The ever-valuable Neuroskeptic, channeling Stanley Kowalski ("I knew a girl once said she was the glamorous type. She said to me, 'I am the glamourous type.' I said, 'So what?'"), asks just WTH it means to show that brain scans of earthquake survivors show that "trauma alters brain function."
The authors link their findings to previous work with frankly vague statements such as "The increased regional activity and reduced functional connectivity in frontolimbic and striatal regions occurred in areas known to be important for emotion processing". But anatomically speaking, most of the brain is either "fronto-striatal" or "limbic," and almost everywhere is involved in "emotion processing" in one way or another. So I don't think we understand the brain much better for reading this paper.
Further work, building on these results, might give insights. We might, say, learn that decreased connectivity between Regions X and Y is because trauma decreases serotonin levels, which prevents signals being communicated between these areas, which is why trauma victims can't use X to deliberately stop recalling traumatic memories, which is what Y does.I just made that up. But that's a theory which could be tested. Much of today's neuroimaging research doesn't involve testable theories - it is merely the exploratory search for neural differences between two groups. Neuroimaging technology is powerful, and more advanced techniques are always being developed. What with resting state, functional connectivity, pattern-classification analysis, and other fancy methods, the scope for finding differences between groups is enormous and growing. I'm being rather unfair in criticizing this paper; there are hundreds like it. I picked this one because it was published last week in a good journal.
Another way of putting it, sort of: It's well-established by now that experience changes the brain. So big experiences would be expected to leave substantial or at least noticeable changes in function, which would show up in fMRI. Simply remarking that fact, and showing it in pictures, adds ... not so much. Yet as Neuroskeptic notes, scores, even hundreds (thousands?) of fMRI studies do only that and little more. They show us the brain changed, but not what the changes amount to.
Of course dramatic experiences change us (and our brains). Hell, college changes your brain, and so does playing tennis or Tetris or studying the violin or spending a summer sailing or hanging sheetrock. We'd expect living through earthquakes or war or a divorce would as well. But simply to show brain changes from experiences doesn't mean the changes have made you dysfunctional. This is one of too many studies that simply shows the change, without any real insight (or only the fuzziest insight, so vague as to be meaningless) into what the change amounts to. It gets printed because it has to do with trauma -- but i gives no real insight into the nature or extent of trauma's effect.
Catch it at Neuroskeptic
- Log in to post comments
Uh oh, someone finally named the white elephant in the room.
I'm only mildly annoyed with the current observational approach to neuroscientific correlations. We have to wander over the land for quite a while before we get a decent map. It seems somewhat akin to Emil Kraeplin's longitudinal, observational studies which led to a useful discrimination between schizophrenia and bipolar disorder.
The main issue in hypothesis testing with constrained variables is that so many goddam things change in the brain that it's hard to articulate what variable is actually being constrained, which is being measured, and what the confounder is.
It's disheartening to think that an experimental subject merely being annoyed because of getting a parking ticket right before coming to the scanner will heavily influence the findings, but so it goes when you study that damn brain.