why don't we routinely do LPs to look for decreased serotonin in the csf

[nancysinatra]

---

Members don't see this ad.

Does anyone know? I learned while making my futile and dispiriting annual last ditch effort to improve my prite score that there are decreased levels of serotonin in the csf of depressed people. And some urinary measures too! (Now I can't remember what they are.)

Ok I can see that the risk probably outweighs the benefits most of the time. But I could think of cases where you'd want a hard answer.

If we HAD a reliable biomarker for depression or any other psychiatric illness, where would we be? We'd still have all these people complaining of symptoms, and a lot of them might not have the lab results to confirm the illness.

Then we'd have psychosomatic psychiatric illnesses.

 

[dl2dp2]

If we HAD a reliable biomarker for depression or any other psychiatric illness, where would we be? We'd still have all these people complaining of symptoms, and a lot of them might not have the lab results to confirm the illness.

Believe me you are not the only one who thought of this. The short answer is all attempts in developing a useful biomarker in psychiatry for any practical purpose have FAILED miserably either because there's too much of a overlap between control and patients or because the assay is unreliable.

In your case, on AVERAGE 5HT in the CSF in a depressed patient is lower. But that's AVERAGE. the overlap is HUGE. If you measure a bunch of CSFs you'll eventually hit that p<0.05 and get a paper out. Doesn't mean any iota of it is useful clinically. You need to think of a realistic clinical scenario where such a biomarker is useful: could CSF 5HT level differentiate a malingerer from someone who's genuinely depressed? HELLS NO. Could the CSF 5HT level give you an indicator as who's gonna respond to SSRIs? No way. You need to start thinking about how a viable biomarker is developed, ROC analysis, etc.

The same problem plagues the neuroimaging literature. Oh the dorsalateral PFC in schizophrenics is LESS active than normal controls. Can we PLEASE make a biomarker out of it? NO. Guess why? Because the overlap is ridiculously large. I'm not sure why people are still doing this kind of research...haven't we "generated" enough "hypotheses" already?



If you can somehow cleverly solve this problem you'll likely be a Jedi.

 

[nancysinatra]

Believe me you are not the only one who thought of this. The short answer is all attempts in developing a useful biomarker in psychiatry for any practical purpose have FAILED miserably either because there's too much of a overlap between control and patients or because the assay is unreliable.

In your case, on AVERAGE 5HT in the CSF in a depressed patient is lower. But that's AVERAGE. the overlap is HUGE. If you measure a bunch of CSFs you'll eventually hit that p<0.05 and get a paper out. Doesn't mean any iota of it is useful clinically. You need to think of a realistic clinical scenario where such a biomarker is useful: could CSF 5HT level differentiate a malingerer from someone who's genuinely depressed? HELLS NO. Could the CSF 5HT level give you an indicator as who's gonna respond to SSRIs? No way. You need to start thinking about how a viable biomarker is developed, ROC analysis, etc.

The same problem plagues the neuroimaging literature. Oh the dorsalateral PFC in schizophrenics is LESS active than normal controls. Can we PLEASE make a biomarker out of it? NO. Guess why? Because the overlap is ridiculously large. I'm not sure why people are still doing this kind of research...haven't we "generated" enough "hypotheses" already?



If you can somehow cleverly solve this problem you'll likely be a Jedi.

So it's neither sensitive nor specific?

I just want to be able to make accurate diagnoses, like surgeons who diagnose appendicitis and rule out cholecystitis, and then go and ACTUALLY rule it out. I don't like how psychiatry is this thing where people say they have 25 different problems and we never figure out what they actually have and then they just go on disability anyway. I also realize there are many psychological ailments that will always be there but sometimes that's really what it IS and it would be great to have a test to confirm that. Bipolar is where we need a test the most!

I also don't like pan-labbing everyone who walks in the hospital just to "see" if there's something wrong with them when we're not looking for something specific or monitoring for certain drugs that we use.

 

[loveoforganic]

If what I'm saying is obvious, sorry, I don't mean to say basic things - just don't know where your knowledge is at. The specificity and sensitivity aren't inherent to the test - they depend on the thresholds set to get the best combination of the two, while keeping in mind the disease prevalence (and the concomitant effects on positive and negative predictive values) and the associated ethical implications of false positives and negatives.

