Nicely done, Awais! The points you make dovetail very neatly with the piece I did for Psychiatric Times, in which I discuss the nature of the placebo condition; what the "15%" figure in the Stone et al study really means for clinicians; and why a narrow fixation on numbers can obscure the main reason psychiatric medications are worthwhile; i.e., because they reduce, even if modestly, the suffering and incapacity of extremely debilitating and sometimes lethal illnesses.
I think there is an inconsistency in this argument. It was stated that the maintenance efficacy data is severely confounded. However, I believe the justification for this, primarily withdrawal effects, also casts doubt on the RCTs that were referenced in support of antidepressant efficacy.
Stone et al. (2022)¹ stated that one of the limitations of their paper was that "the effects of treatment history of antidepressants or discontinuations before study entry on our results are unknown". They cited Hunter et al. (2015)², who found that previous exposure to antidepressants caused a greater separation between drug and placebo in a subsequent trial.
The vast majority of trials do not exclude people who have had previous exposure to antidepressants. This means that any drug-placebo difference could be due to this and may not be a true treatment effect that would manifest in people who are drug-naive.
While withdrawal effects from antidepressants are poorly understood, there are many patient reports of symptoms persisting for months and sometimes years after discontinuation. Presumably lasting changes have occurred in these people, which is the mechanism behind these persistent symptoms. These changes may alter their response to antidepressants at a later date. It's also possible that those in the placebo arm may have a decreased response, due to experiencing withdrawal from prior treatment.
With this in mind, I think it's difficult to argue that the rest of the evidence base is less confounded than the maintenance trials. Unless a large RCT is conducted on purely drug-naive people, we can't know what is a true drug effect and what is confounded by prior antidepressant use.
¹ Stone M B, Yaseen Z S, Miller B J, Richardville K, Kalaria S N, Kirsch I et al. Response to acute monotherapy for major depressive disorder in randomized, placebo controlled trials submitted to the US Food and Drug Administration: individual participant data analysis
² Hunter AM, Cook IA, Tartter M, Sharma SK, Disse GD, Leuchter AF. Antidepressant treatment history and drug-placebo separation in a placebo-controlled trial in major depressive disorder.
Nicely done, Awais! The points you make dovetail very neatly with the piece I did for Psychiatric Times, in which I discuss the nature of the placebo condition; what the "15%" figure in the Stone et al study really means for clinicians; and why a narrow fixation on numbers can obscure the main reason psychiatric medications are worthwhile; i.e., because they reduce, even if modestly, the suffering and incapacity of extremely debilitating and sometimes lethal illnesses.
https://www.psychiatrictimes.com/view/antidepressants-placebos-and-lithium-some-parting-thoughts
All the best,
Ron
Ronald W. Pies, MD
Professor Emeritus of Psychiatry
SUNY Upstate Medical University
I think there is an inconsistency in this argument. It was stated that the maintenance efficacy data is severely confounded. However, I believe the justification for this, primarily withdrawal effects, also casts doubt on the RCTs that were referenced in support of antidepressant efficacy.
Stone et al. (2022)¹ stated that one of the limitations of their paper was that "the effects of treatment history of antidepressants or discontinuations before study entry on our results are unknown". They cited Hunter et al. (2015)², who found that previous exposure to antidepressants caused a greater separation between drug and placebo in a subsequent trial.
The vast majority of trials do not exclude people who have had previous exposure to antidepressants. This means that any drug-placebo difference could be due to this and may not be a true treatment effect that would manifest in people who are drug-naive.
While withdrawal effects from antidepressants are poorly understood, there are many patient reports of symptoms persisting for months and sometimes years after discontinuation. Presumably lasting changes have occurred in these people, which is the mechanism behind these persistent symptoms. These changes may alter their response to antidepressants at a later date. It's also possible that those in the placebo arm may have a decreased response, due to experiencing withdrawal from prior treatment.
With this in mind, I think it's difficult to argue that the rest of the evidence base is less confounded than the maintenance trials. Unless a large RCT is conducted on purely drug-naive people, we can't know what is a true drug effect and what is confounded by prior antidepressant use.
¹ Stone M B, Yaseen Z S, Miller B J, Richardville K, Kalaria S N, Kirsch I et al. Response to acute monotherapy for major depressive disorder in randomized, placebo controlled trials submitted to the US Food and Drug Administration: individual participant data analysis
² Hunter AM, Cook IA, Tartter M, Sharma SK, Disse GD, Leuchter AF. Antidepressant treatment history and drug-placebo separation in a placebo-controlled trial in major depressive disorder.