The paper described makes a statistically rigorous attempt to recover common psychiatric diagnoses (like major depressive disorder MDD) from raw unbiased data about people’s experiences. And seems to conclude that the Venn diagram for common presentations of these conditions doesn’t line up well with the official diagnoses. Not that the conditions don’t exist. But that we are grouping together distinct conditions under the same name. They give examples of anhedonia and suicidality as distinct clusters in their data even though both would be considered MDD under the DSM.
To me this begs the question of whether the DSM authors might actually know something real and useful that the data don’t show. For example that anhedonia often progresses to suicidality even though the two might not coexist in the same person often. (Doesn’t sound right to me, but that’s how I read the article.) I think it’s plausible that the direct implication of this research is actually wrong. The obvious conclusion is that the DSM is full of it and doesn’t match real people’s experiences. But I suspect it might be that the DSM captures useful correlations and progressions that this method didn’t collect. Perhaps because the data here are from a single point in time, not a progression.
To me the conclusion is that the DSM is out of date. Not the document, or specific iteration itself but the mindset driving it. When DSM was simpler and we didn’t know as much about mental illness, then what DSM was trying to achieve wasn’t such a tall order, and it was a useful diagnostic companion. However it seems to me it’s gone from being merely useful to being considered canon, and through successive iterations has tried to fill those boots to the point where people are calling out the DSM as being entirely flawed when in fact what is going on is it’s just stepping out of its bounds.
Even still, wouldn't youvwant to separate X & Y are the same thing from X eventually progresses to Y.
I think the most obvious conclusion is not that the DSM is bunk, but that some of things it groups together might be better to group separately. The real question of course is which grouping corresponds better to the group responding the same way to a similar intervention. After all the point diagnosing is to put people into groups so we can learn from the patterns of similar people to figure out better treatments.
Fair point. I agree. If there’s structure in people, the DSM should capture it. MDD phase 1,2,3 would be a useful update.