This article has a correction

Please see: Evid Based Med 2013;18:120

Evid Based Med 18:27-28 doi:10.1136/eb-2012-100735
  • Therapeutics
  • Randomised controlled trial

A small proportion of people with dementia and neuropsychiatric symptoms experience clinically significant worsening when antidepressants are discontinued

  1. Anne Corbett2
  1. 1King's College London, London, UK
  2. 2Alzheimer's Society, London, UK
  1. Correspondence to: Clive Ballard
    Wolfson Centre for Age-Related Diseases, Wolfson Wing, Hodgkin Building, King's College London, Guy's Campus, London SE 1 1UL, UK; clive.ballard{at}

Commentary on: Google Scholar


Depression is common in dementia with prevalence over 20%,1 causing distress, reducing quality of life, exacerbating impairment, increasing mortality and increasing caregiver stress.2 Treating depression is a clinical priority to improve well-being, quality of life and function in people with dementia. Depression may also underpin other neuropsychiatric symptoms such as agitation.3 The overall evidence does not, however, suggest that antidepressants are an effective treatment for depression in Alzheimer's disease (AD),4–6 although there is preliminary evidence for potential benefits of some antidepressants in the treatment of agitation.


Bergh and colleagues conducted an important and well-designed 25-week double-blind placebo-controlled withdrawal trial in 128 people with AD, vascular dementia or mixed AD/vascular dementia taking the antidepressants escitalopram, citalopram, sertraline or paroxetine. Patients with a diagnosis of a concurrent depressive disorder were excluded, although 20–25% of participants had Cornell Depression Scale Scores within the range usually considered to represent at least minor depression.


Over the 25 weeks, the group continuing to take the antidepressants had a 1.6 point significant advantage on the Cornell Depression Scale compared with the placebo group, although 86% of people remained below the threshold for significant depression. There was also a modest non-significant advantage for the antidepressant-treated group with respect to overall neuropsychiatric symptoms. A higher proportion of people in the discontinuation group withdrew because of the worsening of neuropsychiatric symptoms (21% vs 6%), and there were no other significant differences in adverse events.


The results are difficult to interpret for several reasons. Firstly, participants had different types of dementia, which may have responded differently to treatment. In addition, although reasonable for this type of study and population, there was a moderately high drop-out rate (37%). It is also unclear what proportion of people had significant depression at baseline. The main conclusion is, however, that only a small proportion of people experienced a clinically significant worsening in depression or overall neuropsychiatric symptoms when their antidepressant was discontinued.

For clinical practice, the results reinforce that only a small proportion of people with dementia without a clear history of depressive disorder will experience a worsening of depression or neuropsychiatric symptoms if antidepressants are discontinued. The results also highlight important research questions pertaining to the potential value of antidepressants in the overall treatment of agitation and in vascular dementia.

The evidence base for the treatment of depression in dementia was summarised in a Cochrane review which concluded that despite its clinical seriousness, there was only weak evidence for the effectiveness of antidepressants in dementia. Subsequently, there have been two larger randomised controlled trials. The DIADS-II study compared 67 people with AD prescribed sertraline with 64 people given placebo, finding no benefit of sertraline at 12 or 24 weeks, even in the sub-group with more severe depression.5 A more recent study in the United Kingdom, HTA SADD, randomised 326 patients with clinically significant depression in the context of AD to sertraline, mirtrazapine or placebo in equal ratios for 9 months. Neither antidepressant was significantly more efficacious than placebo, and sertraline was associated with a significant increase in adverse events.6

In the present study, the modest benefits and small proportion of people developing clinically significant depression when antidepressants were discontinued do not provide sufficient evidence to justify the long-term use of antidepressants in people with dementia. The main clinical interpretation should, therefore, be that it is clinically safe to discontinue long-term antidepressants in the majority of people with dementia, although clinicians should monitor for worsening depressive symptoms.


  • Competing interests None.


Free Sample

This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of EBM.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.