Skip to main content

View Diary: NIMH rejects DSM-V (68 comments)

Comment Preferences

  •  Even with as little respect as I have for the DSM, (18+ / 0-)

    I sense an implied view in your "translation" that things like depression and anxiety disorders etc. can now be (or, rather, relatively soon can be) diagnosed solely in terms of neuroscience.

    Ain't gonna happen. First, it may just be the wrong level to look at, like your auto mechanic figuring out what's wrong with your car using basic physical laws of motion and electromagnetism, or a plumber solving equations in fluid dynamics.

    Second, It's a mistake to think that psychological disorders are necessarily caused by the underlying neuroscience. Even what you quote singles out "cognitive" factors: How people think. If you think that nothing you can do will make things better because of your past experience (brought up in an abusive household, etc.), that's not a neuroscience issue.   It may, in principal it may be represented in terms of the neuroscience, but likewise, in principal it may also be represented in terms of quantum mechanics. Which gets us back to my first point.

    There are two kinds of people in this world: Those who fit into one of two mutually exclusive categories, and those who don't.

    by zhimbo on Sat May 11, 2013 at 02:36:17 PM PDT

    •  I think the point is that we don't know (3+ / 0-)
      Recommended by:
      Sunspots, elmo, Ralphdog

      that looking at lower levels such as genes is the "wrong" level. Maybe instead of boosting serotonin levels for "depression" we should be identifying specific genes that need to be upregulated and upregulating them. E.g. talk therapy could be -- when allied with lifestyle changes -- a roundabout and frequently ineffective way of upregulating those genes.

      •  Interestingly, there are two regions that are (7+ / 0-)

        generally different on imaging studies, one in the cortex and one further back involved in emotions. Cognitive behavioral therapies can lift depression and imaging shows changes in the cortex; medication can lift depression and imaging shows changes in the other area. When both are used, both areas change and the relapse rate is much lower than for either alone. I don't have the studies at hand for a cite, though.

        Is it true? Is it kind? Is it necessary? . . . and respect the dignity of every human being.

        by Wee Mama on Sat May 11, 2013 at 02:53:14 PM PDT

        [ Parent ]

        •  Exactly -- and maybe calling both "depression" (3+ / 0-)
          Recommended by:
          Sunspots, Geenius at Wrok, elmo

          i.e. trying to force the data into DSM nosology -- isn't helpful here.

          Perhaps there are two different processes -- one dependent on epigenetic factors and resulting in differences in neural receptor profiles in the serontonergic parts of the brain, and another cortical process involving cognition and more responsive to talk therapy. Perhaps the two should not be lumped together.

          I'm not claiming that this is exactly right -- I don't know. But it is the sort of understanding that we need to improve, and just binning these processes according to the DSM classifications -- "depression" in this case -- is not helpful. It is just to crude and subjective. We need data and measurement based ways to discuss the processes.

      •  Wouldn't it be pretty to think so. (7+ / 0-)

        I would say that the appropriate response is: "A plague on both your houses." Being retired and long out of practice, I haven't bothered to get a copy of DSM-V, but the universal reaction seems to be that the committee has engaged in massive violations of Occam's Razor -- I.e., extensively  multiplying entities needlessly.   On the other hand, NIMH seems to be throwing out the proverbial baby in a move toward hyper-reductionism that doesn't fit what is known as the systemic nature of much psychopathology, a reversion to
        the pursuit of the mythical schizococchus.  

        And speaking of systems, this situation certainly would seem to be a disaster for clinical researchers, since it looks like one set of diagnostic criteria for clinical/insurance diagnosis and an extremely different one for research.


        "If you don't read the newspapers, you're uninformed. If you do read the newspapers, you're misinformed." -- M. Twain

        by Oliver St John Gogarty on Sat May 11, 2013 at 03:02:31 PM PDT

        [ Parent ]

      •  Well, we should certainly be looking everywhere (8+ / 0-)

        because we have a long way to go in our understanding.

        (This is one of my soap box issues, though, so bear with me...)

        But there are inherently psychological questions.  To step away from mental disorders, questions like: "What's the best way to study for an exam?" or, "What's the most effective layout for a an airplane cockpit?" may, perhaps, be partially illuminated by looking at underlying neuroscience, but so far the best answers to those questions by far are framed in cognitive terms, and it's hard to even imagine a useful answer that isn't primarily in the language of psychology as opposed to biology.

        Likewise, we shouldn't assume as a default view that advancing our understanding of mental problems necessarily implies a more biological/genetic understanding.  In many cases, that's going to be barking up the wrong tree.

