|Do Antidepressants Really Work?|
During the past couple of decades, drugs from the class of antidepressants called SSRIs, as well as newer drugs with novel modes of action, have been given credit for helping many people reclaim their lives from the devastating grips of depression. These drugs have also been responsible for making their manufacturers very wealthy. But, what if these drugs don't really provide significant relief for depression at all? What if the placebo effect is responsible for most of their effectiveness?
Kirsch, a psychologist at the University of Connecticut, pooled together data from 38 studies of six drugs approved by the FDA between 1987 and 1999 (Prozac, Zoloft, Paxil, Serzone, Celexa and Effexor). The studies were placebo-controlled clinical trials in which some patients were given the antidepressant while others were given an inert pill called a placebo. What he found was surprising. On the average, the placebo groups improved by eight points on the 50-point Hamilton Depression Scale, while those given the active ingredient improved by an average of 10 points. This difference, argues Kirsch, is "clinically negligible".
What Does This Mean?
According to Kirsch, although mean differences between the antidepressants and placebo were small, the differences were in favor of the drugs and they were statistically significant. What this means is that the drug did appear to have some effect on mood which could be proven by the data. This small difference can be accounted for in two ways. One, the active drug actually has an effect, although not a large one; or, two, the difference resulted from an enhanced placebo effect.
An enhanced placebo effect may occur with these particular antidepressants because, despite efforts to keep patients unaware of which group they are in, it is well-known that antidepressants produce side-effects. If a patient is having side-effects, they know they must have received the active drug and therefore have expectations of feeling better. In fact, it has been shown that the ability of patients to deduce whether they are in the drug or placebo group in antidepressant trials does exceed chance. While this enhanced placebo effect is small, one must bear in mind that the difference in improvement between the drug and placebo groups was also small--only 2 points on the HAM-D.
Possible Avenues for Future Exploration
One problem with the research, admits Kirsch, is the fact that his conclusions are based upon the assumption that placebo and drug effects are additive, meaning that he is assuming one can simply subtract the placebo's effect from the drug's effect and the difference between them is what the drug caused. It is possible that the placebo effect played a smaller role in the patients taking the active drug, therefore meaning that the drug had a greater effect than the difference between the two figures would indicate. If it turns out that the effects are not additive, conventional methods are not appropriate for testing these drugs. Kirsch suggests that it may be necessary to alter future study designs to determine whether the model of additive effects is appropriate for evaluating antidepressant drugs. If it is determined that effects are additive, this means we will have to reevaluate our current arsenal of drugs. If they are not additive, new methods will have to be designed to appropriately evaluate new drugs coming down the pipeline.
The Emperor's New Drugs: An Analysis of Antidepressant
Medication Data Submitted to the U.S. Food and Drug Administration.
Irving Kirsch, Thomas J. Moore, Alan Scoboria, and Sarah S. Nicholls.
Prevention and Treatment July 15, 2002, Vol. 5.