Intended for healthcare professionals


New antidepressants for old people?

BMJ 1999; 318 doi: (Published 19 June 1999) Cite this as: BMJ 1999;318:1640

The evidence that newer drugs are much better than the old is thin

  1. Martin G Livingston, Senior lecturer in psychological medicine (mgl2w{at},
  2. Hilary M Livingston, Consultant psychiatrist
  1. Southern General Hospital, Glasgow G 51 4TF

    Depression is common in elderly people. Old people are at greater risk of developing adverse events while taking any medication—including antidepressants—because of concurrent illness, consumption of other prescribed and over the counter drugs, dose miscalculation because of forgetfulness, and altered drug kinetics.1 The efficacy and safety of antidepressants in elderly people is therefore an important treatment issue. In all age groups the use of tricyclic antidepressants for depression has been declining in favour of serotonin reuptake inhibitors.2 Although views differ on whether serotonin reuptake inhibitors should be used as first line treatment in depression,3 the drop out rates from treatment are similar for both classes of drug—about a third in each group.4 What is the evidence for efficacy and safety of serotonin reuptake inhibitors and other newer classes of antidepressants drugs inelderly depressed patients?

    Newer, non-tricyclic antidepressants are often claimed to be as effective as but safer than tricyclic antidepressants and hence drugs of choice in elderly people. Such drugs include the selective serotonin reuptake inhibitors (SSRIs) citalopram, fluoxetine, fluvoxamine, sertraline, and paroxetine; reversible inhibitors of monoamine oxidase A (RIMAs) such as moclobemide; selective serotonin and noradrenaline reuptake inhibitors (SNRIs) such as milnacipran and venlafaxine; the combined 5HT2 antagonist and 5HT reuptake inhibitor nefazodone; mirtazapine, which antagonisesα2presynaptic receptors and blocks 5HT2 and 5HT3 receptors;and the noradrenaline reuptake inhibitor (NARI) reboxetine. The wide range of actions of these drugs on the central nervous system shows that a coherent theory of the biochemical basis of depression continues to elude us.

    It is claimed that fluoxetine,5 citalopram,6 and moclobemide7 are more effective than placebo in trials involving subjects aged 65 and over who are defined as depressed using either the ELDRS score8 or DSM-IIIR criteria.9 In virtually all comparator trials of antidepressants both of the drug treatments show similar efficacy. As might be expected, fluvoxamine,10 milnacipran,11 paroxetine,12 sertraline,13 and venlafaxine14 are said to show similar efficacy when compared with a tricyclic antidepressant in the treatment of DSM-IIIR defined depression.

    All the trials cited excluded subjects who had physical diseases, such as cardiac, renal, and hepatic disorders and prostatism, common in people aged over 64. No trials comment on the sequelae of overdoses, and little information exists on drug-drug interactions, a crucial omission for this population. Fluoxetine is the only newer antidepressant that has been evaluated clinically in depressed patients with organic brain disorder,5 which is important because depressive symptoms accompany dementia in 19% of cases.15 Drop out rates in the trials comparing older antidepressants with new ones are available for sertraline versus amitriptyline (48% v 49%),13 paroxetine versus amitriptyline (21% v 34%),12 venlafaxine versus dothiepin (20% v 15%),14 fluvoxamine versus dothiepin (35% v 27%),10 and milnacipran versus imipramine (46% v 37%).11 Only one trial gives comparative drop out rates for a newer antidepressant compared with placebo in the over 64s exclusively—citalopram versus placebo (39% v 33%).6 The high drop out rate on placebo probably reflects the high level of somatic complaints among elderly people with depression. None of the available trials would have sufficient power to detect a 20% difference in efficacy between old and new antidepressants (assuming 80% power, α=0.05, then n required would be 788). For placebo trials, assuming 80% power to detect a 50% difference in efficacy of the two compounds, a sample size of 128 would be necessary. Only the trials involving citalopram6 and moclobemide7 against placebo in the over 65swould meet this criterion.

    Which of the newer drugs should therefore be selected to treat depression in older people? Citalopram, moclobemide, and probably fluoxetine are more effective than placebo in older depressed patients. The serotonin reuptake inhibitors fluvoxamine, paroxetine, and sertraline as well as milnacipran and venlafaxine are probably (but not unequivocally) as effective as older antidepressants in this population. Fluoxetine is also effective in treating elderly patients with dementia and depressive symptoms. So far, however, trials have failed to establish that the new non-tricyclic antidepressants are safer than the older tricyclics in elderly people, with the possible exception of paroxetine. Tricyclic antidepressants, especially amitriptyline and dothiepin,16 are known to pose a high risk of death in overdosage. These drugs should therefore be avoided in older people whose medication is not supervised and who are at risk of taking an overdose. Beyond that it is hard to recommend the use of the newer drugs for old people on safety grounds.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    View Abstract

    Log in

    Log in through your institution


    * For online subscription