Antidepressants of Limited Benefit in Dementia Patients

Clinical Context

The prevalence of depression in patients with dementia is more than 20%, as reported by Ballard and colleagues in the June 1996 issue of the International Journal of Geriatric Psychiatry. The first-line treatment is often a selective serotonin reuptake inhibitor or a noradrenergic and specific serotonergic antidepressant. However, a 2002 Cochrane review by Bains and colleagues found that the evidence of the effectiveness of antidepressants in dementia was weak.

This multicenter, parallel-group, double-blind, placebo-controlled, randomized trial by Banerjee and colleagues assesses the efficacy and safety of sertraline, a selective serotonin reuptake inhibitor, and mirtazapine, a noradrenergic and specific serotonergic antidepressant, compared with placebo for the treatment of depression in patients with dementia.

Study Synopsis and Perspective

Antidepressants may provide little benefit but appear to increase adverse effects in individuals with dementia and comorbid depression, suggesting that clinicians should reconsider use of these agents in this population, new research suggests.

In a randomized, controlled trial of more than 300 elderly patients with probable or possible Alzheimer’s disease (AD), those who were treated with sertraline or mirtazapine showed no significant difference in depression reduction after 3 months compared with the placebo group. In addition, individuals receiving active treatment reported more moderate and severe treatment-associated adverse events.

“I am surprised by just how unequivocal our findings are,” lead author Sube Banerjee, MD, professor of mental health and aging at King’s College London, Institute of Psychiatry, United Kingdom, told Medscape Medical News.

On the basis of these results, the investigators recommend clinicians “reframe” the way they treat these patients and “reconsider routine prescription of antidepressants.”

“The present practice of use of these antidepressants with usual care for first-line treatment of depression in Alzheimer’s disease should be reconsidered,” they write.

The study was published online July 18 in The Lancet.

Commonly Prescribed Agents

According to the study, dementia affects 35 million people worldwide and costs “1% of global gross domestic product every year.” There is also a more than 20% prevalence of depression in these patients.

“Depression is one of the most important comorbidities in dementia.  It is a source of great distress yet the treatments we use are not proven,” said Dr. Banerjee.

The study’s aim was to assess efficacy and safety of “2 of the most commonly prescribed drugs,” explain the researchers.

The Health Technology Assessment Study of the Use of Antidepressants for Depression in Dementia (HTA-SADD) included 326 elderly patients (mean age, 79 years) between January 2007 and December 2009 from 9 centers providing old-age psychiatry services in England.

All participants were diagnosed with comorbid depression for 4 or more weeks and had a score of 8 or more on the Cornell scale for depression in dementia (CSDD). Participants were randomly assigned to receive 150 mg of sertraline (n = 107, 68% female), 45 mg of mirtazapine (n = 108, 71% female), or placebo (n = 111, 64% female) daily.

Reduction in CSDD score at 13 weeks was the primary outcome measure. CSDD score and adverse events at 39 weeks were among the secondary outcomes.

No Between-Group Benefits

Results showed that although the CSDD scores decreased for all groups at 13 weeks, there were no significant differences between those receiving sertraline or mirtazapine compared with those receiving placebo (mean difference, 1.17 and 0.01, respectively) or compared with each other (mean difference, 1.16).

There were also no significant between-group differences in CSDD scores at the 39-week check-point.

In addition, significantly fewer patients receiving placebo reported adverse reactions compared with those receiving sertraline (26% vs 43%; P = .010) or mirtazapine (41%; P = .031).

Gastrointestinal reactions, especially nausea, were the most common events reported by those receiving sertraline, and drowsiness and sedation were the most commonly reported by those receiving mirtazapine.

Compared with both treatment groups, the placebo group also had significantly less severe adverse events (P = .003).

A total of 15 participants died by the 39th week, 5 in each treatment group.

“The message is to think before using antidepressants for depression in dementia.  It may well be that these symptoms will resolve with the problem-solving and information-giving that is implicit in good-quality dementia care,” added Dr. Banerjee.

Furthermore, the investigators suggest that antidepressants be reserved for “individuals whose depression has not resolved within 3 months of referral, apart from those in whom drug treatment is indicated by risk or extreme severity.”

Largest Trial to Date

In an accompanying editorial, Henry Brodaty, MD, from the Brain and Aging Research Program and Primary Dementia Collaborative Research Center at the University of New South Wales in Sydney, Australia, notes this study is the “largest trial of antidepressant drugs in dementia ever and almost equalling the combined total of previously published work.”

Although negative, he adds, the findings are important because depression in dementia is common and results from previous trials looking at this issue are “unconvincing.”

“The absence of efficacy of antidepressants in the [study’s] dementia population raises questions about whether there are different pathogenic mechanisms at play in depression in Alzheimer’s disease,” writes Dr. Brodaty.

He points out that the study recommends “a stepped care approach of watchful waiting, followed-up by low-intensity psychosocial interventions and, if unsuccessful, by more complex and intense interventions.”

Dr. Brodaty also stresses that the trial does not advocate the abandonment of antidepressants in these patients.

Rather, he notes, the trial data underscore “the need for clinicians to think about creative alternatives to drug treatment for management of depression in people with dementia, and to use evidence-based techniques and partnerships with family carers.”

The study was funded by the United Kingdom National Institute of Health Research HTA Program. The study authors report several disclosures, which are listed in full in the original article. Dr. Brodaty has disclosed no relevant financial relationships.

Related Link
The Alzheimer’s Society provides a variety of online resources on recognition, diagnosis and management of depression, including a large number of downloadable fact sheets helpful for patient education.

Study Highlights


  • 326 patients enrolled from 9 old-age psychiatry services in the United Kingdom were randomly selected to 1 of 3 groups.
  • Inclusion criteria were probable or possible AD, coexisting depression for at least 4 weeks that might need antidepressant treatment, and a CSDD score of 8 or higher.
  • Exclusion criteria were patients’ condition clinically too critical, absolute contraindication to study drugs, prior antidepressant use, enrollment in another trial, or no caregiver.
  • 111 patients received placebo.
  • 107 received sertraline 50 mg, up to a target total dose of 150 mg per day.
  • 108 received mirtazapine 15 mg, up to a target total dose of 45 mg per day.
  • Doses were increased from 2 to 3 doses daily based on CSDD scores at 4 and 8 weeks or based on clinician recommendations thereafter.
  • Mean sertraline dose was 70 mg, and mean mirtazapine dose was 24 mg.
  • All patients received other treatment as usual.
  • Groups were similar in age (79 – 80 years), sex (29% – 36% men), ethnicity (92% – 94% white), married status (43% – 56%), residence in a care home (12% – 18%), duration and severity of depression, caregiver characteristics, and participant quality of life.
  • The primary outcome measure was reduction of depression based on the CSDD score at 13 weeks.
  • Analysis adjusted for baseline CSDD score, time, and treatment center.
  • Depression scores improved in all 3 groups from baseline to 13 weeks and 39 weeks, with the greatest absolute improvement in CSDD score of −5.6 for the placebo group at 13 weeks.
  • Depression score mean differences at 13 weeks did not differ between the placebo vs the sertraline groups (1.17; P = .10), placebo vs mirtazapine groups (0.01; P = .99), or the mirtazapine vs the sertraline groups (1.16; P = .11).
  • Depression score differences at 39 weeks did not differ among the placebo, sertraline, and mirtazapine groups.
  • Fewer neuropsychiatric symptoms, higher caregiver-rated participant scores for health-related quality of life, and higher caregiver quality-of-life scores occurred in the mirtazapine vs the sertraline group at 13 weeks, but not at 39 weeks.
  • Higher caregiver quality-of-life and mental health scores occurred in the placebo vs the sertraline group at 13 weeks, but not at 39 weeks.
  • Adverse reactions were less common in the placebo group (29 [26%] of 111) vs the sertraline group (46 [43%] of 107; P = .010) or the mirtazapine group (44 [41%] of 108; P = .031).
  • The most common reactions were gastrointestinal tract symptoms in the sertraline group and psychological reactions in the mirtazapine group.
  • Serious adverse events were similar in the placebo vs the sertraline vs the mirtazapine groups (15 vs 12 vs 14), but these effects were rated as severe in only 3 (20%) of patients in the placebo group vs 8 (67%) of patients in the sertraline group and 10 (71%) of patients in the mirtazapine group.
  • Mortality rate was similar for all groups: 5 patients in each group had died by 39 weeks.
  • Withdrawal rates by 39 weeks were 24% in the placebo group vs 35% in the sertraline group vs 29% in the mirtazapine group.
  • Loss to follow-up was greater in the sertraline and mirtazapine groups vs placebo at 13 weeks but was similar by 39 weeks.
  • Study limitations were possible bias from loss to follow-up, revised target sample size, measurement error from cognitive impairment, and lack of data on intervention used by various teams.

Clinical Implications


  • In patients with depression and dementia, sertraline or mirtazapine vs placebo is not more effective in alleviating depressive symptoms at 13 weeks and at 39 weeks.
  • In patients with depression and dementia, adverse events are less common with placebo use vs sertraline use and mirtazapine use. Sertraline use most commonly causes gastrointestinal tract symptoms, and mirtazapine causes psychological reactions.