6 ms·

But the biological theory of mental illnesses is very far from established. We have no reliable way of identifying them and no understanding of their mechanism.

by gruseom 13y ago

But the biological theory of mental illnesses is very far from established. We have no reliable way of identifying them and no understanding of their mechanism. Existing studies of the medications we use to treat them have been subjected to pretty devastating critique; e.g. the most comprehensive review of the SSRI studies (by Irving Kirsch et. al.) found no clinical significance beyond placebo for the vast majority of patients. The belief that mental illness is biological ("chemical imbalance in the brain" and so on) is very widespread and very emphatic. But the gap between that and what we actually know is drastic. What explains that gap? Most likely, it's that (a) people very much want to believe it, and (b) the belief has been heavily marketed.

scott_s 13y ago
Steven Novella addresses the biological theory of mental illness quite well in this 10 minute segment from the podcast The Skeptics Guide to the Universe: http://www.youtube.com/watch?v=-MLdvKzLl6k http://www.youtube.com/watch?v=-MLdvKzLl6k
logjam 13y ago
> the most comprehensive review of the SSRI studies (by Irving Kirsch et. al.) found no clinical significance beyond placebo for the vast majority of patients. No. Biological theories of mental health and mental illness are very much established. And yours is a complete mischaracterization of the work of Kirsch et al. Their meta-analysis found that in ALL cases of depression, there is statistically significant improvement in quantitative measures of depression treated with SSRI. In moderate to severe cases of depression, the improvement was clinically significant (ie a clinician and patient clearly note the improvement). The meta-analysis done on clinical trials were overwhelmingly trials of 6 weeks of SSRI or less. Suicidal patients are almost all excluded from such trials. Treatment groups were treated with SSRIs via research protocols, and not usual clinical practice: there were no attempts made to match patients to an SSRI that would be most likely to help, to wait on a response, to use SSRIs that were helpful based on a family history; dose adjustments were typically prescribed by study protocol and not by patient response, patients were not switched to another SSRI when there was no response or if intolerable side effects developed, often no attempt to treat side effects with other agents were done, etc. In other words, none of the things any physician routinely does to optimize treatment were done in those trials, and yet there were STILL improvements in measures of depression across the board. No reputable psychiatrist or physician will fail to recommend an SSRI in cases of moderate to severe depression. Therapy can also be helpful, but the evidence is overwhelming that SSRIs help in clinically significant ways in moderate to severe depression. There is no credible doubt about it. Despite your assertions about what "people very much want to believe" - in my experience, people very much want to believe that the brain is somehow different - that something beyond biology is at work. Believing that somehow depressed or otherwise mentally ill people can just "snap out of it" or "pull themselves up by their bootstraps" is erroneous, unhelpful, but fits nicely into really completely shitty and cruel narratives about human beings. But please. Perhaps you'd like to advance a theory on how an organ like the brain differs from every other organ and does its work in ways not explainable by biology, chemistry, physics, etc.
- gruseom 13y ago
  And yours is a complete mischaracterization of the work of Kirsch et al. It isn't my characterization so much as a paraphrase of various sources I've read/heard on this, including Kirsch himself. My understanding is that Kirsch and Sapirstein's findings are two: (a) for all but severe depression, the improvement of SSRIs over placebo is measurable but too small to be clinically significant; (b) even that small measurable difference is questionable, because the studies were not really double-blind. (Patients could figure out which group they were in because of the SSRIs' side effects.) Is either of those descriptions of their findings wrong? I haven't kept a list of sources, but one I remember is [1]. Is Kirsch mischaracterizing his own work? I noticed that you didn't use the word 'placebo' in what you wrote. My understanding is that Kirsch et. al.'s findings were specifically about lack of clinical significance beyond placebo. Could that explain the discrepancy between your description and mine? Perhaps you'd like to advance a theory I would not. Do you think that lack of a better theory has anything to do with whether this one is substantiated or not? [1] http://www.huffingtonpost.com/irving-kirsch-phd/antidepressants-the-emper_b_442205.html http://www.huffingtonpost.com/irving-kirsch-phd/antidepressa...
  - Retric 13y ago
    Shouldn't the default position be "we don't know"? No that's a common misunderstanding. Science is the search for the best model/theory as such it's hard to rank something as better or worse than we don't know. Thus the default theory is "There is no relationship between X and Y". Even if not clinically useful there does appear to be a connection and as such you can replace the default theory. After even more research the current theory is something like. "They appear to beat placebo's which have no side effects." Which again is far more useful than "We don't know." As it suggests comparing them to placebos that have side effects. Again not because they are going to help you treat patents directly, but because it tells you more about the disease and possible what research could be useful.
    - gruseom 13y ago
      Science is the search for the best theory as such it's hard to rank something as better or worse than we don't know. Thus the default theory is "There is no relationship between X and Y". For the purposes of this discussion, that's a distinction without a difference. (BTW I edited my comment before your reply showed up, and had deleted the bit you quoted.) More significantly, your suggestion of comparing SSRIs to placebos with side effects strikes me as an excellent idea; it is what Kirsch's study would naturally seem to suggest. I wonder if such studies will ever be undertaken. Edit: I did find at least one such study: http://psychrights.org/research/Digest/CriticalThinkRxCites/Moncrief2004.pdf http://psychrights.org/research/Digest/CriticalThinkRxCites/... Authors' conclusions: The more conservative estimates from the present analysis found that differences between antidepressants and active placebos were small. This suggests that unblinding effects may inﬂate the efﬁcacy of antidepressants in trials using inert placebos. Further research into unblinding is warranted.
