Can prayer heal a sick patient? Though polls show that most Americans believe so, scientists remain sharply divided on the question. In tomorrow's issue of The Lancet, three psychiatrists critique several hundred studies that purport to show a link between religious faith and health benefits. Typically, they say, these studies ignore other factors that may improve health, such as abstinence from tobacco and alcohol. And even the scientifically sound studies, they contend, were inconsistent and don't justify bringing religion into medical practice.
Richard Sloan of Columbia University and his colleagues reviewed every article connecting religion and physical health they could find in Medline, an online service that indexes medical studies. Many of them, he says, focused on such groups as Roman Catholic priests or Benedictine monks, which forbid certain risky behaviors. Others looked at more general populations of churchgoers and found lower disease rates, but failed to take into account that only people who are in fairly good health can go to church. When these confounding factors were taken into account, either by the original researchers in a follow-up study or by Sloan's group, the alleged benefits usually disappeared. Overall, Sloan says, "the evidence is very unconvincing and weak.... Much weaker, for example, than the link between marital status and health."
Although praying for recovery may seem harmless, the scientists say many people blame themselves if their prayers are not answered. Patients and family members "can be wracked with guilt," says Sloan, who is also concerned by the growing movement to encourage doctors to bring up religion and to offer to pray with their patients. "We'd be outraged if a physician said 'I'm recommending you to get married,' " he says. "Religion, like marriage, is a very personal and private matter."
"No one has really shown that religion causes better health," acknowledges Harold Koenig, the director of the Center for Religion/Spirituality and Health at Duke University. In part, he says, that may be due to sloppy science, but it's also because "religiousness" is a tricky concept to quantify. But Koenig asserts that several yet-to-be-published studies do control for confounding variables in the way that Sloan recommends and still show a positive effect of religion on health. The Lancet paper, Koenig says, is important because few people have challenged the link between religion and health. "The field needs that challenge to become stronger and better."