Researchers say the benefits of psychotherapy for depression have been overstated by previous research as a result of publication bias.
Depression is one of the most common mental disorders in the US, with major depressive disorder (MDD) affecting around 6.7% of adults.
Antidepressant medication is the first-line treatment for patients with depression. But according to study coauthor Steven Hollon, professor of psychology at Vanderbilt University in Nashville, TN, and colleagues, previous research has found the benefits of such treatment have been overstated as a result of publication bias.
As an example, they point to one study that identified 74 placebo-controlled antidepressant studies that had been submitted to the US Food and Drug Administration (FDA). Of the 38 studies that showed the benefits of antidepressants, 37 were published. However, of the 36 studies that showed antidepressants in a negative light, 61% were not published, while 31% were “spun” to make them appear positive.
For their study, Hollon and colleagues set out to investigate whether psychotherapy – a treatment that is frequently offered to patients with mild to moderate MDD – may also have been subject to publication bias.
Psychotherapy 25% less effective than published studies show
There are two types of psychotherapy: cognitive behavioral therapy (CBT) and interpersonal therapy (IPT). CBT focuses on helping patients tackle negative thoughts that may be contributing to their depression, while IPT aims to help patients understand and deal with relationships that may be causing or exacerbating their depression.
Numerous studies have hailed psychotherapy for its efficacy in treating depression, and as a result, the National Institute of Mental Health (NIMH) recommend it as an effective treatment option for the condition.
Hollon and colleagues searched National Institutes of Health (NIH) databases to identify all randomized clinical trials that had received grants from the NIH and had tested the effects of psychotherapy against depression between 1972 and 2008.
Of the 55 studies that had received NIH grants, 13 had not been published. The team then contacted the researchers of these studies to get their results, before adding the unpublished results to studies that were published and conducting a series of meta-analyses.
While the team found that psychotherapy was effective for treating depression, reviewing both the unpublished and published studies together resulted in a 25% fall in the efficacy of the treatment, suggesting psychotherapy may not be as beneficial as previously claimed.
Commenting on the results, Hollon says:
“This study shows that publication bias occurs in psychotherapy, mirroring what we’ve seen previously with antidepressants and other drugs.
This doesn’t mean that psychotherapy doesn’t work. Psychotherapy does work. It just doesn’t work as well as you would think from reading the scientific literature.”
Hollon explains that, although studies are assessed through peer review, there are certain loopholes that allow the benefits of treatments to be overstated and possible risks to be understated. “The consumers of this skewed information are health care providers and, ultimately, their patients,” he adds.
In order to avoid this type of publication bias going forward, the team recommends that funding agencies and journals archive the raw data and original protocols of both published and unpublished studies.
Last month, Medical News Today reported on a study published in The BMJ, in which a reanalysis of studies investigating the efficacy of the antidepressant paroxetine found it is unsafe and ineffective for teenagers with depression.
Written by Honor Whiteman