Major Study Finds Antidepressants Work, But May Have Limitations

A meta-analysis of existing trials suggests that the drugs are mostly effective on a short-term basis for patients suffering from acute depression

About 13 percent of Americans over the age of 12 use antidepressants, but how well—or whether—they work is still a question for many patients. Bojan89/iStockPhoto

For years, researchers have been debating the efficacy of antidepressants, which are used by around 13 percent of Americans over the age of 12.  As Aaron E. Carroll reports for the New York Times, a comprehensive new study has shown that the drugs do work, but their benefits may be modest.

The study, published late last month in The Lancet, analyzed 522 double-blind studies that included 116,477 patients and 21 commonly prescribed antidepressants. The paper marks the most comprehensive study on antidepressants to date, and, according to its authors, includes “the largest amount of unpublished data” of any study on the medications.

Using a network meta-analysis technique, which allows scientists to simultaneously compare multiple treatments, the researchers sought to determine antidepressants’ efficacy and their acceptability, or how many patients stuck to taking the drugs at prescribed intervals. The results of the study showed that all 21 of the antidepressants were more effective than placebos in reducing depressive symptoms during the first eight weeks of treatment. Large and small trials did not vary significantly in their results.

The new analysis suggests, however, that the efficacy of the drugs may be limited. For one thing, the benefits applied in the short term, and only to patients who are suffering from acute major depression. “[W]e still do not know how well antidepressants work for those with milder symptoms that fall short of major depression, especially if patients have been on the drugs for months or even years,” Carroll writes. “Many people probably fall into that category, yet are still regularly prescribed antidepressants for extended periods.”

The researchers also found evidence of the “novelty effect:” Antidepressants worked better when they first came onto the market, but were shown to lose efficacy and acceptability as the years went on. Additionally, not all of the drugs were shown to be equally effective. “Some antidepressants were more effective than others, with agomelatine, amitriptyline, escitalopram, mirtazapine, paroxetine, venlafaxine, and vortioxetine proving most effective, and fluoxetine, fluvoxamine, reboxetine, and trazodone being the least effective,” the study authors write. (These are the generic names for antidepressants commonly sold in the U.S., including Prozac, Elavil and Paxil.)

As Olivia Goldhill of Quartz points out, much of the past evidence showing that antidepressants may be no more effective than placebos derives from a 2008 meta-analysis led by Irving Kirsch, associate director of the Program in Placebo Studies at Harvard Medical School. Kircsh used the Freedom of Information Act to obtain data that was sent to the Food and Drug Administration by pharmaceutical companies seeking approval for antidepressants. (Unlike peer-reviewed literature, pharmaceutical companies must submit the results of their trials regardless of the result.) They found 74 trials covering more than 12,500 patients. The researchers showed that around half of those trials showed antidepressants were more effective than a placebo. But as Carroll notes, the published literature did not reflect these results. While all of the “positive” studies were published, only three of the “negative” studies, or studies that did not show statistically significant benefits for antidepressants, had been published in journals. Eleven had been presented to appear positive, and 22 never appeared in medical literature.

In an effort to gain a more accurate picture of the drugs’ effectiveness, the new study analyzed 86 unpublished trials found on trial registries and pharmaceutical company websites. The new analysis includes an additional 15 unpublished studies discovered through “personal communication or [by] hand-searching other review articles,” the study authors write. Researchers also asked for and received unpublished data from more than half of the trials included in the study, in order to account for potentially misleading representations of trial outcomes in published journal articles. But lingering questions of bias remain. Of the 522 trials included in the new analysis, 409 were funded by pharmaceutical companies.

So while the new study suggests that antidepressants are more effective than a placebo, at least in some cases, media reports claiming that the research shows “antidepressants do work, and many more people should take them” are not entirely accurate.

“The clinical (as opposed to statistical) significance of the treatment effects that we detected will continue to be contested,” John Ioannidis, a co-author of the study who has previously argued that the benefits of antidepressants have been inflated, tells Carroll. “[I]t is still important to find ways that one can identify the specific patients who get the maximum benefit.”

Get the latest stories in your inbox every weekday.