andrewcooke 13y ago
i realise it's anecdote, but i grew up with a parent that was depressive. was on lithium, etc. life was pretty damn miserable until the first SSRIs became available - they were the first drugs that had any noticeable effect and they changed our lives.
- gruseom 13y ago
  Anecdote it may be, but I'm glad to hear that. I know both sides of that coin myself.
Retric 13y ago
Mental illnesses like cancer are defined by there effects not there underlying cause. As such they tend to treat a wide range of physical causes under the same general umbrella. And drugs or therapies which work wonders for some people often have little benifit for others. However, there are well studdied physical mechanisms for things like specific types of addiction which do present a physical mechanism. A great example of this is Heroin addiction and alcohol addiction are vary different biochemically. There is even a wide range of Alcholhol addictions, but you will see them treated side by side. The simple truth is we can't really understand why someone has issues but we do have treatments that help people lead productive lives and that's enough to be useful. Just as your dentist does not need to know how pain killers work to use them effectively.
DanBC 13y ago
Your comment is far more nuanced. You're reading the science. While I gently disagree with you I understand what you're saying, and I could fairly easily be persuaded that you're right. The comment I replied to had some worrying flags. It made blanket statements about meds; it suggested that depression has a "root cause". I took that to be a reference to classical psychotherapy -- some event happened a long time in the patient's past and that event must be uncovered by the therapist for the patient to address it and recover. Re-reading the post I see that I might be wrong! Perhaps the post is just talking about regular evidence based cognitive model. > The belief that mental illness is biological Well, for something like bipolar disorder this feels like it's true. I don't know much about BPD, and I don't know many people in real life who have it. But it seems that people need the meds. When talking about depression I recognise that there are various types. I'm not a doctor and have no special knowledge. When people ask I suggest they investigate talking therapies (and probably CBT) first. But I know that people might get benefit from meds, especially if they're on the more severe end. I know that meds can have unpleasant side effects.
- gruseom 13y ago
  It's great when someone replies as thoughtfully as this, and I love your phrase "gently disagree". Would that more disagreements were of this kind. > Well, for something like bipolar disorder this feels like it's true. I agree. The open question is on what level these things are best to be understood. In some sense everything about us is biology, just as in some sense everything is physics. But we don't think of, say, heart disease as best addressed by physics. Similarly, that there is a biological stratum to our actions and feelings does not automatically mean that chronic emotional suffering is best understood as biology. One might as well conclude from the tongue and larynx that language is biology too. And indeed it is, sort of—yet its meaning lies elsewhere. Or you might as well conclude from the importance of neurons to learning that education is brain biology. Why bother with teachers or schools? We should just 'learn' the neurons directly. Such examples are obviously silly, at least given our current knowledge. To know that chronic emotional suffering is a biological disorder requires more than the involvement of biological phenomena in it; it requires an experimentally verified model. My understanding is that we don't have anything close to that. The one that entered the public imagination, the serotonin-deficiency theory, is widely dismissed by experts [1,2,3,4]. The only argument seems to be whether they knew it was false from the beginning or discovered that it was false decades ago. Yet we insist that modern science has discovered that mental illness is biological just the way that cancer is and so on. What do you call a conclusion like that which goes far beyond what we actually know? It's worth realizing that psychiatry has always made this claim. The grounds for it shift every 20 years or so, and the previous grounds are always dismissed as ludicrous if not harmful (think lobotomies etc.). Nevertheless we're perpetually certain that we're beyond all that now. > But I know that people might get benefit from meds I don't think anyone questions that; the question is whether the benefit is that of a placebo or not. My understanding is that SSRIs are no more effective a treatment for depression than tricyclics were a generation earlier; their advantage is rather that they have fewer side effects [5]. So whatever explanation there is for their efficacy must plausibly explain how both of those (presumably very different?) biochemical mechanisms could do it. Given that even inert placebos produce most of the same effect (Kirsch's finding), the placebo explanation is pretty clearly a major candidate. If there's another, I'd like to know about it. (I realize you were talking primarily about bipolar disorder, but I haven't read about that, so I've continued to talk about depression instead. That may lessen the relevance of the above.) [1] http://www.npr.org/blogs/health/2012/01/23/145525853/when-it-comes-to-depression-serotonin-isnt-the-whole-story http://www.npr.org/blogs/health/2012/01/23/145525853/when-it... [2] http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020392.t001/largerimage http://www.plosmedicine.org/article/info:doi/10.1371/journal..., via http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020392 http://www.plosmedicine.org/article/info:doi/10.1371/journal... [3] http://www.psychiatrictimes.com/blogs/couch-crisis/psychiatry%E2%80%99s-new-brain-mind-and-legend-%E2%80%9Cchemical-imbalance%E2%80%9D http://www.psychiatrictimes.com/blogs/couch-crisis/psychiatr..., paywall bypassable by clicking link at http://goo.gl/3LxUW http://goo.gl/3LxUW [4] http://www.madinamerica.com/2012/01/revising-the-history-of-the-serotonin-theory-of-depression-2/ http://www.madinamerica.com/2012/01/revising-the-history-of-... [5] http://www.ncbi.nlm.nih.gov/pubmed/17636689 http://www.ncbi.nlm.nih.gov/pubmed/17636689. This looks like it was withdrawn because it was to be superseded by a larger study, but I couldn't find that one.

