And why shouldn’t religious belief and activity promote health? As Levin points out, religious involvement generally promotes healthy behavior and lifestyles and positive thinking. It offers social support and discourages isolation. Doctors who discount the participation of the hospital chaplain are being foolish, and researchers are realizing the important role stress — mitigated, in many instances, by religious involvement — can have on the immune system. If you don’t believe any of this, you can agree that at the very least, attendance at church or synagogue services — a main measure used by many studies — selects those who are healthy enough to get out of bed and may live more moderate lives. Put most plainly, if you’re out late drinking on Friday or Saturday night, you’re not very likely to make it to synagogue Saturday or church Sunday. And if you’re very ill, you won’t make it, either.

Richard Sloan, a psychologist at Columbia University, is one of the more outspoken critics of this statistical sleight of hand and some of the deeper philosophical problems of the notion that religious belief leads to good health. He has published fairly severe critiques of such ideas in the Lancet and the New England Journal of Medicine. Sloan worries, rightfully so, about the negative effects of religion on health. A patient who believes that illness is the result of insufficient faith or prayer may be plunged into dangerous despair and self-reproach by the diagnosis of a major disease like cancer. And from a statistical point of view, many of the studies cited by proponents of the religion-health link are flawed, either because the researchers did not originally set out to measure religious beliefs or practices or because other factors — lifestyle, diet, genetics — are not adequately taken into account.

Levin acknowledges some of these problems, but barely. To some extent, he’s spinning the data, making an effort to come off as a hard-scientific epidemiologist but falling short. For example, he cites evidence that several religious groups are less likely to die from certain diseases. The death rate from heart disease among one Mormon sect is only about 80 percent of that of other citizens of the sect’s state. Seventh-Day Adventists are less likely to suffer from high blood pressure. Jews are less likely to suffer from cancers of the uterus, cervix and penis.

But there’s a better explanation than beliefs or church attendance at work here. None of these communities are particularly heterogeneous, nor are they random samples of the population. They’re tightly woven groups that discourage intermarriage and have, as a result, concentrated their genes, good and bad. So it’s no surprise that they have lower rates of a particular disease with a strong genetic component like cancer.

This is particularly true of Ashkenazi Jews. Sure, Ashkenazi Jews have a lower rate of some cancers. (Some of that may even be a result of circumcision, in the case of penile cancer, although that’s a controversial theory.) But research also indicates that other groups of Jews have at least a slightly higher risk of breast or colon cancer, risks that also have a genetic component. Why not acknowledge that? Similarly, why not acknowledge that once a group has a lower rate of high blood pressure — whether because of genes or prohibitions against eating meat, drinking alcohol and smoking — it’s pretty much guaranteed a lower rate of heart disease?

But the most significant problem with the religion-health connection — and the most troubling part of Levin’s book, even though it’s really just two chapters — arises when proponents give more credence than they should to the idea that patients are helped by remote prayers spoken by others. It’s a provocative idea, and one that has the backing of several studies. The most-often-cited one was of coronary care unit patients, published in 1988, but a larger study in 1999 confirmed the basic findings of the first, and an even larger study of some 2,000 patients is currently underway at Harvard Medical School. The subject is worthy of more study, even if it sounds outlandish. No one is ready to suggest that doctors prescribe prayer, although at least one survey found that patients wished their doctors would pray with them.

Levin, however, takes it too far. When he starts talking about “psychoenergetic systems” and “interconnections among subtle energies, physiological systems, chakras and meridians,” I think readers can be forgiven for puzzled looks. This stuff is decidedly non-Cartesian and backed up by zero evidence. One of the studies he cites is so bizarre I had to read it twice to make sure I wasn’t hallucinating. In it, doctors actually inflicted 44 male college students with wounds to test a highly questionable “modern subtle-energy-based healing technique” called “noncontact therapeutic touch,” whose powers Levin attributes to the realm of the “superempirical” — rather than the “supernatural.” When all else fails, he invokes “nonlocal” physics to explain how someone praying miles away from a sick patient can heal the patient.

Levin recognizes the value of rigorous peer review in medical journals, but seemingly only when it suits him. When there are no studies that meet his needs published in such journals, he turns to such rigorous experts as New Age health guru Andrew Weil. At another point, he cites “1,385 medical journal articles reporting cases in which cancer and many other serious chronic diseases seemed to miraculously disappear following treatment believed to be inadequate, or with no treatment at all.” Cancers spontaneously remit all the time. That doesn’t prove the placebo effect; it just proves we don’t know enough about cancer.

Levin can also overstate the acceptance of his own theories. In his introduction, he crows about a paper he and colleagues published in the Journal of the American Medical Association in 1997. The honor of being featured in JAMA even makes it into his book jacket bio. But read almost 200 pages further, and you learn that Levin claims to have been “shocked to discover that parts of my final manuscript had been excised.” Well, a few facts should be set straight. Levin published his paper in Pulse, the medical student section of JAMA — of which I was co-editor in chief at the time. The part of the article that was cut was the least substantiated, most speculative section of the piece. As he himself acknowledges in the book, “the evidence linking prayer to healing in humans, in my opinion, is less conclusive.”

Levin does a good job summarizing much of what we know about the connection between spirituality and health. Readers new to the field will probably find themselves willing to believe that religious affiliation and activity are probably a good marker for improved health, and may even promote it. But they should be leery of those who try to take the evidence much further than that, and be conscious of the bad effects religion can sometimes have. Read a history of the crusades for even more evidence on the negative effects of religion on the public’s health.

One member of the clergy attending the Duke conference, prefacing a question by wondering why a conference on faith was being held on a Sunday morning when he and others would ordinarily be ministering to their congregations, expressed doubts about the importance of the subject matter. Religion is for praising God, not for healing, he said. The response from a panelist was that if one of the “side effects” of religious involvement is health, why not acknowledge it?

To that, I say, “Amen.” But don’t risk incurring God’s wrath by pushing it any further.