’s just because I expect them to have that effect, but hey, at least they have positive rather than negative sideeffects 🙂
Just throwing this out there: I have no doubt that benzos and other psych drugs can have long-term consequences and withdrawal can take years. But the farther out you get from the time of the last dose, the less likely it seems that a panic attack or return of symptoms would be a result of withdrawal syndrome, and the more likely it would be caused by something else. So, maybe your panic attacks, etc., seven years after quitting benzos aren’t related to the benzos. Maybe you just straight up had a panic attack. There’s nothing wrong with that. It doesn’t mean you have a disease or have to go back on drugs. But maybe you can look to other ways of dealing with it than just blaming it on the long-past psych drugs and waiting for it to pass.
This is great news. I frequent Retraction Watch in particular and feel like its efforts bring valuable accountability to scientific researchers and publishers alike. I’m glad to see they’ll be able to make the site more comprehensive.
PubPeer is another research accountability effort worth being aware of. I don’t think I’ve seen much activity there with respect to psychology/psychiatry, though.
I’m glad this debate happened. I respect Allen Frances for reaching out to this community and trying to establish common ground. My impression of him is that his APA connections do blind him to the role of APA in the current state of affairs. But it’s good to hear outside voices in dialogue with this community and I think we need more of that 🙂
What’s with the 24.4% increase in prevalence of birth defects? That surprised me.
I was on Klonopin for some seven or eight years, starting at age 14. Definitely longer than 90 days. When I finally got off of it, suddenly my panic attacks virtually vanished—the “antianxiety” drug was largely causing (or at least exacerbating) my anxiety! I feel very fortunate that I was able to wean off a benzo in a matter of months, whereas so many out there suffer for years on end with the withdrawal. Hang in there, humanbeing.
Makes me want to cry hearing your story :’-( Please don’t commit suicide—darkness turns to light. Hang in there, seek friends, start hobbies, ask for help, pray, meditate, read, write, feel the pain, start over, then start again. I hope you find your way forward soon.
I actually think tusu has something of a point when he accuses Mad In America of lacking dialogue. Obviously there is dialogue here, but largely it’s discussions between people who are generally like-minded but disagree in the particulars. There are few commenters, for example, who speak in favor of the medical model, or who speak of their recovery from depression using antidepressant medications.
While I generally doubt the efficacy of mainstream psychiatry and medical model based treatments, I would still like to see more contrary voices around these parts. Perhaps MIA is seen too much as espousing just one side of the debate?
Also, I think your use of reference #3 is a bit problematic, since the study in question demonstrated that the seeming differences between treated and non-treated groups were potentially explained by various confounding variables. However, at least from my quick look over it the study did seem compatible with the notion that antidepressants have no effect.
Can you point me to somewhere online I can read references #2 and #4?
I also find it strange that the ineffectiveness of medications relative to placebo is seen as a problem that must be solved by means of ever-more-clever experimental design, rather than as a finding that should be understood in its own right.
Please remember that antidepressants _do_ work—the scandal is that they work because of the placebo effect. I expect that the delay in antidepressant effect after starting treatment is real and is due to the mechanics of this placebo effect. It takes time for the drug’s side effects (dry mouth, etc.) to appear, and these side effects are a large part of what convinces people that the drug is effective, triggering a placebo response.
Let’s see how many people we can offend in a single post… Republicans, Christians…. Not that you should never offend anybody, but you really gained nothing by these comparisons. I’d suggest you stick to matters of psychiatry and keep your views on American politics and religion separate, as they’ll only serve as red herrings to distract people from the real issues at hand.
I’m not a Republican, I agree many Tea Partiers are crazy, but as someone who believes in freedom and constitutional government I sympathize with a number of their ideas. All political parties make outrageous claims that aren’t backed up by facts. Why make this gratuitous comparison between biological psychiatry and the political movement you most disagree with? Totally counterproductive.
Thanks a lot for your further thoughts, Robert. I feel persuaded after seeing the Harrow study in the context of the others. The Harrow study is observational and thus difficult/impossible to draw causal conclusions from, but randomized experiments appropriate for drawing such causal conclusions have found the same thing, thus a larger picture is being drawn wherein a fat lady can be seen singing 🙂
Looking at the paper, the conclusions section from the executive summary seems to warn against making the sort of strong claim that Whitaker does in this post:
“The 20-year data indicate that, longitudinally, after the first few years, antipsychotic medications do not eliminate or reduce the frequency of psychosis in schizophrenia, or reduce the severity of post-acute psychosis, although it is difficult to reach unambiguous conclusions about the efficacy of treatment in purely naturalistic or observational research. Longitudinally, on the basis of their psychotic activity and the disruption of functioning, the condition of the majority of SZ prescribed antipsychotics for multiple years would raise questions as to how many of them are truly in remission.”
The paper seems to provide evidence that is consistent with the claim that antipsychotics impede recovery from psychosis, but it also does not demonstrate that that claim is in fact the case.
