The International Antidepressant Withdrawal Crisis: Time to Act

January 30, 2019

This Letter to the Editor is in response to the article published in Psychiatric Times, “Sorting Out the Antidepressant ‘Withdrawal’ Controversy,” by Ronald W. Pies, MD and David N. Osser, MD.


Editor’s note: This Letter is in response to the article published in Psychiatric Times, “Sorting Out the Antidepressant ‘Withdrawal’ Controversy,” by Ronald W. Pies, MD and David N. Osser, MD. What follows this letter is a rejoinder by Drs Pies and Osser.


Dr Read is Professor at the University of East London and a member of the International Institute for Psychiatric Drug Withdrawal, Sweden. Dr Davies is Associate Professor at the University of Roehampton and a member of the All Party Parliamentary Group for Prescribed Drug Dependence, London.

We are pleased to see that eminent psychiatrists in the US are beginning to acknowledge and discuss the difficulties millions of people around the world are having when they try to withdraw from antidepressants. However, overall, we fear that rather than “Sorting Out the Antidepressant Withdrawal Controversy,” Drs Pies and Osser have made use of imaginary case-studies, appeals to clinical experience, a biased reading of our recent systematic review,1 and a selective use of literature, in order to try and reassure professionals that antidepressant withdrawal is minimal and easily manageable-a view that is clearly inconsistent with an evidence-based approach to this issue.

For example, while the clinical experience of just two people (Drs Pies and Osser) is considered valid data for determining what is and is not “common,” the personal experiences of thousands of people who have actually tried to withdraw from antidepressants is characterized as “anecdotal” and “extreme.” In our opinion, when clinicians start from the false presumption that in their clincial experinece a problem is rare, this can become a self-fulfilling prophecy that minimizes the problem in perpetuity. Let us remember, it was the “clinical experience” of most psychiatrsts in the 1960s and 1970s that benzodiazapines were not addcitive, which of course turned out to be wrong. Furthermore, dismissing the lived experience of thousands of people in the online layperson withdrawal community does not exactly endear the profession to those they purport to help, and does not take us closer to solutions to the problems they are describing.

With such diametrically opposed experiences between professionals and patients, we must, as always, turn to the research. Here again, we believe Drs Pies and Osser demonstrate bias. We are grateful to them for reporting that our own recent systematic review found that an average of 56% of people experience antidepressant withdrawal symptoms when trying to come off the drugs, and that about 46% of these people rate their symptoms as severe. Drs Pies and Osser correctly state that a commentary by the British National Health Service (NHS)2pointed out that some of the studies used to calculate the percentage of people experiencing withdrawal symptoms utilized online surveys, and that “Online surveys are prone to selection bias, as people are more likely to respond to a survey if they have experienced a problem than if they haven’t. This means the results [of the Davies and Read study1] may overestimate the proportion of people who experience antidepressant withdrawal.”

What the authors fail to report is that the very next sentence of the NHS commentary2 reported that “And some of the studies followed unusually short trials of antidepressants (for example, 8 weeks or 12 weeks), whereas most people are prescribed the drugs for at least 6 months. Short treatment trials might underestimate difficulties seen withdrawing from longer-term treatment.” In addition, Drs Pies and Osser do not inform readers that our review1 directly addressed the possible bias of the online surveys by pointing out that in the two largest surveys 83% and 65% of the participants reported that the antidepressants had helped them, so the samples were, if anything, biased towards people with a positive view of the drugs rather than a negative one.

It may be useful to look at the three types of study our review included to see that, when grouped, they did not differ greatly in terms of withdrawal incidence. The weighted averages are as follows:

• The three online surveys – 57.1% (1790/3137)

• The five naturalistic studies – 52.5% (127/242)

• The six short randomised controlled trials – 50.7% (341/673)

Reaching similar findings from different methodologies is typically seen to strengthen confidence in an overall estimate. In fact, findings from the three methodology types demonstrate that it is broadly safe to conclude that at least half of people suffer withdrawal symptoms when trying to come off antidepressants.

Drs Pies and Osser seem keen to promote use of the term “discontinuation syndrome” rather than refer to withdrawal, as we indicated in our study. The definition of “discontinuation syndrome” that is currently in use emerged from the “Discontinuation Consensus Panel” funded by Eli Lilly in 1996,3 which, for commercial reasons, erroneously separated antidepressant withdrawal from other CNS drug withdrawals.4 We agree with Fava and colleagues,5 who noted in 2015 that the term “discontinuation syndrome” minimizes the vulnerabilities induced by SSRI and should be replaced by “withdrawal syndrome” or, in our view, “withdrawal reaction” or “symptom.”

We do appreciate Pies and Osser acknowledging that “many clinicians-including, unfortunately, some psychiatrists-have underestimated the potential severity and duration of antidepressant discontinuation/withdrawal syndromes.” This is an important first step.

About 37 million in the US are prescribed antidepressants in any given month (about 13% of the adult population) and half of those have been taking them for at least 5 years.6 We now know for certain that millions of people in the US and beyond struggle when they try to come off these drugs. Underestimating the problem is not going to help patients get the accurate information, and the withdrawal support services, they need and deserve.


1. Davies J, Read J. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects: Are guidelines evidence-based?Addict Behav. 2018 Sep 4. [Epub ahead of print].

2. Calls for guidelines to be revised over antidepressant withdrawal symptoms. NHS. October 3, 2018. Accessed January 29, 2019.

3. Schatzberg AF, Haddad P, Kaplan EM, et al. Serotonin reuptake inhibitor discontinuation syndrome: A hypothetical definition. Discontinuation Consensus panel. J Clin Psychiatry. 1997;58(Suppl 7):5-10.

4. Nielsen M, Hansen EH, Gotzsche PC. What is the difference between dependence and withdrawal reactions? A comparison of benzodiazepines and selective serotonin re-uptake inhibitors. Addiction. 2012;107:900–908.

5. Fava GA, Gatti A, Belaise C, et al. Withdrawal symptoms after selective serotonin reuptake inhibitors discontinuation: A systematic review. Psychother Psychosom. 2015;84:72-81.

6. Mojtabai R1, Olfson M. National trends in long-term use of antidepressant medications: Results from the U.S. National Health and Nutrition Examination Survey. J Clin Psychiatry. 2014;75:169-77.

Scroll down for a rejoinder by Drs Pies and Osser >

Response to Davies & Read

We appreciate Prof Read and Dr Davies taking the time to respond to our commentary. We would like to take this opportunity to address some of the misapprehensions contained in the letter from Drs Read and Davies.

First: our piece was, as Drs Read and Davies note, an “opinion piece,” labeled as “Commentary.” It was not our aim to provide a comprehensive review of the literature, nor did we intend an extensive analysis or critique of the Davies and Read paper we cited. Had we intended the latter, we would have cited the comprehensive critique of the Davies and Read paper, provided by two UK psychiatrists, Hayes and Jauhar;1 the rejoinder by Drs Davies and Read;2 and the rejoinders to the Davies/Read rejoinder from Drs Hayes and Jauhar [see comments, ref. 1].

Readers of Psychiatric Times who delve into this contentious debate will see that the issues are complex, and the resolution, uncertain. It was simply beyond the scope of our piece to address these in detail. Rather, our piece was intended to acknowledge that there is a controversy as regards antidepressant withdrawal; to lend our perspective to the issue, based on a selective review of the literature and our combined seven decades of clinical experience with at least a thousand depressed patients; and to provide some clinical guidance to psychiatrists about the optimal duration of antidepressant discontinuation.

To some of the specific claims raised by Drs Read and Davies:

1. The case vignettes we presented were hardly “imaginary.” They were composite cases based on many actual patients we have either treated, consulted on, or encountered in our clinical practice. Composite cases are a well-accepted teaching device in academic medicine.

2. We do not accept the false equivalence Davies & Read proffer; ie, between careful clinical observation and the subjective reports of “thousands of people who have actually tried to withdraw from antidepressants.” There is a radical difference between subjective reports contained, eg, in online surveys; and the careful, week-by-week, month-by-month, clinical observation of patients by psychiatric specialists. Such observation is indeed an “evidence-based approach,” though the nature of the evidence differs from that obtained in surveys or randomized studies.

That said, we do not claim that our perspective is completely free of bias, or that it represents a clear consensus among our colleagues. Nevertheless, we are not prepared to discount the thorough assessments of our patients that we have made over the decades and conclude that somehow, we must have missed all these severe withdrawal cases that others are claiming to be so prevalent.  Nor do we consider it appropriate to advocate that clinicians stop using antidepressants, as some are claiming should be the appropriate response to these reports. 

3. We do not dismiss “…the lived experience of thousands of people in the online layperson withdrawal community.” On the contrary, we explicitly stated,

“. . . we do not deny that severe reactions can and do occur when antidepressants are stopped suddenly (or the dose reduced too rapidly), [but] we also believe that fears of such “excruciating” experiences are greatly overstated, in the context of proper psychiatric care.”

