The study leaders did not anticipate that prayer might be harmful and 'allowed cultural presumption to undermine scientific objectivity,' wrote Dr. Mitchell Krucoff of Duke University Medical Center and others in an editorial.

In fact, the researchers from Harvard Medical School and five other U.S. medical centers found -- to their bewilderment -- that coronary bypass patients who knew strangers were praying for them fared significantly worse than people who got no prayers. The team speculated that telling the patients about the prayers may have caused "performance anxiety," or perhaps a fear that doctors expected the worst.

"Obviously, my colleagues were surprised by the unexpected and counterintuitive outcome," said Rev. Dean Marek, director of chaplain services at the Mayo Clinic in Rochester, Minn., and a co-investigator for the project.

Mixed Results in Earlier Studies

It was a strange end for the mammoth prayer study, which cost $2.4 million and enrolled 1,802 patients who had bypass surgery. The majority of funding came from the British-based John Templeton Foundation, which supports research at the intersection of science and religion.

Previous studies had examined the power of prayer for medical patients, with mixed results. Most did not have the statistical power to reliably detect the effects of prayer, if it had an effect.

The new study, which appears in the April issue of the American Heart Journal, was designed to be large enough to see if patients who knew they were being prayed for had better recoveries.

The people who prayed for the patients were strangers -- either Roman Catholic monks or believers belonging to other Christian denominations. Those who prayed were given the patients' first names and last initials, and instructed to give a simple prayer for a quick recovery with no complications. The researchers said they could not find a non-Christian group that could work with the scheduling demands of their study.

Researchers Puzzled

Bypass patients who consented to take part in the experiment were divided randomly into three groups. Some patients received prayers but were not informed of that. In the second group the patients got no prayers, and also were not informed one way or the other. The third group got prayers and were told so.

There was virtually no difference in complication rates between patients in the first two groups. But the third group, in which patients knew they were receiving prayers, had a complication rate of 59 percent -- significantly more than the rate of 52 percent in the no-prayer group.

Researchers said the study was never intended to prove or disprove the existence of God or to settle theological questions. But they had expected that knowing someone was praying for the patients might help those patients relax and bring about a state of well-being, which can reduce strain on the heart.

"In this study we did not find that was the case," said Dr. Herbert Benson of Harvard Medical School, a principal investigator of the study.

The researchers were at a loss to explain the worsened outcomes in their study. An accompanying editorial in the journal criticized the study authors for taking "an almost casual approach toward any explanation, stating only that it 'may have been a chance finding.'"

Unfair Test of God?

The editorial authors, led by Dr. Mitchell Krucoff of Duke University Medical Center, wrote that the study leaders had not anticipated that prayer might be harmful and had "allowed cultural presumption to undermine scientific objectivity."

In light of the significant findings, the editorial concluded, researchers "must be vigilant in asking the question of whether a well-intentioned, loving, heartfelt healing prayer might inadvertently harm or kill vulnerable patients in certain circumstances."

Any attempt to study the power of prayer objectively runs the risk of scientific and theological problems, said Dr. Daniel Sulmasy, director of ethics at St. Vincent's Hospital and New York Medical College.

"God is not just another therapeutic nostrum in a doctor's black bag," said Sulmasy, who is also a Franciscan friar. "It seems fundamentally sinful to conceive of God as our instrument."

Marek, a Catholic priest, conceded that it may be an unfair test of God to measure whether detailed prayers are granted.

"The best prayer probably is, 'Thy will be done,'" he said.