Tools for Practice

#228 The depressing evidence for antidepressants in the elderly

How effective are antidepressants for treating depression in the elderly?

The efficacy of antidepressants in the elderly is inconsistent and may decrease as patients age. From 80% to 40% of elderly patients will recover with antidepressants, with some studies showing no difference from placebo response rates. Harms of antidepressants are common, with ~20% stopping due to adverse effects.  

CFPCLearn Logo

Reading Tools for Practice Article can earn you MainPro+ Credits

Join Now

Already a CFPCLearn Member? Log in

5 recent systematic reviews of randomized, placebo-controlled trials (RCTs) used different ages for inclusion:  
  • All antidepressants, mean age ~70, followed mostly 4 weeks (range 3-20), statistically different rates of “recovery” (achieving a set change in or reaching a predetermined depression score):1 
    • Tricyclic Antidepressants (TCAs) (10 RCTs): 75% versus 51% (placebo), Number Needed to Treat (NNT)=5. 
    • Selective Serotonin Reuptake Inhibitors (SSRIs) (2 RCTs): 83% versus 72% (placebo), NNT=10. 
  • SSRIs and newer antidepressants only: 10 RCTs, mean ages 68-80, followed 6-12 weeks, statistically different rates of:2 
    • Response (>50% improvement in symptoms): 44% versus 35% (placebo), results inconsistent. 
    • “Remission”: 33% versus 27% (placebo), results inconsistent. 
  • Any antidepressant: 15 RCTs, mean follow-up ~7 weeks:3 
    • Response (>50% improvement in symptoms): decreased with age: 
      • 54% for mean age 44, 42% for mean age 73. 
      • Placebo response rates similar regardless of age (~33-39%).  
        • Post-hoc analysis: no difference from placebo when limited to studies over age 65. 
  • SSRIs only: 12 RCTs, mean ages 70-79, followed for mostly 8 weeks:4 
    • Response or remission: no difference compared to placebo. 
  • Limitations: Often based on secondary analysis. 
  • Likely no difference in efficacy between TCAs and SSRIs, but adverse-effect withdrawals higher with TCAs (24% versus 17%).5 
  • Elderly patients may respond to antidepressants slower than adults, possibly requiring 10-12 weeks before effects seen.2 
  • Chronic illness often co-exists with depression in elderly patients, along with frailty, possibly mitigating effects.6 
  • Cognitive Behavioural Therapy has been inconsistently shown to improve depression symptoms in the elderly.7,8 
  • In the elderly, antidepressants have been associated with a similar fall risk as benzodiazepines.9  
  • TFAntidepressants may not be effective in treating depression in dementia.10 

Latest Tools for Practice

#348 How to Slow the Flow III: Tranexamic acid for heavy menstrual bleeding (Free)

In premenopausal heavy menstrual bleeding due to benign etiology, does tranexamic acid (TXA) improve patient outcomes?
Read 0.25 credits available

#347 Chlorthali-D’OH!: What is the best thiazide diuretic for hypertension?

Which thiazide diuretic is best at reducing cardiovascular events in hypertension?
Read 0.25 credits available

#346 Stress Urinary Incontinence: Pelvic floor exercises or pessary? (Free)

How effective are pelvic floor exercises or pessaries for stress urinary incontinence?
Read 0.25 credits available

This content is certified for MainPro+ Credits, log in to access


  • Adrienne J Lindblad BSP ACPR PharmD
  • Shan Lu MD CCFP

1. Wilson K, Mottram PG, Sivananthan A, et al. Cochrane Database System Rev. 2001; 1:CD000561.

2. Nelson JC, Delucchi K, Schneider LS. Am J Geriatr Psychiatry. 2008 Jul; 16 (7): 558-67.

3. Tedeschini E, Levkovitz Y, Iovieno N, et al. J Clin Psychiatry. 2011 Dec;72(12):1660-8.

4. Tham A, Jonsson U, Andersson G, et al. J Affect Disord. 2013. 205:1-12.

5. Mottram PG, Wilson K, Strobl JJ. Cochrane Database System Rev. 2006;1:CD003491.

6. Vaughan L, Corbin AL, Goveas JS. Clin Interv Aging. 2015. 10:1947-58.

7. Wilson K, Mottram PG, Vassilas C, et al. Cochrane Database System Rev. 2008; 1:CD004853.

8. Jonsson U, Bertilsson G, Allard P, et al. PlosOne. 2016; 11(8):e0160859.

9. Woolcott JC, Richardson KJ, Wiens MO, et al. Arch Intern Med. 2009;169:1952-60.

10. Dudas R, Malouf R, McCleery J, et al. Cochrane Database System Rev. 2018;8:CD003944.

Authors do not have any conflicts of interest to declare.