        The problem with the DSM isn't that it cuts off routes of inquiry, it's that it is, by design, atheoretical. It - on purpose! - doesn't look at causes.   That makes it pretty damn useless indeed for furthering our understanding at any level of analysis.

        There are two kinds of people in this world: Those who fit into one of two mutually exclusive categories, and those who don't.

        by zhimbo on Sat May 11, 2013 at 03:17:42 PM PDT

        [ Parent ]

        •  We have reverse engineered vision (0+ / 0-)

          well enough to build computer systems that can recognize faces in real time. (This turns out to be much easier than one might think -- you actually only need a smallish number of neural classifiers for "basic" faces, and can build classifiers for other faces out of this basic set).

          It may be decades before we have a similar level of understanding for all of human cognition, but this is clearly the direction things are going... hence, the US BRAIN project, the EU's Human Brain Project and many other iniatives.

          •  Sure. (5+ / 0-)

            Vision science is awesome.  My background is in cognitive neuroscience, and I've studied visual recognition earlier in my career. But there's no reason at all to believe the same approach will apply to say, "studying for exams" or "dealing productively with problems" or "becoming comfortable with ones appearance".  And even visual processes can't be understood solely in terms of neuroscience.  There are cultural differences in how visual attention is directed in complex scenes; some visual illusions depend strongly on cultural backgrounds, too. To understand that, you have leave the neuroscience level of analysis.

            Neuroscience alone tells us nothing about psychology.  The language of neuroscience is about ionic flow, action potentials, cortical connectivity, etc.,.You need the psychological level to make neuroscience speak to psychological needs. An fMRI image is just an XYZ graph of statistical values related to blood flow.  Until you make connections with psychological theory, that's pretty much all it is.

            There are two kinds of people in this world: Those who fit into one of two mutually exclusive categories, and those who don't.

            by zhimbo on Sat May 11, 2013 at 04:21:38 PM PDT

            [ Parent ]

            •  There are ways to identify features in fMRI images (0+ / 0-)

              without knowing how to classify what you are seeing in advance. There is quite a range of statistical methods and machine learning methods that can be applied. The future of neuroscience is computation-intensive.

              My guess is that cognitive neuroscience/psychology will continue to cross-fertilize as this research proceeds, and that a combination improved scanning and mapping technologies and computer models based on those neural maps and trained on an individual's scans will have a lot to tell us even about what is going on when we are studying for an exam.

              •  "Identify features" (4+ / 0-)
                Recommended by:
                pigpaste, Samer, RonK, Kimbeaux

                is not the same thing as "useful for psychology".

                Those features must be mapped onto psychological theory to be useful psychologically.  You may not have to start with psychological theory, but if you don't ever end up there, it's just a picture of blood flow.

                I'm not anti-neuroscience, of course. Quite the opposite!  But the idea that "progress in psychology" MEANS "understanding the neuroscience" is simply wrong.

                "will have a lot to tell us even about what is going on when we are studying for an exam."

                "what is going on" in terms of the neuroscience isn't obviously the right kind of information we need to address the needs of a student.   XYZs coordinate of blood flow, no matter how identifiable, don't tell anyone how to study until it is related to psychological theory.

                Ed Smith, who was an early pioneer in using PET and MRI technology to study cognitive issues, was once asked (I was there, this is an eyewitness account) why the new neuroscience methods were so promising and useful. His answer was remarkably simple and mundane: "They provide new dependent measures." Basically, our testing of psychological theory could be more effective with additional ways of testing it. It isn't about replacing psychological theory with neuroscience.

                There are two kinds of people in this world: Those who fit into one of two mutually exclusive categories, and those who don't.

                by zhimbo on Sat May 11, 2013 at 05:02:34 PM PDT

                [ Parent ]

                •  Neuroscience isn't one fixed level of explanation (0+ / 0-)

                  I'm taking a p.o.v. that included genomics (gene variant), epigenetics (gene expression variation), individual neurons, neural circuits, neural columns -- and all the algorithms required to represent what this whole mess computes.

                  I'm not a dualist, so that inevitably adds up to the mind -- with all its psychology.

                  Now, how all that will finally map back onto the cognitive psychology of 2013 I don't know. Unless cognitive psychology is as entrenched and unwilling to change as the American Psychiatric Association (and I have no reason to think it is), it will participate in building the computer models for cortical processes involved in cognition. I expect much will be learned in this area over the next two decades, and that cognitive psychology will come to look very different as a result.

                •  See, for example, Cori Bargmann on the likelihood (0+ / 0-)

                  of never understanding even roundworm learning in terms of circuitry.