With huge overlap and wide variability in the predictive measure in both the normal and diseased populations, your "ideal" mix of the two would be crappy.

Uploaded an incredibly crude pic with 2 curves, one disease status and one normal pop drawn in (doesn't really matter which), and with rough guesses for mean and median of both. You could see how you would find a statistically significant (and legitimate) difference between the two populations in a large study, but not a difference that is amenable to testing.

My understanding of it at any rate

 

[michaelrack]

Ok I can see that the risk probably outweighs the benefits most of the time. .

also painful and expensive. testing CSF for hypocretin is useful under certain clinical circustances for evaluation for narcolepsy, but rarely done because of these reasons (even though you can often get the research lab at Stanford to cover the assay, there is still the cost of the LP itself.)

 

[nitemagi]

How would it change management? Is there evidence that CSF levels return to normal once someone is euthymic? (I'm seriously asking this since I don't know)

 

[OldPsychDoc]

How would it change management? Is there evidence that CSF levels return to normal once someone is euthymic? (I'm seriously asking this since I don't know)

And consider this scenario--
"Doctor, I'm so depressed..."
"No you're not--we did a test and all your brain chemicals are normal."

 

[dl2dp2]

And consider this scenario--
"Doctor, I'm so depressed..."
"No you're not--we did a test and all your brain chemicals are normal."

Exactly, there's no reliable way to tell who's a malingerer and who isn't.

 

[digitlnoize]

And consider this scenario--
"Doctor, I'm so depressed..."
"No you're not--we did a test and all your brain chemicals are normal."

Exactly, there's no reliable way to tell who's a malingerer and who isn't.

Or they're not malingering, really DO feel depressed, but the "test" is normal.

Honestly, this seems a bit extreme. An LP is a pretty invasive procedure, and not without complications. We often stick things into people too quickly for things that can often be determined clinically with reasonable accuracy.

 

[nitemagi]

It's hard enough to get an LP done on a delirious patient!

 

[whopper]

Agree with the above. While the test isn't great, if someone's depressed, and there's good reason to believe so, a test is overkill.

If a person's depressed, whether or not the test has low serotonin isn't going to matter much given what we know. Unless there's reason to believe they are malingering or have factitious disorder, there's no point to it. How would you feel if someone wanted to do an uncomfortable test that could cause meningitis that was not needed?

Even in the case of malingerers, there's not yet been any organized decision making process as to when to include this if at all. Most people who malinger use psychosis, mania, or PTSD. While in that case, I can see some purpose, the sensitivity and specifity aren't great. A person could refuse such a test for a malingering evaluation, and given the nature of the test, it'd be understandable to arbitrators such as a judge as to why they'd refuse it.

Personally, I would consider requesting an LP in a malingering evaluation. While it's not proof positive either way, it could add weight to the evaluation in combination with other factors. I'd see it possibly having some merit if the test showed low serotonin and because the person was willing to sit through an uncomfortable procedure to prove their point.

 

[PETRAN]

Agree with the above. While the test isn't great, if someone's depressed, and there's good reason to believe so, a test is overkill.

If a person's depressed, whether or not the test has low serotonin isn't going to matter much given what we know. Unless there's reason to believe they are malingering or have factitious disorder, there's no point to it. How would you feel if someone wanted to do an uncomfortable test that could cause meningitis that was not needed?

Even in the case of malingerers, there's not yet been any organized decision making process as to when to include this if at all. Most people who malinger use psychosis, mania, or PTSD. While in that case, I can see some purpose, the sensitivity and specifity aren't great. A person could refuse such a test for a malingering evaluation, and given the nature of the test, it'd be understandable to arbitrators such as a judge as to why they'd refuse it.

Personally, I would consider requesting an LP in a malingering evaluation. While it's not proof positive either way, it could add weight to the evaluation in combination with other factors. I'd see it possibly having some merit if the test showed low serotonin and because the person was willing to sit through an uncomfortable procedure to prove their point.

But still, what about if he is not malingering and the result comes back normal? Since there is no good sensitivity/specificity it would be useless and possibly would do more harm than good.