I am also concerned about this weakness in the argument. It would seem that a person’s probability of initially being placed on medication is higher if their symptoms are more intense. That would mean the people not on medication are those who had the least severe symptoms to begin with and thus were more likely to recover anyway.
Great article. One small critique: the author makes a common error in his interpretation of the meaning of P values in statistical hypothesis testing. See http://www.nature.com/news/scientific-method-statistical-errors-1.14700 for more on the problems with this approach to data analysis.
No need to assign ill intent to the field of psychiatry in regard to brain scan studies. The sorts of research errors they likely are making are actually common throughout various fields. Lack of preregistered studies / publication bias, lack of control of confounding variables, mistaking correlation for causation, various forms of statistical fudging—mostly these things can be explained by sloppiness, the academic publish-or-perish mentality, and the genuine difficulty of performing quality research that leads to robust causal conclusions.
I’m not saying there aren’t bad actors. And yes, the incentive structure for researchers is largely messed up, rewarding quantity over quality. But I think mostly it’s genuine mistakes and lack of understanding that lead to these sorts of problems, and not just in psychiatry.
I will once again recommend PubPeer and RetractionWatch as places to get insight into the kinds of problems afflicting scientific research today. Deliberate fraud exists, but so do many other problems.
This is the first I’d heard of a new edition of Anatomy of an Epidemic. Has this already been released, or is that still upcoming?
Great article. Reminds me of the book “Crazy Like Us” which was mind-opening.
“the increase in serotonin tells the body that it’s ‘fully satisfied.'”
This reasoning relies on the problematic assumption that antidepressants actually increase serotonin levels in the brain.
Kudos for bringing together all of these related claims in one place. What I’d really like to see in addition is an explanation of what justifies these claims being considered “graveyard” material. Quotes from corresponding refutations would be especially helpful.
The correlation with risk of joining a gang, and using a weapon after having joined a gang, made me laugh. I guess it’s possible but that feels like such a stretch to me!
I also generally admire Dr. Breggin’s work but feel he goes to far with this latest. There’s certainly a need for guilt, shame, and anxiety to put into context, and it’s important to learn that they are emotional responses to things but aren’t inherently representative of truth. However, I don’t think a complete disregard for these feelings—treating them as evolutionary baggage that needs to be left behind—is the right approach. None of these “negative legacy emotions” is wholly negative. In fact, all can be markedly positive when part of an integrated emotional life that includes moderators such as reason and love (which Breggin mentions).
To sum it up, the old folks were annoying nursing home staff, so the staff fried their brains and wrote up this nice paper to tell the world how helpful they found it.
As police violence is to the general public, nursing home practices are to the elderly: the interests of the authority figures (police, nursing home staff) are seen as paramount and pursued at the expense of the interests of the individuals they are paid to protect.
Philip,
Your post contains some valuable criticisms of the paper you reference. May I suggest summarizing these as a comment on the PubPeer page corresponding to that article? https://pubpeer.com/publications/C83CE05FBB33BF848E8EA862379443
In case you’re not familiar, PubPeer is a site for critical discussion of scientific publications. Commenters on PubPeer have already uncovered numerous instances of fraudulent research, resulting in retractions.
The PubPeer page for the Archarya, et al. article has already gotten a few views and will likely receive more in the future, so if you place a comment there it will become visible to researchers interested in critical discussion of the article in question. Be sure to read the howto first: https://pubpeer.com/howto
As many Mad in America posts involve criticism of published research, I believe it would be good practice as a community for authors to extract the purely factual elements of their criticisms and post them as PubPeer comments. This would be one way to expand the reach of the work being done here beyond the existing readership of MIA.
@Bossco, none of what you have described actually proves the Zoloft is effective due to its chemical properties. It’s just as likely that the Zoloft is effective because you expect that it will be effective (placebo response.) Obviously I can’t prove that, either, but I think it’s a possibility you should consider. I mean, how great would that be to discover you didn’t even need to be taking the Zoloft?
For about a decade I was just as convinced as you that “antidepressants” saved my life. I was equally convinced that I needed to take a benzodiazepine to cure me of anxiety. But eventually I learned that wasn’t true, and now it seems very much like it was my belief that the drugs would work that made them seem to work. Getting off the benzo actually reduced my anxiety level dramatically, and now that I’m halfway off my “antidepressants” and still doing great I am persuaded that I don’t need those either.
Now, exercise and talking openly with people about what’s going on in my life have become my real antidepressants. Maybe that’s just because I expect them to have that effect, but hey, at least they have positive rather than negative sideeffects 🙂
Just throwing this out there: I have no doubt that benzos and other psych drugs can have long-term consequences and withdrawal can take years. But the farther out you get from the time of the last dose, the less likely it seems that a panic attack or return of symptoms would be a result of withdrawal syndrome, and the more likely it would be caused by something else. So, maybe your panic attacks, etc., seven years after quitting benzos aren’t related to the benzos. Maybe you just straight up had a panic attack. There’s nothing wrong with that. It doesn’t mean you have a disease or have to go back on drugs. But maybe you can look to other ways of dealing with it than just blaming it on the long-past psych drugs and waiting for it to pass.
