Extrapyramidal symptoms occasionally have been reported over the last twenty-five years in association with antidepressant monotherapy, and although the correspondents of that anecdotage have implied a causal association, the latter has not been established, not only because it is the exception (by far) bestriding the rule, but because the foundational indication for which antidepressants were approved is itself associated with an extrapyramidal syndrome (Parkinsons disease). its prodromes, and its plausible formes fruste. Antidepressants doubtless are and have been prescribed for depressive symptoms accompanying or presaging fully developed, classic Parkinsons disease, and thereby may antidepressants bottom down on the wrong seat in a game of etiological chairs.
Attributing causation to something by virtue of its association with a “third thing” (tertium quid) associated with one or more other things that also are is an error of logic designated by early Medieval scholastics a tertium quid artifact. Not essentially a confirmation bias, tertium quid artifactivity, like confirmation bias, beclouds other associations that, if considered, might promote an increment of circumspection with just enough inertia to contain an erroneous causal leap. Nettlesome, tertium quid artifactivity, because it can deprive patients of best treatments or make them wait longer for them; nettlesome, too, for researchers snagged on red herrings.
The conclusions of a study by Canadian clinicians (M.Y. Guo and colleagues, Journal of Clinical Psychopharmacology 38:4:349-356, 2018) reminds us that one question need not be confined to one answer.
The authors did a case-control study wherein “cases” were patients, whose last visits billed at a Toronto hospital were coded for Parkinson’s disease or extrapyramidal adverse effects associated with Parkinson’s disease; controls were a larger group treated in the same hospital whose last billing codes did not. Both groups received the same range of treatments with exceptions pertaining to drugs reported to induce extrapyramidal adverse effects, drugs used to mitigate them, and drugs used for treatment of Parkinsons disease. (Accordingly excluded were drugs that included anticholinergics (benztropine, trihexyphenidyl, diphenhydramine), dopaminergic agonists (ropinirole and pramipexole), first-, second-, and subsequent-generation antidepressants, and typical and atypical neuroleptics.) The authors then selected their cases from patients who had a last billing code of parkinsons disease or extrapyramidal symptoms consistent therewith, and their controls from those who did not.
Then calculating incidence rates for receiving antidepressants in th groups, the authors found an association between antidepressant exposure and Parkinsons disease or extrapyramidal symptoms. Antidepressants appeared to differ in the strengths of their. Among the cases, treatment with (and incidence ratios with respect to neurologically normal controls) were duloxetine (5.68), mirtazapine (3.28), citalopram (3.47), escitalopram (3.23), paroxetine (3.07), sertraline (2.57). venlafaxine (2.37), bupropion (2.31), and fluoxetine (2.03).
The authors consider their findings to represent a “harmful association” between antidepressants and Parkinsons disease and/or extrapyramidal symptoms. (Obiter: Yet, as mentioned (supra), while extrapyramidal adverse effects have sporadically been reported in antidepressant-treated patients, a relationship has been documented between Parkinsons disease and depression, and between depression and treatment with antidepressants. More likely to represent a causal relationship in the authors’ study might be that between depression and antidepressants by way of association between Parkinsons disease, its prodromes and formes frustes, and depression.)