Anyway, there is no evidence that depression is related to low serotonin levels per se anyway. This is an ancient, over-simplistic theory which is just wrong. Do they still teach it in psychiatry residencies? There were some imaging studies on serotonin receptors (some of them were supposed to be of a lower number) but still was incoclusive since some studies were finding similar, or even greater numbers of various serotonin receptors (with a huge variability in normal population). The newer theories is about serotonin receptor responses (something about lower or higher adaptation rates, correct me if i'm wrong) but it still seems to be simplistic. What about SSRE's which lower serotonin levels in the synapse? (and still having an anti-depressant/anxiolytic effect) And then there are various complicated interactions with other transmitter systems. Maybe its better to view some mental health problems in terms of the neuropsychology underlying the neural circuity (and its' neurochemistry) but still the localization is very poor. The functional changes almost always involve the whole brain with little consistency between studies (e.g. some studies of depression finding over-activation of frontal lobes others under-activation etc.). As others said, the guy who solves these problems would be a jedi or something.

 

[nitemagi]

Anyway, there is no evidence that depression is related to low serotonin levels per se anyway. This is an ancient, over-simplistic theory which is just wrong. Do they still teach it in psychiatry residencies? There were some imaging studies on serotonin receptors (some of them were supposed to be of a lower number) but still was incoclusive since some studies were finding similar, or even greater numbers of various serotonin receptors (with a huge variability in normal population). The newer theories is about serotonin receptor responses (something about lower or higher adaptation rates, correct me if i'm wrong) but it still seems to be simplistic.

Despite what you're learning in grad school, there is a body of evidence that depressed patients have low serotonin levels in their csf. Yes it's still taught as a hypothesis, rather than following just trends in research it's beneficial to have a perspective on how we ended up where we are.

Your brushing off numerous research studies as simplistic presumes that psychiatrists are following a causational model, rather than an association. There is evidence for a correlation, even with improvement in csf levels in those that respond. Causation, sensitivity, specificity, and change in treatment is much more debatable. It's more likely a useful biomarker in specific subgroups we haven't yet teased out (like suicide attempters). Serotonin is probably not the direct cause of depression or response to a med, but may be a piece of one pathway (some like the BDNF hypothesis as a common pathway for all treatments).

It's all simplistic, but if You have some better hypotheses that somehow synthesizes thousands of neuroscience studies, I'd say we're all ears.
Pro--
http://www.ncbi.nlm.nih.gov/pubmed/9616798
http://www.ncbi.nlm.nih.gov/pubmed/20708058
http://www.ncbi.nlm.nih.gov/pubmed/19032905
http://www.ncbi.nlm.nih.gov/pubmed/18940259
http://www.ncbi.nlm.nih.gov/pubmed/17299512
http://www.ncbi.nlm.nih.gov/pubmed/16713589
http://www.ncbi.nlm.nih.gov/pubmed/15380838
http://www.ncbi.nlm.nih.gov/pubmed/12668368
http://www.ncbi.nlm.nih.gov/pubmed/10357041
http://www.ncbi.nlm.nih.gov/pubmed/9018386

Con--
http://www.ncbi.nlm.nih.gov/pubmed/10523046
http://www.ncbi.nlm.nih.gov/pubmed/9004051

 

[whopper]

But still, what about if he is not malingering and the result comes back normal?

This can and does happen. An evaluation must be based on the totality of all the information, not just one test. If a test were say 70% accurate, you're of course going to have outliers that test against the norm.

As for low serotonin in CSF, yes, I also agree that this could one day lead to some possible lab test, but as of right now no. There's also data that low 5HIAA (did I spell it right) also correlates with higher suicide and violence. More data IMHO needs to be done. Remember, doctors aren't just about diagnosing, but also about making a patient feel comfortable and getting them better. Most people wouldn't want an LP because they're already convinced they are depressed, why do they have to prove it to you?

 

[PETRAN]

Despite what you're learning in grad school, there is a body of evidence that depressed patients have low serotonin levels in their csf. Yes it's still taught as a hypothesis, rather than following just trends in research it's beneficial to have a perspective on how we ended up where we are.

Your brushing off numerous research studies as simplistic presumes that psychiatrists are following a causational model, rather than an association. There is evidence for a correlation, even with improvement in csf levels in those that respond. Causation, sensitivity, specificity, and change in treatment is much more debatable. It's more likely a useful biomarker in specific subgroups we haven't yet teased out (like suicide attempters). Serotonin is probably not the direct cause of depression or response to a med, but may be a piece of one pathway (some like the BDNF hypothesis as a common pathway for all treatments).

