up SearchFeedback[help] CPMCnet

R E S E A R C HR E P O R T S

Researchers Find Little Evidence to Support Religion-Health Link

Lead Researcher: Richard P. Sloan

In the first and only comprehensive review published by a major medical journal, researchers at P&S found only very weak empirical support for claims that religious faith promotes physical health. The new study applied standards of medical scientific reliability and statistical soundness to previously published studies examining religion and health. At a time of growing interest in introducing religious activity in medicine, the review raises fundamental scientific and ethical questions.

The research, reported in the Feb. 17, 1999, issue of Lancet, shows that many previous investigations failed to identify other variables that could account for an apparent link between religious practice and health, such as age, sex, functional capacity, education, ethnicity, socioeconomic class, and marital status.

“It is critically important that claims of religious activity be subjected to the same rigorous inquiry as any other medical claims,” says Dr. Richard P. Sloan, associate professor of psychiatry and director of behavioral medicine. “There is no compelling evidence that religious activities promote health.”

Dr. Sloan cites one oft-quoted study that showed a positive association between church attendance and health but did not account for the fact that very ill people cannot get to church to be counted. “It was a classic case of failure to control for essential covariates,” he says. The researchers published a follow-up paper recognizing this, but it is rarely reported.

Another common flaw in attempting to confirm the link is assessing so many measures of religious practice and health outcomes that, by chance, two happen to correlate. “Inappropriate statistical methods appear all the time in these sorts of papers,” Dr. Sloan says. Yet another limitation of the studies is the murky and many definitions of religious activity, which can range from regular church or synagogue attendance to personal meditation or belief in God. For now, the researchers conclude, physicians shouldn’t tell patients that religious belief can improve health because no scientific evidence shows that this is so.

Yet even if a well-designed study could identify causal links between religion and health, the Columbia researchers advise that a physician’s probing this area is inappropriate because religion is an intensely personal and private matter. Dr. Sloan offers the comparison of marriage, which is widely regarded as having health benefits. “The evidence is very solid. But we would never expect a doctor to say, ‘There is compelling evidence that marriage is beneficial to health. My advice is to get married.’ That would be regarded as an outrageous intrusion. It is even more so for religion than marriage.”

Advising a patient that religious activity can promote health may actually challenge the physician’s pledge to “do no harm,” especially when health declines. “If a doctor suggests, or even if a patient simply draws an inference, that a health outcome is a product of religious activity, it could be quite harmful. A patient will think, ‘Did I not pray enough? Am I a moral failure?’ It is bad enough to be beleaguered by illness, but it is worse to have to confront guilt and remorse too,” says Dr. Sloan. A physician must not use his or her considerable authority, the authors wrote, to impose a sense of guilt on patients who may not be religious.

The authors stress that physicians should respect patients who draw upon their faith in times of illness, but they caution against using religion as a medical intervention in the absence of strong empirical support and appropriate ethical guidelines.

The study’s other authors were Dr. Emilia Bagiella, assistant professor of clinical public health, and Dr. Tia Powell, assistant clinical professor of psychiatry.

Return to Table of Contents | Return to Research Reports