The take-home message for me was that the lack of consensus in diagnosing psychiatric disorders interferes with the development over time of truly effective treatments, which I think does reflect the current state of affairs.

Furthermore, I think the diversion of effort over the past couple of decades from behavior toward neurology ("neuroscience") has exacerbated this problem in a parallel manner . That is, attempting to come up with a neurological explanation of a behavioral or psychological disorder, when there is poor consensus regarding the diagnosis of the disorder (or whether there is a disorder, or if so, how many), may be great fun, but is unlikely to be successful.

On the other hand, what may really be going on in psychiatry and psychology is the same kind of incremental process described in the article, but at a much, much earlier stage, where the requisite precise description of the syndromes IS the goal.

Greg Shenaut

Interesting points. Here is my two cents.

There are a variety of differences between psychiatric disorders and medical disorders like cancer or heart disease that make the problems discussed in this article very difficult to solve. I will contrast depression with CVD to illustrate my point.

1) With psychiatric disorders, It is not as easy to see what is broken. CVD is defined by calicification of the arteries. What is the parallel process in depression. We have vague ideas that seratonin is involved and that amygdala/prefrontal cortex (gleaned via neuroscience) may also be involved.

2) With CVD, we have a list of modifyable risk factors for CVD (smoking, overeating, sedentary lifestyle etc)-owing to the Framingham Heart Study, the Alameda County Survey etc. As a society, we have taken steps to educate people about how to minimize their risks. For what it is worth, I suspect the drop in CVD mortality between 1980 and 2000 has more to do with dramatic reductions in smoking during the 1960s/70's.

We have also identified factors that signal vulnerability to depression (parental history of depression, early death of a parent, child abuse, poverty). As you might expect, these issues are much more intractable and more difficult to prevent. Moreover, we seem unwilling as a society to treat mental illness as a public health problem-so there is an absence of primary prevention programs to reduce depression (at least in comparison to CVD prevention).

3) It may be that drugs are NOT the best way to treat (at least some psychiatric problems. Data clearly show that agoraphobia (with and without panic attacks) and specific phobias are best treated with a specific kind of behavioral therapy called "Exposure and Response Prevention (ERP)." Insomnia is better treated with "Cognitive Behavior Therapy for Insomnia (CBT-I)" than Ambian. And depression is best managed with a combination of drugs and CBT.

Let me be clear. I am not saying that drugs are not essential for management of some disorders (schizophrenia, bipolar illness etc). Nor am I saying that drugs are not an important part of managing other psychiatric disorders or may work better than psychotharapy for some people (e.g., depression). Moreover, there is some exciting research that shows that an old antibiotic (sorry can't remember the name) can be used to enhance the effects of ERP thereapy.

However, I would argue that a system wide failure to include large scale longitudinal studies of the effects of behavioral interventions (or combinations of behavioral interventions with medication) has retarded progress on this front.

gshenaut:

The take-home message for me was that the lack of consensus in diagnosing psychiatric disorders interferes with the development over time of truly effective treatments, which I think does reflect the current state of affairs. Furthermore, I think the diversion of effort over the past couple of decades from behavior toward neurology ("neuroscience") has exacerbated this problem in a parallel manner . That is, attempting to come up with a neurological explanation of a behavioral or psychological disorder, when there is poor consensus regarding the diagnosis of the disorder (or whether there is a disorder, or if so, how many), may be great fun, but is unlikely to be successful. On the other hand, what may really be going on in psychiatry and psychology is the same kind of incremental process described in the article, but at a much, much earlier stage, where the requisite precise description of the syndromes IS the goal. Greg Shenaut

I agree with the original message that the main problem in psychiatry is diagnostic, however, I believe the cause of the problem is in the nature of the disorders themselves. Diseases like cancer or tuberculosis are observable phenomena. You can see a cancer, and even physically remove a cancer. If cancer occurs in areas like the colon, you can even remove the colon to eliminate the cancer. Subsequently, the drugs used to treat these disease are observable phenomena themselves. Even the course of the treatment is entirely observable. These issues makes it easy to develop effective treatments from old school drugs, because all of the important effects are visible.

Psychiatric disorders are unobservable phenomena that impact an unobservable entity: the mind. You can't see or remove a depression or a schizophrenia. Instead, the mind is an interpretation of behavioral and neurological observations (depression is not a behavioral disorder because people who are mentally normal can present depressive output). Because of this, diagnosis becomes not the observation of physical phenomena but its interpretation. Furthermore, developing drugs to treat an interpretation becomes even more difficult, as how can you directly measure the impact of the medication on the disorder? You can only interpret the effect of the drug.

Given the nature of mental disorders, diagnosis requires the observation of both brain and behavior, as the only way we can interpret the mind is by observing its output (behavior), and examining its physical processing system (brain) for defects. Its kind of like trying to fix some computer program, but all we have access to is the program's output and the computer's hardware.

Therefore, the shift to neurology is essential for developing better diagnostic assessments and treatments for mental disorders; understanding the brain will give us an edge in interpreting the mind by providing additional information that the output cannot provide.

As you have correctly pointed out, the interpretations we make from neurological data need to reflect the behavioral data as well. We cannot merely come up with a neurological explanation that defines the disorder, because the disorder is defined by the mind not the brain. Regardless, our ability to measure neurological phenomena in the live human brain is new and underdeveloped. We will develop better technologies to better understand our hardware. That information, in turn, will inform and refine our hypotheses regarding behavioral data. Thereby narrowing our diagnostic criteria to precise definitions.