It's all simplistic, but if You have some better hypotheses that somehow synthesizes thousands of neuroscience studies, I'd say we're all ears.
Pro--
http://www.ncbi.nlm.nih.gov/pubmed/9616798
http://www.ncbi.nlm.nih.gov/pubmed/20708058
http://www.ncbi.nlm.nih.gov/pubmed/19032905
http://www.ncbi.nlm.nih.gov/pubmed/18940259
http://www.ncbi.nlm.nih.gov/pubmed/17299512
http://www.ncbi.nlm.nih.gov/pubmed/16713589
http://www.ncbi.nlm.nih.gov/pubmed/15380838
http://www.ncbi.nlm.nih.gov/pubmed/12668368
http://www.ncbi.nlm.nih.gov/pubmed/10357041
http://www.ncbi.nlm.nih.gov/pubmed/9018386

Con--
http://www.ncbi.nlm.nih.gov/pubmed/10523046
http://www.ncbi.nlm.nih.gov/pubmed/9004051

First of all, i'm not learning that stuff in "grad school", i have a personal interest in the subject-matter, thats all. Well, the majority of research says that "lower serotonin levels" must be false, because if the "serotonin levels" were low, then the SSRIs-or whatever- would directly correct "the imbalance" and bring instant relief. As far as we know thats not the case and it takes several days/weeks until the effect starts to take place, possibly meaning that there must be some gradual cellular/receptor change taking place. Then there are drugs such as the SSREs that also bring relief but work in the opposite way-lowering serotonin levels in the synapse. How would you respond to these criticisms?


As far as i know, the "serotonin-level" hypothesis is dead, but the hypothesis still lives in various "serotonin-receptor" models with various proposals of increased/dicreased number of 5-HT receptors, increased/decreased sensitivity, up- or down-regulation of pre-synaptic auto-receptors etc. (also the BDNF hypothesis that you state, which could be a common pathway or a finding completely unrelated to mood regulation). Well, i guess that if you search hard enough you will eventually find some decreased serotonin levels-whatever that means. There would probably be some "depressed sub-populations" showing some correlation, but this is still very weak scientifically speaking. I could find hundreds more of pro- and con- correlational studies like the ones you post. Furthermore, the purely biological proposals are hugely inadequate IMO, since they say nothing about the neuroscientific/neuropsychological mechanisms per se. E.g. does the serotonin receptor activity correlates with mood state, mood intensity, mood experience, mood excitation/inhibition, mood what? What is the role of the serotonergic system in normal mood and emotional experience? You would need to explicitly state these issues if you want a valid biological test showing correlations between a chemical level and a psychological level (e.g. x level of serotonin reveals y level of change in that aspect of mood experience/intensity/level). This would be truly valid. But maybe i ask for too much.

 

[PETRAN]

As for low serotonin in CSF, yes, I also agree that this could one day lead to some possible lab test, but as of right now no. There's also data that low 5HIAA (did I spell it right) also correlates with higher suicide and violence. More data IMHO needs to be done. Remember, doctors aren't just about diagnosing, but also about making a patient feel comfortable and getting them better. Most people wouldn't want an LP because they're already convinced they are depressed, why do they have to prove it to you?

Yes, i completely agree with you. My response was about the "forensic" aspect. The test would be valid if there was a systematic relationship between-say- 5HIAA and mood intensity, positivity/negativity or something. Like "doctor i'm very depressed", "-oh yes i can see that, your 5-HIAA levels revealed 58 "moodions" which is very low" (or at least "it usually correlates with 2 sds below the standard mood levels or something) 😛 😎

 

[nitemagi]

Well, the majority of research says that "lower serotonin levels" must be false, because if the "serotonin levels" were low, then the SSRIs-or whatever- would directly correct "the imbalance" and bring instant relief...I could find hundreds more of pro- and con- correlational studies like the ones you post.

Many current hypotheses presume further downstream changes after serotonin levels change are responsible for improvement. This doesn't negate that literature exists on low serotonin levels in depressed patients. If you want to cite serious studies that counter the correlation, then do so. Your opinion doesn't counter evidence that there are biological correlates of depression, even if what we have now is very preliminary.

