This WSJ article on increasing use of antidepressants illustrates at least part of the problem: Readers naturally start thinking about patients with depressive disorders, and the article alludes to recent media attention to possible lack of effect on mild cases. Only near the end of the article does the author remind us of the wide variety of uses of these drugs beyond treatment of depressive illness, some of which enjoy FDA approval. Bupropion helps with smoking cessation. Fluoxetine gained approval for Bulimia Nervosa. I frequently prescribe mirtazapine, off label, for insomnia. FDA has approved various SSRI's for anxiety disorders like Panic Disorder and PTSD.
Did I say SSRI? Here comes another dimension. SSRI refers to a mechanism of action, or just action. SSRI's (starting with fluoxetine in the US, fluvoxamine in Europe) represented an apparent improvement over the older tricyclic antidepressants. But tricyclic, like tetracyclic (trazodone) refers to chemical structure. Other chemical classes include benzodiazepine and barbiturate.
Had enough yet? I struggle with yet another category of drug class. Even if you leave out chemical class and action, and attend to what I call clinical class, which clearly includes antidepressant, anxiolytic, and anti-psychotic, several other classes seem distinct. These include sedative-hypnotic, psycho stimulant, and neuroleptic. To my way of thinking clinical implies illness or symptom. Antidepressant means attacks depression, a symptom. But neuroleptic refers to no illness or symptom, even though we usually use that class of drugs to treat psychotic disorders. I propose calling these "effect" classes and separating them from the clinical classes. Should clinical classes be a subset of effect class or a separate class on the same hierarchical level?
Clinical classes also suffer from the too frequent assumption of all or none status. Once FDA grants approval for treatment of depression few would argue with membership of the drug in the antidepressant class. Enter the controversy surrounding the evidence that antidepressants can precipitate mania in patients with Bipolar Disorder, and take for example the anti-epileptic drug gabapentin. Anecdotal reports in the literature describe cases of apparent antidepressant effect. Should we classify the drug as an antidepressant based on such scant evidence? Does inclusion in the antidepressant class imply risk that the drug may precipitate mania in Bipolar? Just how should we determine whether a drug deserves admission to a given clinical club? For many drugs it seems the original category sticks despite evidence for inclusion in other categories.
We can see the same problem with action. We may call a drug a dopamine antagonist because that action seems to dominate, but the same drug may have histamine antagonist (anti histamine) action, and others, as well.
Sometimes the context determines the category. FDA first approved divalproex for treatment of epilepsy (Think clinical class.), but when discussed in psychiatric circles we usually classify it as a mood stabilizer (Think effect class: There's no direct mention of illness or symptom.).
Separating effect classes from clinical classes will not solve the problem. Ultimately we must maintain awareness of the limitations of the designations. The need to categorize and the complexities of the task permeate human psychology and language. For an exhaustive and fascinating exploration read: