Antidepressants in the elderly

Clinical question

How effective are antidepressants for treating depression in the elderly?

Bottom line

Response to antidepressants in the elderly varies widely from 45% to 80%, compared with 35% to 70% for placebo, with inconsistencies in part owing to use of secondary analysis, variable age cutoffs, and short trials. Efficacy might decrease with patient age. Harms of antidepressants are common, with approximately 20% of patients stopping owing to adverse effects.

Evidence

Five recent systematic reviews of RCTs used different ages for inclusion.

• For all antidepressants¹ (mean age about 70, mostly followed for 4 [range 3 to 20] weeks), there were statistically different rates of “recovery” (achieving a set change in or reaching a predetermined depression score).

  • -Tricyclic antidepressants (TCAs) in 10 RCTs: 75% versus 51% (placebo), number needed to treat of 5.

  • -Selective serotonin reuptake inhibitors (SSRIs) in 2 RCTs: 83% versus 72% (placebo), number needed to treat of 10.

• For SSRIs and newer antidepressants only² (10 RCTs, mean ages 68 to 80, followed for 6 to 12 weeks), there were statistically different rates of response, defined as a 50% or more improvement in symptoms, (44% vs 35% [placebo]) and “remission” (33% vs 27%), although results were inconsistent.

• For any antidepressant³ (15 RCTs, mean follow-up of about 7 weeks), response (> 50% improvement in symptoms) decreased with age: 54% for mean age 44, 45% for mean age 70, 42% for mean age 73.

  • -Placebo response rates were similar regardless of age (about 33% to 39%).

  • -Post hoc analysis found no difference from placebo when limited to studies of those older than 65.

• For SSRIs only⁴ (12 RCTs, mean ages 70 to 79, followed for mostly 8 weeks), there was no difference in remission or response versus placebo.

Context

• There is likely no difference in efficacy between TCAs and SSRIs, but withdrawals from treatment owing to adverse effects are higher with TCAs (24% vs 17%).⁵

• Elderly patients might respond to antidepressants more slowly than younger adults, possibly requiring 10 to 12 weeks before effects are seen.²

• Chronic illness often coexists with depression in elderly patients, along with frailty, possibly mitigating effects.⁶

• In the elderly, antidepressants have been associated with a fall risk similar to that with benzodiazepines.⁷

• Antidepressants might not be effective in treating depression in dementia.⁸

Implementation

Late-life depression affects 15% to 20% of people older than 65 and is associated with reduced quality of life and function, and increased risk of death.⁹ When starting antidepressants, setting goals can help determine response in a practical way (eg, improved socialization, sleep, anxiety, appetite, energy). Medical conditions and their symptoms should be optimized, pain treated, and strategies employed to improve socialization. In patients with dementia, apathy is common (50% to 90%) and can often be misconstrued as depression.¹⁰ Interventions such as cognitive-behavioural therapy and exercise have been inconsistently shown to improve depressive symptoms.^11,12