No one here has said the test is diagnostic. That doesn't mean there isn't some validity to the research.

 

[PETRAN]

Many current hypotheses presume further downstream changes after serotonin levels change are responsible for improvement. This doesn't negate that literature exists on low serotonin levels in depressed patients. If you want to cite serious studies that counter the correlation, then do so. Your opinion doesn't counter evidence that there are biological correlates of depression, even if what we have now is very preliminary.

No one here has said the test is diagnostic. That doesn't mean there isn't some validity to the research.

But correlation studies say nothing scientifically conclusive seriously. Correlation studies, factor analysis and stuff like that are done for "invesigative" reasons. They are used in multivariate phenomena (like "depression") that you know nothing about and you want some idea about what is happening. As you said they don't say anything about causations and prediction. A few studies showing "decreased serotonin levels" or decreased metabolites or whatever in a sub-population could mean that a thousand other extraneous variables are leading to the decreased level. From other neurochemical changes relating to serotonin (which would be more directly responsible for mood regulation) to dietary or sleep changes of a few depressed patients.


Ofcourse there is *some* validity in the research since all research has some validity ( unless the papers are written by pharma ghost writers lol) but this doesn't mean that what we can conclude from this research-if we can conclude anything at all from this type of research- is *valid*.

 

[nitemagi]

But correlation studies say nothing scientifically conclusive seriously. Correlation studies, factor analysis and stuff like that are done for "invesigative" reasons. They are used in multivariate phenomena (like "depression") that you know nothing about and you want some idea about what is happening. As you said they don't say anything about causations and prediction. A few studies showing "decreased serotonin levels" or decreased metabolites or whatever in a sub-population could mean that a thousand other extraneous variables are leading to the decreased level. From other neurochemical changes relating to serotonin (which would be more directly responsible for mood regulation) to dietary or sleep changes of a few depressed patients.


Ofcourse there is *some* validity in the research since all research has some validity ( unless the papers are written by pharma ghost writers lol) but this doesn't mean that what we can conclude from this research-if we can conclude anything at all from this type of research- is *valid*.

No one has said it's predictive, but it's a far cry to discount consideration of serotonin or other monoamines as having roles in depression just because it seems simplistic. Your counter-argument might as well negate all mental health literature, since it's all "simplistic," has infinite confounds, too many dropouts, use depression as a construct. Why research anything at all? If you have serious studies to mention that invalidate any contribution of monoamines to mental states, then cite them. Otherwise "your opinion" isn't really adding much here. Your deconstructionist approach to all research is the same garbage that people use to say that psychosis doesn't exist.

 

[PETRAN]

No one has said it's predictive, but it's a far cry to discount consideration of serotonin or other monoamines as having roles in depression just because it seems simplistic. Your counter-argument might as well negate all mental health literature, since it's all "simplistic," has infinite confounds, too many dropouts, use depression as a construct. Why research anything at all? If you have serious studies to mention that invalidate any contribution of monoamines to mental states, then cite them. Otherwise "your opinion" isn't really adding much here. Your deconstructionist approach to all research is the same garbage that people use to say that psychosis doesn't exist.

Who said that i discount the hypothesis that "monoamines have role in depression"? You completely misinterpreted what i have said. I have specifically talked about (well not me, the majority of research) the "serotonin-level" hypothesis of depression is just wrong- that depression is the result of literally too little serotonin quantity in the synapse. I don't discount (again not me, research) that "the serotonin system" has a role in depression. It is very different to say that "neural circuits (or maybe some of the neural circuits) that use serotonin, or some of the serotonin receptors have a role in mood intensity as a mediator or last common pathway" from "depression results from too little serotonin quantity". There are some stronger studies that show some causality (e.g. the tryptophan depletion studies, interestingly it induces low moon only to those who already have a depressive episode meaning that the serotonergic system on its own is not the first cause but a mediator-or even a last pathway-in a chain of events).


My approach is far from deconstructionist, it is just scientific. And those many studies that try to pass correlations as causations is just not good science.

