Stopping antidepressants

It is a positive step when a patient is ready to stop taking antidepressants, but as Dr Cosmo Hallström warns, this should not happen abruptly

Withdrawal reactions on stopping antidepressants have been recognised for many years¹. More recently, this issue has reached public awareness following an article called 'the antidepressant web' written by Charles Medawar². The concern arises out of the natural reluctance of patients to take medication for what is seen as a psychological problem and the past spectre of addiction to tranquillisers and is a reaction against the rising tide of antidepressant prescribing.

Symptoms

The main characteristics of the antidepressant withdrawal syndrome are listed in Table 1. The main features are gastrointestinal symptoms such as nausea and vomiting, flu-like symptoms such as myalgia, fatigue and chills and dizziness and vertigo. Patients also complain of sleep disturbance and vivid dreams, sensory disturbances such as electric shocks and movement disorders. Other symptoms have also been reported³. The syndrome described on withdrawal from selective serotonin reuptake inhibitors (SSRIs) and other new antidepressants are broadly similar to that on discontinuing tricyclics, although movement disorders such as dystonia and paradoxical activation such as mania, appears to be a more specific anticholinergic withdrawal syndrome and is associated mainly with tricyclic antidepressants.

The mechanism of these withdrawal reactions would appear to be cholinergic rebound, at least in part. Even the SSRI paroxetine (Seroxat), has a strong affinity for muscarinic receptors. Other mechanisms may include a functional deficiency of serotonin but other neurotransmitter systems may also be involved, such as dopamine and opioid receptors. Withdrawal reactions seem to be more common with short half-life SSRIs such as paroxetine, sertraline (Lustral) and venlafaxine (Efexor), rather than the longer acting ones such as fluoxetine (Prozac). Also, withdrawal reactions are less likely if the drug is tapered off over a few days rather than being stopped abruptly. Withdrawal reactions are uncommon if the drug has been taken for less than eight weeks. There have even been case reports of withdrawal reactions in new-born children of mothers taking antidepressants.

These withdrawal reactions are distinct from the underlying depression or other condition being treated. They are not like the benzodiazepine or alcohol withdrawal reactions. Nor are they associated with addictive type syndromes such as drug seeking behaviour or dose escalation and are not associated with 'addiction'. They are simply a rebound phenomenon.

These reactions also need to be distinguished from relapses or recurrences of the underlying depressive illness which may occur on discontinuation of the antidepressants⁴. Because of this risk, there is often a case for staying on antidepressant medication for some months after treating the acute episode, and in some patients with frequent relapses in whom the condition is disabling if untreated, long-term treatment may be advisable.

Conclusion

Withdrawal reactions to antidepressants do occur. They are distinct from an addiction and dependence phenomena and represent a pharmacological rebound. They are generally mild and short-lived, usually lasting for two or three days, rarely a week or more. Reactions can be minimised by tapering off the antidepressants over a week or two, as would be good practice when discontinuing all long-term medication. If withdrawal reactions are troublesome, they can be stopped by restarting the antidepressants and discontinuing them more slowly. They probably occur with all antidepressants but the true incidence is not really known, although they appear to be more common with the highly potent short-acting compounds and less so with longer half-life compounds.

As with all therapeutic decisions, the risks of this rare and minor side-effect of treatment needs to be balanced against the true benefits of treatment with antidepressants; a clinical judgement for the clinician. Patients and doctors need to be aware of this and other potential problems of the treatment. This minor problem of possible withdrawal reactions should not prevent doctors from treating depressive illnesses vigorously (with all available therapies), since depression and allied conditions remain common, cause considerable morbidity and are often highly treatable.

Cosmo Hallström is Consultant Psychiatrist and Medical Director, Charter Clinic Chelsea

References
1 Current problems in pharmacovigilance. The Medicines Control Agency/Committee on Safety of Medicines, 1993
2 Medawar C. The Antidepressant Web. Int J Risk and Safety in Medicine. 1997; 10: 75-126
3 Schatzberg et al. Antidepressant discontinuation syndrome. J Clin Psych. 1997; 589 (suppl): 3-27
4 Edwards G. Long-term pharmacotherapy of depression. Br Med J 1998; 316: 1180-1