by gruseom 13y ago

scott_s 13y ago

Steven Novella addresses the biological theory of mental illness quite well in this 10 minute segment from the podcast The Skeptics Guide to the Universe: http://www.youtube.com/watch?v=-MLdvKzLl6k http://www.youtube.com/watch?v=-MLdvKzLl6k

logjam 13y ago

> the most comprehensive review of the SSRI studies (by Irving Kirsch et. al.) found no clinical significance beyond placebo for the vast majority of patients. No. Biological theories of mental health and mental illness are very much established. And yours is a complete mischaracterization of the work of Kirsch et al. Their meta-analysis found that in ALL cases of depression, there is statistically significant improvement in quantitative measures of depression treated with SSRI. In moderate to severe cases of depression, the improvement was clinically significant (ie a clinician and patient clearly note the improvement). The meta-analysis done on clinical trials were overwhelmingly trials of 6 weeks of SSRI or less. Suicidal patients are almost all excluded from such trials. Treatment groups were treated with SSRIs via research protocols, and not usual clinical practice: there were no attempts made to match patients to an SSRI that would be most likely to help, to wait on a response, to use SSRIs that were helpful based on a family history; dose adjustments were typically prescribed by study protocol and not by patient response, patients were not switched to another SSRI when there was no response or if intolerable side effects developed, often no attempt to treat side effects with other agents were done, etc. In other words, none of the things any physician routinely does to optimize treatment were done in those trials, and yet there were STILL improvements in measures of depression across the board. No reputable psychiatrist or physician will fail to recommend an SSRI in cases of moderate to severe depression. Therapy can also be helpful, but the evidence is overwhelming that SSRIs help in clinically significant ways in moderate to severe depression. There is no credible doubt about it. Despite your assertions about what "people very much want to believe" - in my experience, people very much want to believe that the brain is somehow different - that something beyond biology is at work. Believing that somehow depressed or otherwise mentally ill people can just "snap out of it" or "pull themselves up by their bootstraps" is erroneous, unhelpful, but fits nicely into really completely shitty and cruel narratives about human beings. But please. Perhaps you'd like to advance a theory on how an organ like the brain differs from every other organ and does its work in ways not explainable by biology, chemistry, physics, etc.