This is great news. I frequent Retraction Watch in particular and feel like its efforts bring valuable accountability to scientific researchers and publishers alike. I’m glad to see they’ll be able to make the site more comprehensive.
PubPeer is another research accountability effort worth being aware of. I don’t think I’ve seen much activity there with respect to psychology/psychiatry, though.
I’m glad this debate happened. I respect Allen Frances for reaching out to this community and trying to establish common ground. My impression of him is that his APA connections do blind him to the role of APA in the current state of affairs. But it’s good to hear outside voices in dialogue with this community and I think we need more of that 🙂
What’s with the 24.4% increase in prevalence of birth defects? That surprised me.
I was on Klonopin for some seven or eight years, starting at age 14. Definitely longer than 90 days. When I finally got off of it, suddenly my panic attacks virtually vanished—the “antianxiety” drug was largely causing (or at least exacerbating) my anxiety! I feel very fortunate that I was able to wean off a benzo in a matter of months, whereas so many out there suffer for years on end with the withdrawal. Hang in there, humanbeing.
Makes me want to cry hearing your story :’-( Please don’t commit suicide—darkness turns to light. Hang in there, seek friends, start hobbies, ask for help, pray, meditate, read, write, feel the pain, start over, then start again. I hope you find your way forward soon.
I actually think tusu has something of a point when he accuses Mad In America of lacking dialogue. Obviously there is dialogue here, but largely it’s discussions between people who are generally like-minded but disagree in the particulars. There are few commenters, for example, who speak in favor of the medical model, or who speak of their recovery from depression using antidepressant medications.
While I generally doubt the efficacy of mainstream psychiatry and medical model based treatments, I would still like to see more contrary voices around these parts. Perhaps MIA is seen too much as espousing just one side of the debate?
Also, I think your use of reference #3 is a bit problematic, since the study in question demonstrated that the seeming differences between treated and non-treated groups were potentially explained by various confounding variables. However, at least from my quick look over it the study did seem compatible with the notion that antidepressants have no effect.
Can you point me to somewhere online I can read references #2 and #4?
I also find it strange that the ineffectiveness of medications relative to placebo is seen as a problem that must be solved by means of ever-more-clever experimental design, rather than as a finding that should be understood in its own right.
Please remember that antidepressants _do_ work—the scandal is that they work because of the placebo effect. I expect that the delay in antidepressant effect after starting treatment is real and is due to the mechanics of this placebo effect. It takes time for the drug’s side effects (dry mouth, etc.) to appear, and these side effects are a large part of what convinces people that the drug is effective, triggering a placebo response.
Let’s see how many people we can offend in a single post… Republicans, Christians…. Not that you should never offend anybody, but you really gained nothing by these comparisons. I’d suggest you stick to matters of psychiatry and keep your views on American politics and religion separate, as they’ll only serve as red herrings to distract people from the real issues at hand.
I’m not a Republican, I agree many Tea Partiers are crazy, but as someone who believes in freedom and constitutional government I sympathize with a number of their ideas. All political parties make outrageous claims that aren’t backed up by facts. Why make this gratuitous comparison between biological psychiatry and the political movement you most disagree with? Totally counterproductive.
Thanks a lot for your further thoughts, Robert. I feel persuaded after seeing the Harrow study in the context of the others. The Harrow study is observational and thus difficult/impossible to draw causal conclusions from, but randomized experiments appropriate for drawing such causal conclusions have found the same thing, thus a larger picture is being drawn wherein a fat lady can be seen singing 🙂
Looking at the paper, the conclusions section from the executive summary seems to warn against making the sort of strong claim that Whitaker does in this post:
“The 20-year data indicate that, longitudinally, after the first few years, antipsychotic medications do not eliminate or reduce the frequency of psychosis in schizophrenia, or reduce the severity of post-acute psychosis, although it is difficult to reach unambiguous conclusions about the efficacy of treatment in purely naturalistic or observational research. Longitudinally, on the basis of their psychotic activity and the disruption of functioning, the condition of the majority of SZ prescribed antipsychotics for multiple years would raise questions as to how many of them are truly in remission.”
The paper seems to provide evidence that is consistent with the claim that antipsychotics impede recovery from psychosis, but it also does not demonstrate that that claim is in fact the case.
I am also concerned about this weakness in the argument. It would seem that a person’s probability of initially being placed on medication is higher if their symptoms are more intense. That would mean the people not on medication are those who had the least severe symptoms to begin with and thus were more likely to recover anyway.
Great article. One small critique: the author makes a common error in his interpretation of the meaning of P values in statistical hypothesis testing. See http://www.nature.com/news/scientific-method-statistical-errors-1.14700 for more on the problems with this approach to data analysis.
Bayesian statistics are one approach to hypothesis testing that gets away from the binary significant/not-significant distinction. See http://www.indiana.edu/~kruschke/BEST/BEST.pdf
I expect that a great deal of psychiatric/psychological research rests on similarly flawed statistical analysis.