Stopping antidepressants

Introduction

By 2020, major depression is projected to become second only to heart disease as the leading cause of morbidity.¹ The lifetime risk of depressive illness is 12-26% for women and 4-12% for men.² It is usually a chronic and recurrent illness (75-80% of treated patients have recurrences) that frequently requires long-term maintenance treatment. Depression is often both unrecognised and under treated. The aim of treatment is a full remission and long-term recovery rather than short-term response.

When is the right time to stop treatment?

The decision to stop treatment should be made after an assessment of the patient's current mood state and other factors that may indicate the likelihood of a relapse or recurrence of depression. These factors include the number and severity of previous episodes, success of treatment of earlier episodes, the risk of suicide if another episode were to ensue and the disruption caused by depression to the lives of the patient and their family. Discussion of these factors with the patient (and a key family member, if possible) would be essential in coming to a decision about discontinuing therapy.³

How to stop treatment

If antidepressants are withdrawn, higher doses should be gradually tapered off, unless there are medical indications for an abrupt cessation. These indications could include pregnancy, severe adverse reactions or inability to take oral medications. Precise guidelines concerning the time needed to taper off the dose are lacking. A gradual reduction is recommended to prevent discontinuation effects and to allow adaptation at the receptor level. A rule of thumb is 6-8 weeks after 6-8 months treatment³ or 3-6 months after maintenance therapy. Many patients, particularly those on lower doses, may be able to stop more quickly without adverse effects.

If the response to treatment has been unsatisfactory, a switch to a different antidepressant may be necessary. The prescriber should check the product information to see if the two drugs interact, but reducing the dose over a 1-2 week period may be adequate. Patients taking high doses of an antidepressant or who are on an antidepressant with a shorter half-life (e.g. paroxetine and venlafaxine) are more likely to develop discontinuation symptoms during short taper periods. Patients must be educated about the importance of supervised dose reduction when discontinuing antidepressants and about what symptoms they may experience. They can be reassured that these symptoms will remit with time.

Monitoring

Education is a critical aspect of treatment and enhances compliance with medication. The patient and their family should be informed that adverse effects are common, but are usually mild and resolve on continued treatment, and that the depression is likely to recur if treatment is stopped too soon.⁴ Other educational messages that are associated with better compliance include advice to take the medication every day and to continue even when feeling better.

Patients must be warned that as depression is typically a recurring disorder, stopping medication is always a trial and may lead to symptoms reappearing. They also need to be instructed to contact their doctor as soon as any symptoms start to recur. The assistance of a key family member can play a crucial role in this respect. The doctor may need to continue to monitor patients periodically after medication has been stopped.

Problems associated with discontinuation

Discontinuation reactions may have physical or psychological symptoms, which appear after stopping or reducing the dose of medication. The symptoms may start within 1-10 days, but usually within three days of stopping treatment. These are distinct from the symptoms of depression, which can also recur within hours to days after cessation of treatment. However, recurrences are less likely than discontinuation reactions to occur in the first week after stopping treatment. Discontinuation reactions are more common in patients who have been treated for more than eight weeks and with higher dosages of antidepressants. Discontinuation symptoms must also be distinguished from an intercurrent illness. They are often overlooked in the acute hospital setting.

Discontinuation symptoms from abrupt cessation of tricyclic antidepressants (TCAs) (see Table 1) and monoamine oxidase inhibitors (MAOIs) have long been recognised, but features of addiction such as tolerance and addictive use are rare.⁵ Gastrointestinal effects, flu-like symptoms, affective symptoms and sleep disturbance are the most common problems after stopping a TCA. Discontinuation effects are also common after withdrawal of MAOIs and include disorientation, confusion, myoclonus, ataxia, agitation, cognitive impairment, catatonia, paranoid delusions, aggressiveness, hallucinations, depression, suicidality, slowed speech and sleep disturbance.

The commonest cessation effects of SSRIs are dizziness, light-headedness, nausea, lethargy and headache (see Table 2). Distinguishing a discontinuation syndrome from a recurring depression can be difficult (see 'Medicinal mishaps: serotonin states' Aust Prescr 1998;21:63). The cessation effects of SSRIs are generally less frequent than those of the TCAs. Reports vary from 33% for clomipramine to 80% for amitriptyline, while the rate is 35% for paroxetine and much less (2-14%) for other SSRIs.⁶ The commonest withdrawal symptoms are also different for each class of antidepressant. Two symptoms which are prominent after stopping an SSRI are balance and sensory abnormalities. These do not occur after a TCA is stopped.

There are few reports in the literature about the cessation of newer antidepressants. Stopping venlafaxine has resulted in symptoms of dizziness, light-headedness, irritability, dysphoria, insomnia and sweating.⁶

Effects when changing treatment

If a patient is not responding to an antidepressant, or relapses, a different drug may be necessary. The effects of discontinuing the first drug may not appear until after starting the new drug. It is important not to confuse the symptoms of discontinuation with the adverse effects of the new drug. If time permits, it is helpful if the patient has 3-4 days off medication before starting the new drug. This allows discontinuation symptoms to be identified and distinguished from new adverse events. Advice for changing from one antidepressant to another is published in Therapeutic Guidelines: Psychotropic.⁷

Management of discontinuation reactions

Problems occurring on cessation of an antidepressant may be minimised by preventative measures, supportive treatment and, if necessary, specific treatment. Preventative measures include emphasising the need for a supervised reduction of the dosage, advising the patient of the risk of discontinuation reactions and warning of possible symptoms that may occur. If possible, avoid high doses and abrupt cessation of medication.

Usually supportive treatment is sufficient. The patient should be reassured that symptoms are not life-threatening and that they will resolve spontaneously within 1-2 weeks. If symptoms are severe, resuming therapy may be necessary. The discontinuation syndrome will then typically resolve within 24 hours or so. A slower reduction of the dose may minimise cessation reactions the next time withdrawal is attempted.

Conclusion

Depressive illness is now recognised as a major health problem. Recent guidelines recommend long-term maintenance treatment for patients with recurrent depression. Higher doses and longer treatment periods may lead to the more frequent occurrence of discontinuation reactions in future. Approximately one in three patients do not respond to the first antidepressant they are prescribed and are switched to another. This changeover period is a risk time for discontinuation reactions as well as drug interactions.

Clinicians need to be aware of the symptoms associated with discontinuation of antidepressants and inform their patients what to expect. Abrupt cessation of antidepressants should be avoided unless medically necessary and gradually tapering off the dosage should be the norm.