 

[nitemagi]

Who said that i discount the hypothesis that "monoamines have role in depression"? You completely misinterpreted what i have said. I have specifically talked about (well not me, the majority of research) the "serotonin-level" hypothesis of depression is just wrong- that depression is the result of literally too little serotonin quantity in the synapse. I don't discount (again not me, research) that "the serotonin system" has a role in depression. It is very different to say that "neural circuits (or maybe some of the neural circuits) that use serotonin, or some of the serotonin receptors have a role in mood intensity as a mediator or last common pathway" from "depression results from too little serotonin quantity". There are some stronger studies that show some causality (e.g. the tryptophan depletion studies, interestingly it induces low moon only to those who already have a depressive episode meaning that the serotonergic system on its own is not the first cause but a mediator-or even a last pathway-in a chain of events).


My approach is far from deconstructionist, it is just scientific. And those many studies that try to pass correlations as causations is just not good science.

You discounted all research on low serotonin in CSF on LP's, as if anyone was saying that finding was the same as low serotonin=cause of depression. No one implied cause, but you jumped in to refute the serotonin hypothesis which no one here was advocating.

 

[Ibid]

Your deconstructionist approach to all research is the same garbage that people use to say that psychosis doesn't exist.

I think the point is that even though a general agreement about what psychosis refers to exists, it is essentially a value judgement.

 

[nitemagi]

I think the point is that even though a general agreement about what psychosis refers to exists, it is essentially a value judgement.

As is every determination by a physician, about what qualifies as normal vs. abnormal. We judge in every little thing we do, whether you admit it to yourself or not, no matter how PC everyone wants to be.

 

[Ibid]

As is every determination by a physician, about what qualifies as normal vs. abnormal.

Yes, we will just have to agree to agree about that.
To clarify though psychosis is unique in that the context is everything. You can not ever say “behold this psychotic thought”. A psychotic thought has no intrinsic difference to an ordinary thought. You can’t say that for any other judgement you might make in practice. e.g. a malignant tumour.

 

[billypilgrim37]

To clarify though psychosis is unique in that the context is everything. You can not ever say “behold this psychotic thought”. A psychotic thought has no intrinsic difference to an ordinary thought. You can’t say that for any other judgement you might make in practice. e.g. a malignant tumour.

I can appreciate the sentiment, but while the definitions are a bit fuzzy, it's clear that people who have thoughts that we judge as psychotic have and have always had completely different prognoses than people who do not. And your last statement, re: the malignant tumor, well, it's not something somebody who knew much about medicine could really say. There's tons of diagnostic uncertainty. Before CT scans were common practice, even an appendicitis diagnosis was wrong about 15% of the time. You could never say, "behold this appendicitis" either, until you had someone knocked out and their abdomen cut open. And even with CTs, there are still appropriately performed appendectomies-turned-exlaps. Again, no "behold this appendicitis," even when the shoe otherwise fits.

Rheumatology as a field is always the best comparator for me. Though rheumatology does have a few biomarkers which are only somewhat sensitive and specific, for the most part diagnosing these very physically experienced illnesses are not that much different than what we do in psychiatry. Subjective symptoms get thrown into a set of criteria. When you look at things like a Capgras syndrome, I think it's a whole lot easier to say, "behold that psychotic thought!" than it is to say "behold this seronegative autoimmune disorder!"

Over time, folks with psychotic thoughts demonstrate substantially more neuronal loss. They tend to lose the ability to take care of themselves more rapidly than others. They have higher risk for other neurological disorders (and ones that might seem a bit more, say, objective.). There are plenty of sociologically determined things we might say about psychosis and psychotic disorders, but promoting the notion that there is nothing intrinsically different about psychotic thought and normal thought is, well, a bit oversimplified and something you have to decide to believe in for political purposes. Is it a bit of a value judgment? Perhaps, but in that regard psychosis is simply not THAT different from many other things we diagnose in medicine, but I don't think that's something that folks who haven't gone to medical school can understand easily.

And I say this in the context that you know from past posts that I definitely appreciate your perspective on mental health.

 

[Ollie123]

Why does serotonin need to have a causal relationship with depression to be diagnostically useful? I mean, I obviously agree that right now it isn't useful as a test and there is limited reason to believe it will be in the future given what we know now. However, unless these are two separate points, you seem to be implying that we need to know something plays a causal role to use it for diagnostic purposes. If we're going to limit diagnoses to criteria that have been demonstrated to play a causal role, we basically need to just shut the entire field of mental health down and stop diagnosing anything...ever. Few things in the DSM define clear causal pathways and even those are often built on shaky foundations. Most "symptoms" could be causes, consequences, or even just co-morbid conditions that we haven't yet been able to appropriately differentiate.

