Blog Editor


  • Copyright 2005-20012 by Adam Kolber
    All rights reserved.

« Singer on Ashley's "Pillow Angel" Treatment | Main | On the limits of neuroscience and the value of "magical thinking" (Connors) »


Better biological tests for PTSD will involve stress hormone levels. I suppose, though, we could get relatively good data looking at the amygdala. In either case, you would need to expose the person to the trigger, which would differ from person to person. It's interesting how we have this hierarchy: brain response is really real, hormone level is pretty real, skin conductance is less real still, and simply asking the person (probably the single most reliable test) is not real at all.

I don't know, maybe it could be one of biological makers (hormones could be great as Neil said) but I think it could be a waste of money. It will cost to much for the results. I don't know how much clincical screening costs, but at 400$ a hours for fMRI, I'm almost sure that fMRI is more expensive !

Thanks for the comments, Neil and Ben!

Neil, I'm equally fascinated by the way that brain imaging has taken on the mantel of medical objectivity - by the idea that an enhanced amygdala response is, as you say, "more real" proof of PTSD than is an individual's description of his or her own experience. I'm doing a research work at LSE on how this perception of brain science objectivity has developed... Marmar's comments to the Department of Defense certainly help maintain it!

As to the possibilities of amygdala scanning, the idea of using fMRI data to prove psychiatric state is particularly suspect given its renowned variability. A bad scanner day (i.e., the magnet needed to be calibrated, the technician aligned the subject incorrectly, the experimental software crashed in the middle of the experiment), a bad subject day (the soldier was tired, drunk, fell asleep, was moving, sneezed, etc), problems processing the data, and problems interpreting the data all seem to be sources of significant error. The fact that, as you point, out, the experimental stimuli might have to be hand-tailored to each soldier also seriously complicatesthe practical conduct and interpretation of these tests. I would love to see a full report of exactly how Marmar handled these issues in his testimony - which, unfortunately, probably won't be out for a while.

Like Ben, I'm also wondering how the appropriations subcommittee will respond to the question of cost. Don't forget someone will also have to be paid to process the imaging data - which takes time and skill!

A few more questions to throw out for discussion:

--How will these data be interpreted to make a diagnosis - and who gets to make the call?

--How much amygdala response is too much? Where does the interpreter draw the line between normal, heightened, and pathological response?

--Does it matter how the individual responded pre-trauma - and if so, is it a problem that there is no "before" picture to compare?

--What happens if a subjectively ill individual has a null imaging result? If no clinical interview is undertaken, what happens to soldiers who don't show a typical (if there is such a thing, and I'd have serious questions that there is) PTSD brain response?


You may or may not be interested to know that I am likely headed towards a dissertation on pain and pain management in which one of my aims is to evaluate claims that fMRI will reveal (for the first time) "objective" evidence of pain and therefore transform the way we conceive of pain at all.

Short answer: I'm exceedingly dubious.

The comments to this entry are closed.