Science Claims for Mediation Effects Overstated
Linda Heuman interviews scientist and Buddhist Willoughby Britton:
As a scientist and as a Buddhist, what do you make of the AHRQ report? The report sounds pretty fair. This review—and pretty much every one before it—has found that meditation isn’t any better than any other kind of therapy.
The important thing to understand about the report is that they were looking for active control groups, and they found that only 47 out of over 18,000 studies had them, which is pretty telling: it suggests that there are fewer than 50 high-quality studies on meditation.
What are active control groups and why are studies based on them of higher quality? There are different levels of scientific research, different levels of rigor. I think this is a place where the public could use a lot of education. Because they don’t know how to interpret science, they assume much higher levels of evidence.
The first level is a “pre-post” study, which looks something like this: We go learn to meditate for eight weeks and at the end of it we feel better. We took a stress and anxiety scale before and after, and our stress or anxiety improved. So we say, “Meditation helped me!” That is actually not a valid conclusion. The conclusion you can make in science is that something helped. We didn’t control for the idea that just deciding to do something is going to help. Just that factor—intentionally deciding to make a commitment to your health and well-being—can make a big difference.
One problem is that just filling out the questionnaire changes you. In my recent sleep study, I had people fill out a questionnaire and keep a sleep diary. That is all they did for eight weeks. They didn’t meditate. And their sleep improved a lot. So, you have to control for the effect of taking the questionnaires.
You also have to control for the passage of time. Sometimes people just feel better after two months compared with when they started. So you can’t actually conclude that meditation had anything to do with it. A lot of the studies on meditation are pre-post studies like this. They shouldn’t count at all as evidence.
The next level of rigor is “wait-list controls.” Half the participants begin meditating immediately while the other half acts as a control group, and only later participate in the actual meditation. Those in the control group might be thinking, “I’m in the study. I’m going to learn to meditate!” They’re psyched. Their depression is already getting better because they’ve decided to do something about it. These are effects of expectation; they aren’t doing meditation.
But even at this level the study is not considered in any way conclusive. If I have an inspiring teacher, for example, it can be a helpful factor that is not meditation. Even to know that somebody felt depressed and anxious at one point and then got better is helpful. There’s the normalization of my symptoms. There’s the social support. I meet other people who have my problem. I thought I was the only person in the world who had anxiety, and now there are all these other people who have anxiety and we’re all talking about it. And I really get along with them. So I’m making friends. I’m less lonely. That’s not meditation either. There are all these things that are not meditation that could be helping me feel better.
If we really want to be able to say that meditation was the active ingredient, the control group has to do everything the other group is doing except meditation, and they can’t know that they are in the control group. This level of scientific study is called “active control groups.” But that largely isn’t what is happening in meditation research, although it’s starting to.