gruseom 13y ago
And yours is a complete mischaracterization of the work of Kirsch et al. It isn't my characterization so much as a paraphrase of various sources I've read/heard on this, including Kirsch himself. My understanding is that Kirsch and Sapirstein's findings are two: (a) for all but severe depression, the improvement of SSRIs over placebo is measurable but too small to be clinically significant; (b) even that small measurable difference is questionable, because the studies were not really double-blind. (Patients could figure out which group they were in because of the SSRIs' side effects.) Is either of those descriptions of their findings wrong? I haven't kept a list of sources, but one I remember is [1]. Is Kirsch mischaracterizing his own work? I noticed that you didn't use the word 'placebo' in what you wrote. My understanding is that Kirsch et. al.'s findings were specifically about lack of clinical significance beyond placebo. Could that explain the discrepancy between your description and mine? Perhaps you'd like to advance a theory I would not. Do you think that lack of a better theory has anything to do with whether this one is substantiated or not? [1] http://www.huffingtonpost.com/irving-kirsch-phd/antidepressants-the-emper_b_442205.html http://www.huffingtonpost.com/irving-kirsch-phd/antidepressa...
- Retric 13y ago
  Shouldn't the default position be "we don't know"? No that's a common misunderstanding. Science is the search for the best model/theory as such it's hard to rank something as better or worse than we don't know. Thus the default theory is "There is no relationship between X and Y". Even if not clinically useful there does appear to be a connection and as such you can replace the default theory. After even more research the current theory is something like. "They appear to beat placebo's which have no side effects." Which again is far more useful than "We don't know." As it suggests comparing them to placebos that have side effects. Again not because they are going to help you treat patents directly, but because it tells you more about the disease and possible what research could be useful.
  - gruseom 13y ago
    Science is the search for the best theory as such it's hard to rank something as better or worse than we don't know. Thus the default theory is "There is no relationship between X and Y". For the purposes of this discussion, that's a distinction without a difference. (BTW I edited my comment before your reply showed up, and had deleted the bit you quoted.) More significantly, your suggestion of comparing SSRIs to placebos with side effects strikes me as an excellent idea; it is what Kirsch's study would naturally seem to suggest. I wonder if such studies will ever be undertaken. Edit: I did find at least one such study: http://psychrights.org/research/Digest/CriticalThinkRxCites/Moncrief2004.pdf http://psychrights.org/research/Digest/CriticalThinkRxCites/... Authors' conclusions: The more conservative estimates from the present analysis found that differences between antidepressants and active placebos were small. This suggests that unblinding effects may inﬂate the efﬁcacy of antidepressants in trials using inert placebos. Further research into unblinding is warranted.

andrewcooke 13y ago

i realise it's anecdote, but i grew up with a parent that was depressive. was on lithium, etc. life was pretty damn miserable until the first SSRIs became available - they were the first drugs that had any noticeable effect and they changed our lives.

gruseom 13y ago
Anecdote it may be, but I'm glad to hear that. I know both sides of that coin myself.

Retric 13y ago

Mental illnesses like cancer are defined by there effects not there underlying cause. As such they tend to treat a wide range of physical causes under the same general umbrella. And drugs or therapies which work wonders for some people often have little benifit for others. However, there are well studdied physical mechanisms for things like specific types of addiction which do present a physical mechanism. A great example of this is Heroin addiction and alcohol addiction are vary different biochemically. There is even a wide range of Alcholhol addictions, but you will see them treated side by side. The simple truth is we can't really understand why someone has issues but we do have treatments that help people lead productive lives and that's enough to be useful. Just as your dentist does not need to know how pain killers work to use them effectively.

