ORLANDO – The bad news about antidepressant therapy for patients whose first episode of depression occurs in late life is that the randomized trials show it’s no better than placebo. The good news is that placebo is pretty darn effective.
What are clinicians to make of these research findings?
Bruce Jancin/Frontline Medical News
Dr. J. Craig Nelson
In everyday practice, make sure to give patients with late-life depression the sort of TLC they would have received in the placebo arm of a randomized drug trial. Only don’t call it placebo therapy. First of all, it’s not ethical to give patients a placebo outside of a formal clinical trial, and secondly, what goes on in the control arm of a randomized antidepressant trial is not really a placebo anyway, Dr. J. Craig Nelson explained at the annual meeting of the American Association for Geriatric Psychiatry.
"I don’t know that it’s a mistake to give those people antidepressants. I think it is a mistake, though, not to provide the elements of clinical management that those people really seem most responsive to. So if you do use an antidepressant, don’t just prescribe it and then see that person 2 months later. You want to see the patient on a regular basis and take advantage of what’s often called a placebo response but is really a response to clinical management," said Dr. Nelson, professor of psychiatry and director of geriatric psychiatry at the University of California, San Francisco.
"Provide those elements of supportive care that are important," he urged. "It’s being human, but to be more specific it’s being human and providing attention, reassurance, support, education about depression, and monitoring of symptoms and side effects. You want to see the patient regularly and take advantage of that effect."
In his meta-analysis of 10 randomized, placebo-controlled trials of antidepressant therapy, using data on more than 2,300 outpatients aged 60 years or older with late-life major depressive disorder, he was able to identify one specific subgroup with a markedly better response to antidepressant medication than to placebo.
Among the 385 patients with at least a 10-year duration of depression along with at least moderately severe disease as defined by a baseline Hamilton Depression Rating Scale score of 21 or more, there was a 58% response rate to medication, compared with a 31% response to placebo. The number needed to treat with an antidepressant rather than placebo to achieve one additional response in this group was a very respectable four. For all others, however, the number needed to treat was a whopping 21 (Am. J. Psychiatry 2013;170:651-9).
Antidepressants are clearly ineffective in patients with depression and dementia. Dr. Nelson demonstrated this in a meta-analysis of seven randomized trials with 330 patients, in which antidepressants showed no significant benefit over placebo in terms of response or remission rates (J. Am. Geriatr. Soc. 2011;59:577-85). Moreover, a subsequent large, placebo-controlled randomized trial of sertraline or mirtazapine involving 326 patients with depression and dementia also showed no suggestion of an advantage for drug therapy over placebo (Lancet 2011;378:403-11).
Another group for whom randomized trials have shown antidepressant medications are likely to fail is patients with late-life depression plus executive dysfunction. This is a population for whom problem-solving therapy has proved to be quite effective, the psychiatrist noted.
He highlighted what he called "arguably the best controlled study of psychotherapy in older depressed adults." It included 221 subjects aged 60 years and older with executive dysfunction who were randomized to 12 weekly sessions of problem-solving therapy or supportive therapy. The two groups showed comparable reduction of depressive symptoms for the first 6 weeks, then the group assigned to problem-solving therapy pulled ahead.
At week 9 they had a response rate of 47%, compared with 29% in the supportive therapy group. By week 12 the margin had grown to 57% vs. 34%. The week 12 remission rates were 46% and 28% (Am. J. Psychiatry 2010;167:1391-8). The group that received problem-solving therapy also showed a significant advantage in terms of improvement in disability, as measured using the 12-item WHO Disability Assessment Schedule II (Arch. Gen. Psychiatry 2011;68:33-41).
Dr. Nelson reported serving as a consultant to Bristol-Myers Squibb, Eli Lilly, Lundbeck, Otsuka America, Pfizer, Shire, and Sunovion.