A clear causative pathway would certainly provide the strongest evidence, but we lack that for nearly everything in this field and even when we do find a causal relationship it typically accounts for a relatively small portion of the variance. The evidence that serotonin is in some way involved in depression seems pretty undeniable to me, though the majority of the details of that relationship are still up in the air. That certainly contributes to its inadequacy as a diagnostic test and may indicate it is too far removed to ever be a useful biomarker, but that seems quite separate from the issue of diagnostic utility. If we're throwing serotonin out because there isn't a causal relationship, we certainly need to get rid of things like the clinical interview, SCID, MMPI, and basically anything that is used to diagnose depression.

 

[fiatslug]

An interesting read on biomarkers (which I checked out from our library but did not finish 😳) is http://www.amazon.com/Endocrine-Psychiatry-Solving-Riddle-Melancholia/dp/0199737460/ref=sr_1_1?ie=UTF8&qid=1318872032&sr=8-1 Endocrine Psychiatry: Solving the Riddle of Melancholia by Edward Shorter and Max Fink. As with Shorter's other books, he weaves the history of psychiatry in with an eloquent argument for why we should be continuing to use a biomarker (the dexamethasone suppression test) in psychiatry. I need to check it out again and finish writing my notes on it!

 

[Ibid]

I can appreciate the sentiment, but while the definitions are a bit fuzzy, it's clear that people who have thoughts that we judge as psychotic have and have always had completely different prognoses than people who do not. And your last statement, re: the malignant tumor, well, it's not something somebody who knew much about medicine could really say. There's tons of diagnostic uncertainty. Before CT scans were common practice, even an appendicitis diagnosis was wrong about 15% of the time. You could never say, "behold this appendicitis" either, until you had someone knocked out and their abdomen cut open. And even with CTs, there are still appropriately performed appendectomies-turned-exlaps. Again, no "behold this appendicitis," even when the shoe otherwise fits.

Yes, it was a bit clumsy of me, I should just have said beauty is in the eye of the beholder or something like that. I'll add a health warning to my posts as soon as I can come up with a form of words san expletives.

Rheumatology as a field is always the best comparator for me. Though rheumatology does have a few biomarkers which are only somewhat sensitive and specific, for the most part diagnosing these very physically experienced illnesses are not that much different than what we do in psychiatry. Subjective symptoms get thrown into a set of criteria. When you look at things like a Capgras syndrome, I think it's a whole lot easier to say, "behold that psychotic thought!" than it is to say "behold this seronegative autoimmune disorder!"

Very elegant.

Not Capgrass but I know of a man who was detained for a month because he claimed his wife was sleeping with a man from Mars. And so it transpired, she was, probably.

Over time, folks with psychotic thoughts demonstrate substantially more neuronal loss. They tend to lose the ability to take care of themselves more rapidly than others. They have higher risk for other neurological disorders (and ones that might seem a bit more, say, objective.).

http://weblaw.usc.edu/contact/contactinfo.cfm?detailid=300

If cases like Prof Saks are not 99% down to social factors rather than biology I will eat my hat. I'll stand by with a big bottle of tomato sauce but I don't think i'll need it.

There are plenty of sociologically determined things we might say about psychosis and psychotic disorders, but promoting the notion that there is nothing intrinsically different about psychotic thought and normal thought is, well, a bit oversimplified and something you have to decide to believe in for political purposes. Is it a bit of a value judgment? Perhaps, but in that regard psychosis is simply not THAT different from many other things we diagnose in medicine, but I don't think that's something that folks who haven't gone to medical school can understand easily.

I agree with that. There is perhaps a great deal of wishful thinking about the degree of certainty with regard to the rest of medicine. The extent to which the subject matter of psychiatry is an altogether messy situation is something you are best placed to judge the size of. I'm only pointing at the gap rather than trying to fill it with rubbish. Honest.

And I say this in the context that you know from past posts that I definitely appreciate your perspective on mental health.

I can say the same with appreciation and all due respect.