Journal of Religion and Health, Vol. 26, No. 1, Spring 1987 Is There a Religious Factor in Health? JEFFREY S. LEVIN and PRESTON L. SCHILLER ABSTRACT: This paper reviews epidemiologic studies employing religion as an independent con- struct, and finds that most epidemiologists have an extremely limited appreciation of religion. Af- ter a historical overview of empirical religion and health research, some theoretical considerations are offered, followed by clarification of several operational and methodological issues. Next, well over 200 studies are reviewed from nine health-related areas: cardiovascular disease, hyperten- sion and stroke, colitis and enteritis, general health status, general mortality, cancer of the uter- ine corpus and cervix, all other non-uterine cancers, morbidity and mortality in the clergy, and cancer in India. Finally, an agenda for further research is proposed. Historical background It has been nearly 150 years since Benjamin Travers remarked that he had never seen a case of cancer of the penis in a Jew, 1 and almost that long since Rigoni-Stern first observed that Catholic nuns in Verona, Italy, were at signif- icant risk for breast cancer yet significantly protected against uterine cancer. 2 Half a century later, in 1891, John Shaw Billings first considered religious af- filiation to be a potential factor in the study of differential rates of morbidity and mortality among social groups2 Three-quarters of a century have now elapsed since William Osler broached the issue of a "faith that heals, TM a senti- ment recently echoed by Frank, ~ and nearly sixty years have passed since Paulsen reviewed "religious healing" movements in JAMA.8 Over a quarter of a century has gone by since White expounded upon the interactions between the psyche and the soma, 7 and it has been more than a decade since several in- fluential researchers attempted to stimulate research into the effects of illness behavior on religious beliefs 8 and the relations between religious institution attendance and mortality2 Finally, nearly another decade has elapsed since Vaux attempted to promote research into the "intersection of religion and health behavior in order to better delineate their contribution to American life ''1~ and since Vanderpool asserted the "therapeutic significance" of reli- gion. 1~ Although many epidemiologists continue to collect some information about subjects' religious preference, background, or practice as part of their inquir- Jeffrey S. Levin, M.P.H., is with the Division of Sociomedical Sciences, Department of Preven- tive Medicine and Community Health, at The University of Texas Medical Branch in Galveston, Texas. Preston L. Schiller, Ph.D., is with the RDI: Research/Development/Information, 6502 106th Ave., N.E., Kirkland, WA 98033. The authors wish to thank Dr. C. David Jenkins for his com- ments on an earlier draft of this manuscript. 9 9 1987 Institute of Religionand Health