We've all heard of people who try an antidepressant for a while but then report no improvement in their mood. You may have even read in the New York Times a few weeks ago about a new study that showed that antidepressants just don't work. If you read that
We've all heard of people who try an antidepressant for a while but then report no improvement in their mood. You may have even read in the New York Times a few weeks ago about a new study that showed that antidepressants just don't work.
If you read that article carefully, though, you saw that these medications actually do work quite well—in people with moderate to severe depression. People who have "mild" depression will not notice any improvement. Any kind of depression is painful, whether mild, moderate, or severe, so perhaps "mild" is not a term anyone would choose to describe his or her own depression.
How can you tell whether you have depression that will respond to antidepressants?
The same way a general practitioner knows that antibiotics work quite well for someone who has a cough that's due to pneumococcal pneumonia, but not for people who have a cough from post-nasal drip. Antidepressants aren't effective against the type of sadness or mild depression that is a single symptom, or an understandable emotional response to the everyday hard knocks of life: loss, frustration, or disappointment. Antidepressants are only intended to treat the more serious depression that is a part of a persistent and pervasive mental illness.
How did the researchers define mild depression?
They combined data from 6 different studies with over 700 people, after first making sure that all 6 had used the same criteria to judge how severe someone's depression was. (In this case, all 6 studies had used an instrument called the Hamilton Rating Scale for Depression).
This type of a research study, which combines data from a number of experiments that have used comparable methods, is known as a meta-analysis, and it aims to discover truths that may have been hidden in the smaller individual studies. And the question the scientists were asking is not whether the participants got better, but whether they got significantly better on active medication than on placebo.
Using the Hamilton Rating Scale, how bad would a person's depression have to be before an antidepressant would be expected to help? Pretty severe, actually. Here is list of symptoms from the Hamilton scale that, if in a mild form, would still be considered under the threshold where antidepressants can be shown to work better than placebo:
Complaints about bad mood. Feeling life is not worth living. Feeling like a disappointment to others. Feeling tense and irritable. Not feeling up to work or hobbies. Sometimes taking more than a half hour to fall asleep. Feeling restless and disturbed at night. Having trouble sleeping through the night. Feeling somewhat worse in the morning than later. Seeming to move or talk a bit slow. Fidgeting, some stomach troubles, nerve troubles, or other physical symptoms. Some worries about physical well-being. Low appetite, but still eats. Losing a little bit of weight. Feeling somewhat heavy, achy, tired out. Some loss of interest in sex. Blaming other people or things for the depression. Feeling somewhat out of touch with reality. Suspicious of others. Preoccupied with intrusive thoughts or impulses.
But if a person had nearly all of these characteristics of depression in mild form, as well as a few other items at moderate or high severity, then that person would be considered likely to respond to antidepressant medication.
My point is, this study was not giving the antidepressants a free ride; anyone who was helped by them and showed improvement was starting out with moderate or severe illness. Another way to think about it is this: As a rule of thumb in any medical study, the closer the patients are to being well, the harder it is to measure a difference in treatment response.
Some final thoughts: