History

To our knowledge, the term discontinuation syndrome, with respect to cessation of an SSRI, was first used in a single article published in 1993.^{Reference Black, Wesner and Gabel26} However, it appears next only in 1997, with the publication of a supplement issue in the Journal of Clinical Psychiatry. ^{Reference Kaplan19–Reference Schatzberg, Haddad, Kaplan, Lejoyeux, Rosenbaum and Young22,Reference Haddad27–Reference Schatzberg, Haddad, Kaplan, Lejoyeux, Rosenbaum and Young31} This supplement issue was based on a closed symposium titled ‘SSRI Discontinuation Events’; both the symposium and the supplement were sponsored by Eli Lilly.^{Reference Schatzberg, Haddad, Kaplan, Lejoyeux, Rosenbaum and Young31} Immediately after this supplement, the term became more frequently used in the literature (Fig. 1). The Journal of Clinical Psychiatry published another supplement issue in 2006 that was based on a 2004 meeting funded by Wyeth Pharmaceuticals.^{Reference Shelton, Blier, Tremblay, Fava, Delgado and Schatzberg32} In this supplement, authors advocated for replacing the term withdrawal with antidepressant discontinuation syndrome in the literature as the former was regarded as ‘particularly confusing to patients, many of whom are already concerned about becoming addicted to antidepressants’.^{Reference Shelton23} According to one of the articles, ‘the distinctions between discontinuation symptoms and drug withdrawal are clear. Thus, the use of proper terminology when discussing this phenomenon with patients will help to alleviate concerns and stop the spread of common misperceptions’.^{Reference Shelton23}

SSRI, selective serotonin reuptake inhibitor.

Fig. 1 Number of publications per year from 1993 to 2018 that mention the term ‘discontinuation syndrome’ in relation to SSRI antidepressants. Articles retrieved from Medline, EMBASE and PsychInfo, using the search terms ‘antidepressant’ AND ‘discontinuation syndrome’. The search retrieved 183 publications once duplicates and irrelevant results (i.e. results not related to SSRI cessation) were removed.

Defining antidepressant discontinuation syndrome

As the term antidepressant discontinuation syndrome came into use, there were various attempts to define it and distinguish it from withdrawal. In an article from the Journal of Clinical Psychiatry 1997 supplement, Schatzberg described five ‘hallmark features’ of discontinuation syndrome as proposed by a panel of experts: ‘It is not attributable to other causes; It is emergent upon abrupt discontinuation, frequent nonadherence (missed doses), and, less often, after dose reduction; It is generally mild and short-lived but can be distressing; It can be reversed by the reintroduction of the original medication or one that is pharmacologically similar; It is minimized by a slow taper or by using a drug that has an extended half-life’.^{Reference Schatzberg21} There are several limitations in this description of discontinuation syndrome. First, none of the above criteria serve to distinguish discontinuation syndrome from withdrawal syndrome. Second, there is much evidence that adverse effects following antidepressant cessation are often neither ‘mild’ nor ‘short lived’.^{Reference Fava, Gatti, Belaise, Guidi and Offidani11–Reference Davies and Read13,Reference Read, Cartwright and Gibson33} From a systematic literature review on withdrawal symptoms following discontinuation of SSRIs and benzodiazepines, Nielsen et al concluded that adverse effects following SSRI cessation include potentially debilitating symptoms that resemble those associated with benzodiazepines.^{Reference Nielsen, Hansen and Gøtzsche12} In a more recent systematic review, the authors found that symptoms following discontinuation of antidepressants occur at a weighted average rate of 56%, and are described as severe in 46% of patients experiencing them.^{Reference Davies and Read13} Lastly, although the severity of withdrawal symptoms can presumably be ‘minimized by a slow taper or by using a drug that has an extended half-life’,^{Reference Schatzberg21} they may still be a significant challenge despite such strategies.^{Reference Fava, Gatti, Belaise, Guidi and Offidani11,Reference Wilson and Lader34}

