Have studies proven that prayer can help heal the sick?

SDStaff DavidB replies:

This subject has seen more than its fair share of press in recent months. Indeed, even Christ has written about this topic! (Well, Theodore Christ, that is, as a co-author on an article, “God’s HMO: Prayer, Faith, Belief & Physical Well-Being,” in Skeptic magazine, vol. 8, no. 2 — see below.)

As with so many things, the hype about prayer definitely outshines the reality. To date there has not been a single good study showing that prayer has any value for helping sick people. A couple studies appeared to show such an effect, if you believe the media accounts, but I’ll get to that in a little bit.

First, let’s clarify what we’re talking about. Some studies have shown that people who pray on their own behalf or know they are being prayed for show an improvement in their health. However, because of the placebo effect, these results don’t prove anything about the power of prayer as such.

Two of the authors of the Skeptic magazine article mentioned earlier, William J. Matthews and Jim Conti, recently conducted a study on this very point. They got a group of sick volunteers to choose to receive either intercessory prayer (prayer on behalf of another person) or non-religious “positive visualization.” However, only one-third of each group actually got what they asked for, with the other third getting the other method, and the final third getting neither. Unsurprisingly to those who have seen the placebo effect in action, those who expected intercessory prayer felt better than those who asked for the visualization, no matter what they actually got.

To prove that prayer really works, we need a properly-conducted double-blind study (i.e., neither the subjects nor the test administrators know who’s getting the treatment and who’s getting a placebo), just as it would be for any other claimed treatment. Otherwise, if the sick person believes the prayer will help, it may, just as a sugar pill may help if a doctor tells a patient it contains powerful medicine.

So, we need to find a way to study prayer without the people being prayed for knowing about it. Results of a study like this were published in 1988 by Randolph Byrd. While proponents have claimed it is a “landmark study” proving the effectiveness of prayer, others have found significant problems (for example, Irwin Tessman and Jack Tessman in “”Efficacy of Prayer: A Critical Examination of Claims,”” Skeptical Inquirer, March/April 2000). Specifically, while the test was supposed to be double-blind and the article describing it claimed it was, a number of investigators have found that this was not true. Byrd himself determined who did better, those who were prayed for or those who were not, and he determined it after he knew who was in which group. Furthermore, the coordinator of the study was not blinded. Double-blind studies are done so those running the experiments don’t accidentally contaminate the results with their own viewpoints. The failure to properly blind this study calls the results into serious question.

Matthews, Conti, and Christ further point out that there was no difference between those who were prayed for and those in the control group in terms of objective measures such as length of stay in intensive care, stay in the hospital overall, or number of medications that were necessary at discharge. Indeed, as pointed out by Gary Posner, M.D. (Free Inquiry magazine, Spring 1990, available at www.hcrc.org/contrib/posner/byrd.html), the length was unaffected even though there were specific prayers for a rapid recovery. Furthermore, there was no effect seen on mortality, despite prayers “for prevention of … death.”

So what was the basis of the claim that prayer was effective? The study looked at a large number of criteria or “variables” — for example, how the subjects compared in terms of rates of congestive heart failure, cardiopulmonary arrest, pneumonia, etc. With so many different comparisons to choose from, it’s not surprising that a few were found to show differences, especially when there was no prediction ahead of time as to which ones should be different or for what reasons. These differences are likely the result of chance.

The most recent study — the one you probably heard about — was done by W.S. Harris et al. and published in a recent issue of Annals of Internal Medicine. It was somewhat based on the Byrd study and received plenty of press, but many people have pointed out problems with this study as well. For example, Tessman and Tessman, as noted above, point out that of the three criteria measured, only one showed any apparent significance, and even that one was questionable. The Harris study looked at speed of recovery, adverse condition scores, and overall outcome. Speed of recovery and overall outcome showed essentially no effect, while the adverse condition scores showed a small advantage at the boundary of what would be considered chance results. Looking at the study as a whole, the Tessmans comment that one borderline positive result out of three criteria studied “is well explained by pure chance.”

Matthews, Conti, and Christ also address flaws in prayer studies, pointing to an article that appeared in the February 20, 1999, issue of The Lancet. They note that confounding variables such as age, sex, socioeconomic status, etc., weren’t taken into account, making it impossible to be sure you’re comparing apples to apples. Similarly, the studies didn’t control for multiple comparisons, which could cause errors like the ones mentioned in the Byrd and Harris studies. When you look at multiple criteria without predicting ahead of time which should change and why, you may find something caused by chance alone.

Of course, there is a question whether a true test of prayer is even possible. As Eric Stockton pointed out in a letter to the editor of Skeptical Inquirer (July/August 2000), if prayer works because of God’s intervention, and God is the omniscient deity of Christianity (or most any major religion), then He knows He is being tested. As such, He could accept or reject whatever prayer is offered, and either choose to give or not give evidence that it works. It would be impossible to properly blind such an experiment if it’s the deity we’re talking about. If it is supposed to be the prayer itself that heals, rather than God intervening, then we don’t have that issue, but we instead have to wonder how it might be that such prayer might work — if we ever get a decent study that shows it does, that is.

Other issues we may have to deal with: Are Jewish or Muslim prayers as effective as Christian ones? Within Christianity, are Catholic prayers better than Protestant ones? (The Harris study only used Christians, and did not take into account the different branches within Christianity.) Nobody would ever design a study that just asked, “Are drugs better than no drugs?” Which drugs? In what doses? The same questions could reasonably be asked about prayer, if we assume that some form of prayer actually works.

In summary, we have no good evidence of the effectiveness of intercessory prayer in which the person does not know he is being prayed for. Those who believe prayer will help them and know they are being prayed for may indeed get better, thanks to the placebo effect. The same could be said of giving pets to the elderly who like animals (which research has shown is related to both physical and psychological improvement). However, as Matthews, Conti, and Christ note, “if a patient did not like cats, for example, it would seem inadvisable to put one on an elderly lap.” Similarly, “the current research does not suggest that atheists facing heart surgery should be told by their physicians to start praying.”

There you have it. Christ has spoken.

SDStaff DavidB, Straight Dope Science Advisory Board

Send questions to Cecil via cecil@straightdope.com.

STAFF REPORTS ARE WRITTEN BY THE STRAIGHT DOPE SCIENCE ADVISORY BOARD, CECIL’S ONLINE AUXILIARY. THOUGH THE SDSAB DOES ITS BEST, THESE COLUMNS ARE EDITED BY ED ZOTTI, NOT CECIL, SO ACCURACYWISE YOU’D BETTER KEEP YOUR FINGERS CROSSED.