                  •  Citation? (0+ / 0-)

                    I doubt he is advocating dualism or vitalism for C. elegans . More likely he is advocating the use of models with many biological levels.

                    •  She, and she's a geneticist and (0+ / 0-)

                      molecular biologist. Why do you assume anyone would be advocating vitalism?

                      •  Citation? (0+ / 0-)

                        I see her lab does a lot of cell signaling work. Where are her remarks about the impossibility of understanding C. elegans learning? Or perhaps you saw something about the importance of cell signaling in C. elegans learning, and thought that must mean neurons aren't involved? Neurotransmitters are signal molecules, you know.

                        •  I'm well aware of her work, thanks. (1+ / 0-)
                          Recommended by:

                          Btw, it's always dumb to make condescending assumptions about any kossacks' knowledge and talk down to them in the absence of evidence that they don't know what they are talking about. When you know nothing about another person here, it's always a good idea to start by crediting them with some education and expertise at least as great as your own, especially on technical subjects.

                          She made that comment in an interview, but I haven't found it on the web.

              •  No, what we should be working on is whole (0+ / 0-)

                brain emulation at the cellular level.  That combined with the ability to copy a person's entire brain will let us collect all the data we need to make a perfect model using only neuroscience.  Not to mention some measure of immortality as well.

                You have watched Faux News, now lose 2d10 SAN.

                by Throw The Bums Out on Sat May 11, 2013 at 05:13:48 PM PDT

                [ Parent ]

                •  The European Human Brain Project (1+ / 0-)
                  Recommended by:
                  Throw The Bums Out

                  is somewhat like that -- an attempt to model the whole brain -- but I don't think they are trying to model housekeeping functions of neurons nor all the immune system-like properties we are finding in glial cells.

                  The American BRAIN project is more about improving mapping and scanning technology. It was called the Brain Activity Map (BAM) project until the Obama administration renamed it.

                  Both sorts of approaches are needed.

                  •  But will it emulate a human brain well enough (0+ / 0-)

                    to actually achieve full human level sentience including the ability to know right from wrong independent of any programming?  Or even well enough to actually upload a copy of a real person's brain immediately after their "death" (currently theoretically possible, but it requires slicing the brain and scanning each slice with an electron microscope) to it and keep them alive and fully conscious indefinitely?

                    You have watched Faux News, now lose 2d10 SAN.

                    by Throw The Bums Out on Sat May 11, 2013 at 05:33:07 PM PDT

                    [ Parent ]

                    •  I doubt that the issue of machine sentience (1+ / 0-)
                      Recommended by:

                      will be facing us at any time in this project, during the next 10 years.

                      And unless some really incredible improvements in scanning technology appear, uploading people isn't in the prospects for the next 10 years, either.

                      My personal suspicion is that we are going to be much more concerned with global warming issues in this century than with the Singularity.

                      •  However, if you have whole brain emulation thanks (0+ / 0-)

                        to the European Human Brain Project and the  American BRAIN project working on scanning technology then isn't it only a matter of time before the Singularity?  I mean, if I told you 20 years ago that you would have a computer with 2GB of RAM and a 64GB hard drive in your pocket you would have said I was crazy.

                        As for uploading a person, you would probably have to remove their brain and scan it slice by slice with an electron microscope unless the American BRAIN project comes up with something really impressive.  But still, we will never reach the Singularity with the DSM getting in the way.  Whether or not it is better to use the existing DSM as a stopgap or to get rid of psychiatric treatment altogether until the new NIMH model comes out is the real question and one I am not equipped to answer.

                        You have watched Faux News, now lose 2d10 SAN.

                        by Throw The Bums Out on Sat May 11, 2013 at 05:48:27 PM PDT

                        [ Parent ]

                        •  In the case of the Singularity (0+ / 0-)

                          I'd be a lot more worried about global warming getting in the way than the DSM.

                          What actually becomes of the DSM in the near future is that it gets replaced for medical purposes by ICD10. ICD10 is an international list of diseases maintained by WHO. The key difference is that the American Psychiatric Association doesn't control ICD and can't put things in it because some American drug company bribes them to.

                          ICD will work as the stopgap until better, science-based diagnoses for mental illnesses are developed. But the APA will not be in that loop.

                          They have made themselves irrelevant to the future of medicine.

                  •  Oh, and have the people working on the American (0+ / 0-)

                    BRAIN project thought about a way to write a brain image map onto a real brain once it is done being manipulated or do they only care about capturing minds (or souls, or dynamic full brain process inventories, or whatever) and not being able to write them back to a real brain?

                    You have watched Faux News, now lose 2d10 SAN.

                    by Throw The Bums Out on Sat May 11, 2013 at 05:38:02 PM PDT

                    [ Parent ]

            •  Zhimbo, I could not agree more. We know far too (3+ / 0-)
              Recommended by:
              Kimbeaux, pigpaste, princesspat

              Little about mental disorder at this point to go directly to the level of behavior or cognition equals x or y neuron or circuit/ pathway. Without a coherent theory, we are groping in the dark for correlations. We are a long way from specific bio markers for each (or any)' mental disorder.

              "There is nothing - absolutely nothing - half so much worth doing as simply messing about in boats ..." - Kenneth Grahame -

              by RonK on Sat May 11, 2013 at 08:24:26 PM PDT

              [ Parent ]

              •  We are not a long way from being able to research (0+ / 0-)

                how the biomarker and neural data cluster, independently of the diagnostic categories cooked up in DSM. That is what NIMH is doing. And it represents a measurment driven approach to a future classification system.

                The fact that we haven't got the results of that research yet is not a reason to believe that the categories of DSM-V are true or fundamental.

                Brain science is certainly not lacking a coherent theory to create a new classification system based on measurements. Frankly, the old clinical psychology and personality theory domains that sort of vaguely underpin DSM lacked any coherent theory.

      •  And much 'mental illness' is socioeconomic. (0+ / 0-)

        It's astonishing how high the rate of depression and anxiety are among formerly middle class people who are laid off, evicted, unemployed or otherwise crushed by the corporate juggernaut. Or among the working poor who have been systematically excluded from the 'American dream' for decades now.

        A real solution would be labor rights, a living wage, universal health care and economic justice.

        Instead BigPharma has a DSM-V label and a pill. A pill that costs hundreds to thousands of dollars per year and has lots of side effects, frequently including ruinous weight gain.  

        •  I agree (0+ / 0-)

          DSM is a pill-pushers manual written by Pharma salesmen. They are not even interested in developing new pills for "mental illness" or understanding what would be required to do that, just in expanding markets for the pills they already know how to make.

          Is a person homeless because there is no work? Send her to a "program" to teach her Resume Writing, give her an SSRIs and pep talks about Empowerment and Positive Mental Attitude. It's so much cheap than a real social safety net. And it puts what little money is available in the hands of Pharma and the balkanized non-profit sector that has mushroomed, like a fungus infection, in the hole where government services used to be.

    •  This is an issue (5+ / 0-)

      A biomarker is useful if it correlates with something, typically a symptom or a condition, but that correlation doesn't necessarily exist.

      The fields of genomics and neuroscience have both been burned repeatedly when it turned out that cluster analysis of gene expression data and functional MRI data respectively found  non-existent relationships due to artifacts in the data.

      I'm strongly in favor of evidence based medicine - but only when it is strong, repeatable evidence.

      •  I agree with the need for repeatable evidence (1+ / 0-)
        Recommended by:

        but the DSM-V has such a low reliability in terms of different practitioners arriving at the same diagnosis that its classifications are useless. There is no point in correlating biomarkers to "symptoms" that are not themselves reliable diagnostic indicators.

        •  The classifications are useless only if (0+ / 0-)

          successful therapy is absolutely dependent on them.

          •  Useless for research (0+ / 0-)

            I suppose psychiatrists can take their subjective diagnoses down the route astrologers have taken with their horoscopes, claiming they are useful for therapy even if those nasty reductionist scientists don't understand why.

            •  There are drugs that really work for clearly (2+ / 0-)
              Recommended by:
              AaronInSanDiego, HiBob

              organic diseases and yet we don't know how they do the job, but we still use them as we research their mechanisms. As far as I can tell from the diary, Insel isn't saying current descriptions based on symptoms are useless for therapy and should be abandoned in therapeutic situations, he's saying those diagnoses are not a sound basis for research categories. Two different things. This is just happening in cancer, so it's not like psychiatry is centuries behind medicine.  In practice, psychologists pick and choose from the DSM anyway.

              •  I would further distinguish between psychiatrists (1+ / 0-)
                Recommended by:

                and the APA, a professional organization that exists to enhance the incomes of psychiatrists.

                There are psychiatrists who do medical research and would be considered part of the brain science community.

                What is really at stake here is the political power of the APA within medicine. The APA has so discredited itself with DSM-V that the rest of medical science is ready to give them the boot. We don't need DSM -- MDs can use ICD10 instead for billing purposes and prescriptions. The difference is ICD10 is not controlled by the APA, and the APA can't inject new "diagnoses" into it as quid pro quo for drug company junkets and "consulting" fees.