The latter qualification is critical to understanding why our perspective differs from that of Read and Davies, who provide no evidence in their paper that the subjects in either the online surveys or the naturalistic studies received “proper psychiatric care.” Given that the vast majority of patients prescribed antidepressants in both the US and UK are not treated by psychiatrists, it seems unlikely that this was the case.

4. We acknowledge that short trials of antidepressants “might underestimate difficulties withdrawing from longer-term treatment” (as per the UK National Health Service statement3), and Read and Davies are correct to point this out, with respect to their study. We regret the paucity of longer-term studies of antidepressant use and discontinuation and believe this must be remedied urgently.

5. Re: the issue of “bias” in online surveys, Read and Davies argue that because most survey participants in their study reported that the antidepressants had “helped” them, this somehow indicates that, if anything, the survey samples were “biased towards people with a positive view of the drugs, rather than a negative one.” This reasoning is flatly fallacious. The mere fact that patients reported that, at some point in their treatment, the antidepressant “helped” them does not negate the potential bias pointed out by the NHS; namely, “Online surveys are prone to selection bias, as people are more likely to respond to a survey if they have experienced a problem than if they haven’t. This means the results [of the Davies & Read study] may overestimate the proportion of people who experience antidepressant withdrawal.”,3 italics added]

6. Drs Read and Davies argue, based on their review, that “. . . it is broadly safe to conclude that at least half of people suffer withdrawal symptoms when trying to come off antidepressants . . . ” Though this might be the case, we do not believe that their review validates this claim. First of all, the bulk of the Davies/Read data were drawn from online surveys. These can, at most, provide information on what patients retrospectivelyreport to survey takers. Such data are not able to establish the actual incidence (or prevalence) of the phenomenon in question, which would require (1) a validated set of uniform diagnostic criteria for “withdrawal”; and (2) careful-ideally, contemporaneous-clinical observation to confirm that subjects actually meet the specified criteria. Unfortunately, we do not yet have a uniform set of diagnostic criteria for the phenomenon in question; rather, a group of very heterogeneous symptoms are now lumped together, under the rubric of “antidepressant withdrawal.”

To be clear: we do not discount the reported experiences of survey participants, many of whom almost certainly experienced severe discomfort in the course of drug tapering. We are merely pointing out that the epidemiological determination of incidence and prevalence require different scientific standards than surveys can provide.

7. Drs Read and Davies claim, in their letter, that we are “very keen to promote use of the term ‘discontinuation syndrome’ rather than refer to withdrawal.” We reject that characterization of our commentary. At least seven times in our commentary, we use the composite term “discontinuation/withdrawal” in order to signal that the matter of labeling this phenomenon is contested and controversial. Nowhere in the piece do we “promote” the term “discontinuation syndrome,” though we note that this is the term used in the DSM-5, and in the review by Jha and colleagues.4

8. Even if we stipulate that the figures Drs Davies and Read provide are accurate from an epidemiological standpoint-ie, that about 1 in 4 people treated with antidepressants experience serious withdrawal symptoms-their data do not address, nor do they refute, our main contention; namely, that “. . . serious [antidepressant] withdrawal symptoms are extremely rare when tapering periods of 2 to 6 months are used.” And here is another way of viewing the Davies and Read data: the overwhelming majority of patients-about 74%-do not experience serious withdrawal effects when antidepressants are discontinued.

We suspect that Prof Read and Dr Davies would agree with us that much more research is needed to test and perhaps confirm this hypothesis; eg, randomized, controlled studies of antidepressant discontinuation periods of 2 to 4 weeks; 4 to 8 weeks; 8 to 12 weeks, etc. Unfortunately, at present, we do not have such controlled data. In the meantime, we can all agree that physicians must be much more vigilant in suspecting, detecting, and managing distressing reactions to discontinuation of antidepressants. The Davies and Read paper rightly reinforces this point. That said, we stand by our main conclusion:

“. . . when managed appropriately, discontinuation of antidepressants need not pose a significant clinical problem, and should not discourage depressed patients from using these beneficial medications.”

Ronald W. Pies, MD
David N. Osser, MD

1. Hayes J, Jauhar S. Antidepressant withdrawal: reviewing the paper behind the headlines. The Mental Elf. October 18, 2018. Accessed January 29, 2019.
2. Antidepressant withdrawal review: authors respond in detail to Mental Elf critique. May 11, 2018. Council for Evidence Based Psychiatry. Accessed January 29, 2019.
3. Calls for guidelines to be revised over antidepressant withdrawal symptoms. NHS. October 3, 2018. Accessed January 29, 2019.
4. Jha MK, Rush AJ, Trivedi MH. When Discontinuing SSRI Antidepressants Is a Challenge: Management Tips. Am J Psychiatry. 2018;175:1176-1184.