I often go on rants about how terrible clinical diagnosis is, and worse that we create descriptions of conditions based on clinical diagnosis. The jist of those rants is that clinical diagnosis, no matter how many structured surveys we create, are ultimately subjective and that subjectivity obviates any chance at reliability.
In the first article below, researchers found that a condition (progressive supranuclear palsy) quantified by specific pathology is being under diagnosed on a kind of jaw dropping scale. For context, the work estimates the incidence rate of PSP to be an order of magnitude higher than Parkinson’s.
More shockingly, nearly ALL incidence of Parkinson’s related symptoms coincide with PSP pathology. Think for a moment about how much effort we’ve dumped down the hole on Parkinson’s research.
And despite the futility in “treating” the condition, despite the weak, heterogeneous etiologies, despite still describing these conditions in terms of “probability”, the consensus thought still isn’t questioning whether or not the subjective clinical diagnostic criteria could possibly be flawed.
This work will be replicated consistently with all “types” of dementia. Alzheimer’s is just as bullshit as Parkinson’s. To be clear I’m not asserting that the pathology which drives the symptoms of these conditions don’t exist, I’m asserting the subjective descriptions established by individuals who had no clue how nervous system pathology plays out might be the barrier to resolving the issue.
The second article is very similar, after two decades of ever widening etiological nets regarding “autism” forcing the introduction of the completely absurd “spectrum” conceit, we are finally starting to see work which illustrates that there are indeed discrete phenotypical groups rather than “blend”.
The article clearly points out just how different the assumed etiology of these conditions must be, describing them as “subtypes”, without ever taking the actual leap of questioning what actually makes them “subtypes”. Without the assumption that they were researching the “causes” of “autism”, I can’t imagine any argument would be taken seriously asserting these “subtypes” were related.
This is so much of neuroscience right now, we keep throwing good effort after bad effort based on these ideas of function which give us nothing but futility with regard to actually addressing the “conditions”. Many ideas born before we even knew what DNA was, let alone how molecular interactions drive behavior. There is still a tremendous mindshare which believes memory is made out of magic ephemeral electricity or something, and that human brains for some reason mechanically process the world differently than the rest of the natural world.
Every single time there’s a big paradigm shift in a field of science, we always wonder how much further along we would be if we weren’t so enamored by that old, silly construct which through the advantage of hindsight was completely ridiculous. We wonder how so many people could have believed such silly concepts, despite all the obvious contrary evidence.
Well, this is how. Right now, at this moment. Our cognitive dissonance is perfectly clear and obvious. For whatever reason, whether it be desire for consensus, or nostalgia, or just fear, every day face the futility of our beliefs and cling to them anyway.
The remodeled brain will become obvious through it’s efficacy and consistency, and the sooner we get there, the better the outcome for us all.
Edit: I really underplayed how insane the first study is.
Imagine there’s a pathology out there which is more common than Alzheimer’s, Parkinson’s, Fronto-Temporal Dementia, and Vascular Dementia combined.
We don’t actually research this pathology much because we are really terrible at clinically diagnosing it.
So terrible in fact, that it’s diagnosed less than 10% of the time it probably exists.
On the obverse, nearly every single time it is actually diagnosed it turns out to be accurate. There are almost no false positives!
This compared to the clinical dementias, which still are so unreliable that we describe them in terms of probability. There are lots of false positives!
Even worse, the etiologies which we think drive those clinical dementias are worthless without clinical history, as half the people with those etiologies do no present with dementias (e.g. Asymptomatic Alzheimer’s).
It’s the parable of the blind men and the elephant playing out right in front of us, if we’d only take the time to look.
Molecular and network-level mechanisms explaining individual differences in autism spectrum disorder
Rainwater Charitable Foundation criteria for the neuropathologic diagnosis of progressive supranuclear palsy – Christ, this just… grrr.
The criteria showed high sensitivity (0.97) and specificity (0.91), as well as almost perfect inter-rater reliability for diagnosing PSP and differentiating it from other tauopathies (Fleiss kappa 0.826).
Like, HOLY SHIT. In the relevant work above, 86% of the time this pathology presented with dementia (quoted from above the first study):
Of these, 87 were termed clinicopathological PSP as they also had either dementia or parkinsonism or both.
I’ve been arguing that most dementias have a shared etiological basis, but didn’t expect it to be this damn resounding.
More than anything, you know what absolutely blows me away?
Most cases (17/25) had 100% agreement across all 14 raters.
PERFECT AGREEMENT AMONG A LARGE POOL OF RATERS. Even with cherry picked cases, that’s BONKERS. For reference, most psychiatric work, which actually drives the overwhelming majority of neuroscience research, considers 70% GREAT inter-rater with only TWO raters. Imagine something like “Fetal Alcohol Syndrome”, where even with cherry picked cases (like all of them are pre-existing diagnoses of FASD), interrater reliability is less than 20%.
This criteria may still be inaccurately measuring the etiological basis for both PSP and dementias as a whole, but it’s utility is in a completely different league than current diagnostic criteria.