                In the meantime, scientific medicine can develop a new, measurement driven nosology for mental illnesses. No doubt, there will be MDs in psychiatry involved in that. But the APA won't be running it.

                And shouldn't be running it.

                •  Don't most professional organizations (0+ / 0-)

                  exist primarily to benefit their members?

                  Gondwana has always been at war with Laurasia.

                  by AaronInSanDiego on Sat May 11, 2013 at 07:37:11 PM PDT

                  [ Parent ]

                  •  It's as if the American Physical Society (1+ / 0-)
                    Recommended by:

                    were discovered to be inserting new particles into the standard model because some industry was bribing them to do so.

                    Except that couldn't happen in physics because the international community in physics is stronger than any national organization.

                    However, psychiatry is frankly so backward it is still nationalistic, and the American Psychiatric Association still issues a manual of "mental illnesses" that has power in America (and Canada) -- but nowhere else.

                    The American drug companies have bribed the APA into expanding markets for their existing pharmaceutical products. They use this to try to leverage their products into markets in Europe and elsewhere.

                    A good example is ADHD and ritalin. A ridiculous number of American children take ritalin. It makes no evolutionary sense the such a large fraction of the children of a species would have such a disorder and need a drug for it. But Americans will believe what the APA tells them. In Europe, and other parts of the world, there is much less readiness to believe that large numbers of children have this American illness ADHD and need to take a drug that enriches American drug companies.

    •  Well you enjoy thinking that while thanks to (0+ / 0-)

      the NIMH I will not only end up being fully cured but immortal as well thanks to a greatly increased knowledge of neuroscience.

      You have watched Faux News, now lose 2d10 SAN.

      by Throw The Bums Out on Sat May 11, 2013 at 05:41:16 PM PDT

      [ Parent ]

    •  Rec'd for bringing up the point about (0+ / 0-)

      levels of analysis. I'm a software developer. When I'm trying to fix a bug in a program, a better understanding of solid-state physics won't help me fix it. This is the case even though everything that happens in a computer ultimately comes down to solid-state physics, which by the way is far, far, far better understood than anything involving the human brain.

      What's happening here is that the execution of my program is an emergent phenomenon of the physical properties that allow the logic gates in my CPU chip to work. The very nature of an emergent phenomenon is that, while it's completely and absolutely dependent on the underlying phenomena it emerges from, it behaves completely differently, on a macroscopic scale, from the way the underlying phenomena behave. And that means that understanding it cannot be achieved solely by understanding the underlying phenomena. To claim that it can is to commit the logical fallacy of composition (which, along with its converse, the fallacy of division, falls into the class of "category mistakes".

      Thus in order to debug my program, I have to take an attitude that might strike some people as "dualist", even though it does not require true dualism. And I think the same is true of human behavior. Neuroscience can certainly identify constraints on how we can behave; it sets conditions that are necessary, but not sufficient, for any proposed explanation of human behavior to be true (trivial example: some therapists claim to have helped patients "recover" visual memories of being a blastocyst traveling through their mom's Fallopian tubes; it's neuroscience (and not particularly advanced neuroscience) that tells us that you simply can't form visual memories when you don't have a brain or eyes and there's no light in your environment).

      But it's a giant leap from that to assume that a full understanding of neuroscience is equivalent to a full understanding of human psychology. Emergent phenomena just don't work that way, any more than Shakespeare's plays can be completely understood in terms of the chemistry of paper and ink.

      Sometimes truth is spoken from privilege and falsehood is spoken to power. Good intentions aren't enough.

      by ebohlman on Sun May 12, 2013 at 12:21:26 AM PDT

      [ Parent ]

      •  Except that neuroscience includes the algorithms (0+ / 0-)

        not just the biochemistry of ion channels. It's the study of how and what neurons compute. All human "behavior" involves neural computation. Even body functions that are not under conscious control, such as digestion, are regulated by the parasympathetic nervous system.

    •  Maybe that's just the point (0+ / 0-)

      . . . that some psychological conditions are biologically/chemically based, while others are simply based in fallacious thinking that can be unlearned, and the DSM would lump them all together as conditions that should be treated with medication.

      "The great lie of democracy, its essential paradox, is that democracy is the first to be sacrificed when its security is at risk. Every state is totalitarian at heart; there are no ends to the cruelty it will go to to protect itself." -- Ian McDonald

      by Geenius at Wrok on Sun May 12, 2013 at 06:12:01 AM PDT

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site