In 1998, Rosenbaum et al developed the Discontinuation Emergent Signs and Symptoms checklist, entailing 43 items to assess for discontinuation syndrome.^{Reference Rosenbaum, Fava, Hoog, Ascroft and Krebs35} Two years later, Black et al proposed a new set of diagnostic criteria for antidepressant discontinuation syndrome after examining 46 case reports published from 1986 to 1997 inclusive.^{Reference Black, Shea, Dursun and Kutcher36} These criteria include two or more of the following symptoms developing within 1–7 days of discontinuation or reduction in dose of an SSRI after at least 1 month's use: dizziness, light-headedness, vertigo or feeling faint; shock-like sensations or paraesthesia; anxiety; diarrhoea; fatigue; gait instability; headache; insomnia; irritability; nausea or emesis; tremor and visual disturbances.^{Reference Black, Shea, Dursun and Kutcher36} However, in the case of antidepressants with a longer half-life, such as fluoxetine, such symptoms may not be present within 7 days of discontinuation. The Journal of Clinical Psychiatry 2006 supplement included an article describing an alternative system, characterising SSRI discontinuation syndrome ‘according to the following syndromic features: neurosensory (e.g., vertigo, paresthesias, shock-like reactions, myalgia, other neuralgia); neuromotor (e.g., tremor, myoclonus, ataxia, visual changes); gastrointestinal (e.g., nausea, vomiting, diarrhea, anorexia); neuropsychiatric (e.g., anxiety, depressed mood, intensification of suicidal ideation, irritability, impulsiveness); vasomotor (e.g., diaphoresis, flushing); and other neurologic (e.g., insomnia, vivid dreaming, asthenia/fatigue, chills)’.^{Reference Shelton23} None of these various proposed criteria justify the distinction between discontinuation syndrome and withdrawal. Instead, the symptoms described are typical of patients experiencing withdrawal and indicate the various ways in which patients can become dependent upon their SSRIs, thus making it difficult to reduce their use of antidepressants.

Distinguishing discontinuation syndrome from withdrawal

A common argument for the use of the term discontinuation syndrome is that it must be distinguished from withdrawal since it is claimed that antidepressants, such as SSRIs, do not cause dependence. One author has argued that substance dependence is distinct from antidepressant discontinuation in the DSM-IV since the diagnostic criteria for substance dependence, including a loss of control, excessive time using the drug or development of tolerance toward the drug ‘rarely occur in patients taking antidepressants’.^{Reference Haddad37} However, the DSM-IV – and, for that matter, the DSM-5 – includes further criteria for dependence that are inconsistent with this argument. The substance dependence criteria also include withdrawal, as manifested by either the characteristic withdrawal syndrome (outlined elsewhere in the DSM-IV), or the same substance being taken to relieve or avoid withdrawal symptoms. By this definition, antidepressants can be considered to cause dependence since patients do experience characteristic withdrawal symptoms upon discontinuation, and may continue to take their antidepressants to avoid withdrawal symptoms.^{Reference Read, Gee, Diggle and Butler9,Reference Stockmann, Odegbaro, Timimi and Moncrieff10} Furthermore, it must be stressed that, although SSRI cessation may not involve addictive behaviours such as excessive time using the drug (as often associated with opiate or alcohol dependence), the evidence for significant symptoms associated with cessation of SSRI drugs is consistent with dependence.

There have been several studies refuting the use of the term discontinuation syndrome in place of withdrawal, in relation to SSRIs. A recent article by Baldessarini and Tondo described reactions following reduction of antidepressants, particularly SSRIs, as being similar to the ‘withdrawal-like syndromes [that] occur with most sedatives and antianxiety agents, as well as opioids, alcohol, and other psychoactive agents’.^{Reference Baldessarini and Tondo38} Furthermore, after evaluating the literature on benzodiazepines and SSRIs, Nielsen et al ^{Reference Nielsen, Hansen and Gøtzsche12} found that 37 out of 42 withdrawal symptoms were described for both benzodiazepines and SSRIs, thus indicating that the withdrawal reactions of SSRIs are comparable with those of benzodiazepines. They concluded that it ‘makes no sense’ to describe only benzodiazepines as being associated with dependence. Using these findings alongside other studies included in a systematic review, Fava et al ^{Reference Fava, Gatti, Belaise, Guidi and Offidani11} concluded that using the term withdrawal syndrome to refer to the symptoms caused by benzodiazepine, antipsychotics and tricyclic antidepressants while using the term discontinuation syndrome for those pertaining to SSRIs has ‘minimized the potential vulnerabilities induced by SSRI and has provided the ground for misleading indications’. Similarly, the authors of a recent systematic review on the topic argue against use of the term discontinuation syndrome: it can be misleading since relevant symptoms can occur without discontinuing the drug (e.g. with dose reduction), and, ‘the term “syndrome” subtly medicalises withdrawal by associating it with a disorder endogenous to the person than with a non-dysfunctional reaction to the cessation of a drug’.^{Reference Davies and Read13} In addition to these arguments in the literature, several medical organisations, including the Royal College of Psychiatrists³⁹ and British Medical Association,⁴⁰ have released statements in which the term discontinuation syndrome to describe symptoms following SSRI cessation has been abandoned in favour of the term withdrawal. The Royal College of Psychiatrists’ recent statement on antidepressants and depression has called for a ‘greater recognition of the potential in some people for severe and long-lasting withdrawal symptoms on and after stopping antidepressants in NICE guidelines and patient information’,³⁹ a contention that is well supported in the literature.^{Reference Davies, Read, Hengartner, Cosci, Fava and Chouinard41}