DanBC 13y ago

Your comment is far more nuanced. You're reading the science. While I gently disagree with you I understand what you're saying, and I could fairly easily be persuaded that you're right. The comment I replied to had some worrying flags. It made blanket statements about meds; it suggested that depression has a "root cause". I took that to be a reference to classical psychotherapy -- some event happened a long time in the patient's past and that event must be uncovered by the therapist for the patient to address it and recover. Re-reading the post I see that I might be wrong! Perhaps the post is just talking about regular evidence based cognitive model. > The belief that mental illness is biological Well, for something like bipolar disorder this feels like it's true. I don't know much about BPD, and I don't know many people in real life who have it. But it seems that people need the meds. When talking about depression I recognise that there are various types. I'm not a doctor and have no special knowledge. When people ask I suggest they investigate talking therapies (and probably CBT) first. But I know that people might get benefit from meds, especially if they're on the more severe end. I know that meds can have unpleasant side effects.

gruseom 13y ago
It's great when someone replies as thoughtfully as this, and I love your phrase "gently disagree". Would that more disagreements were of this kind. > Well, for something like bipolar disorder this feels like it's true. I agree. The open question is on what level these things are best to be understood. In some sense everything about us is biology, just as in some sense everything is physics. But we don't think of, say, heart disease as best addressed by physics. Similarly, that there is a biological stratum to our actions and feelings does not automatically mean that chronic emotional suffering is best understood as biology. One might as well conclude from the tongue and larynx that language is biology too. And indeed it is, sort of—yet its meaning lies elsewhere. Or you might as well conclude from the importance of neurons to learning that education is brain biology. Why bother with teachers or schools? We should just 'learn' the neurons directly. Such examples are obviously silly, at least given our current knowledge. To know that chronic emotional suffering is a biological disorder requires more than the involvement of biological phenomena in it; it requires an experimentally verified model. My understanding is that we don't have anything close to that. The one that entered the public imagination, the serotonin-deficiency theory, is widely dismissed by experts [1,2,3,4]. The only argument seems to be whether they knew it was false from the beginning or discovered that it was false decades ago. Yet we insist that modern science has discovered that mental illness is biological just the way that cancer is and so on. What do you call a conclusion like that which goes far beyond what we actually know? It's worth realizing that psychiatry has always made this claim. The grounds for it shift every 20 years or so, and the previous grounds are always dismissed as ludicrous if not harmful (think lobotomies etc.). Nevertheless we're perpetually certain that we're beyond all that now. > But I know that people might get benefit from meds I don't think anyone questions that; the question is whether the benefit is that of a placebo or not. My understanding is that SSRIs are no more effective a treatment for depression than tricyclics were a generation earlier; their advantage is rather that they have fewer side effects [5]. So whatever explanation there is for their efficacy must plausibly explain how both of those (presumably very different?) biochemical mechanisms could do it. Given that even inert placebos produce most of the same effect (Kirsch's finding), the placebo explanation is pretty clearly a major candidate. If there's another, I'd like to know about it. (I realize you were talking primarily about bipolar disorder, but I haven't read about that, so I've continued to talk about depression instead. That may lessen the relevance of the above.) [1] http://www.npr.org/blogs/health/2012/01/23/145525853/when-it-comes-to-depression-serotonin-isnt-the-whole-story http://www.npr.org/blogs/health/2012/01/23/145525853/when-it... [2] http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020392.t001/largerimage http://www.plosmedicine.org/article/info:doi/10.1371/journal..., via http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020392 http://www.plosmedicine.org/article/info:doi/10.1371/journal... [3] http://www.psychiatrictimes.com/blogs/couch-crisis/psychiatry%E2%80%99s-new-brain-mind-and-legend-%E2%80%9Cchemical-imbalance%E2%80%9D http://www.psychiatrictimes.com/blogs/couch-crisis/psychiatr..., paywall bypassable by clicking link at http://goo.gl/3LxUW http://goo.gl/3LxUW [4] http://www.madinamerica.com/2012/01/revising-the-history-of-the-serotonin-theory-of-depression-2/ http://www.madinamerica.com/2012/01/revising-the-history-of-... [5] http://www.ncbi.nlm.nih.gov/pubmed/17636689 http://www.ncbi.nlm.nih.gov/pubmed/17636689. This looks like it was withdrawn because it was to be superseded by a larger study, but I couldn't find that one.