Before 2000, nearly 1,200 studies had examined the relationship between religion and health, with most studies reporting positive associations.¹ In the past five years many new studies have been conducted that support these findings.² Although there has been criticism of the earlier studies,³ these studies used different designs and methods of analysis, were conducted in different populations, and were carried out by different investigators, both inside and outside the United States. Most studies are cross-sectional and therefore cannot distinguish between cause and effect; however, many prospective cohort studies and more than one dozen clinical trials (mostly examining religious interventions in mental health) support the findings from cross-sectional research. We review some of the best studies below.

Strawbridge and colleagues⁴ conducted a 28-year prospective assessment of more than 5,000 adults. They found that weekly attendance of religious services decreased the relative risk (RR) of dying during follow-up by 36 percent (RR, 0.64; 95% confidence interval [CI], 0.53 to 0.77); adjusting for social connections and health practices reduced the effect to 23 percent (RR, 0.77; 95% CI, 0.64 to 0.93). The effect in women (RR, 0.66; 95% CI, 0.51 to 0.86, adjusted for covariates) approximated the effect of not smoking cigarettes.⁵ Frequent attendance was a predictor of better health behaviors, improved mental health, and increased social connections.⁶

These findings were replicated in a sample of 4,000 older adults followed for six years.⁷ The effect was similar (RR, 0.72; 95% CI, 0.64 to 0.81, adjusted) and was strongest in women (RR, 0.65; 95% CI, 0.55 to 0.76, adjusted). A random survey⁸ of more than 20,000 Americans found that whites who attended religious services regularly lived an average of seven years longer than those who did not, and blacks who attended regularly lived an average of 14 years longer than those who did not. After controlling for multiple covariates and explanatory factors, the risk of dying during the eight-year follow-up was 50 percent higher in persons who never attended religious services than for those who attend more than once per week.

Most recently, Lutgendorf and colleagues⁹ prospectively examined religious attendance and interleukin-6 (IL-6) levels as they relate to mortality rates in 557 older adults. After controlling for multiple covariates and explanatory factors, frequent attendance of religious services reduced the risk of dying in the six-year follow-up period by 78 percent (odds ratio [OR], 0.32; 95% CI, 0.15 to 0.72) compared with nonattendance; this finding seemed to be mediated by decreased serum IL-6 levels (OR, 0.34; 95% CI, 0.16 to 0.73), which replicated earlier findings.¹⁰ High IL-6 levels are an indicator of immune system dysfunction and thus provide a possible biologic mechanism by which religious attendance may influence physical health.

Although attendance at services is the most powerful predictor of health related to religion, it is not the only variable that predicts health outcomes. Studies^11–18 show a connection between religious involvement and several health-related outcomes, including mental health and substance abuse, social health, quality of life, positive health behaviors, disease screening, continuity of care, surgical complications and use of health services, endocrine and immune function, hypertension, coronary artery obstruction, carotid atherosclerosis, survival rates, and positive human traits (e.g., forgiveness, gratitude, meaning and purpose, optimism and hope, altruism). Religious beliefs also influence medical decisions related to chemotherapy,¹¹ do-not-resuscitate status,¹² development of advanced directives,¹³ and end-of-life care.¹⁴ The physical health benefits of religion make sense given what is known about the effects of negative emotions on health outcomes and quality of life, particularly for patients with heart disease^15,16 and cancer.^17,18

Although studies that report negative associations between religion and health are not nearly as common as those with positive findings, recent studies have found associations between religion and obsessive-compulsive disorder in persons living in Italy,¹⁹ an increased risk of breast cancer among women raised in religious homes in California,²⁰ poorer control of diabetes in Muslims in Leeds (U.K.),²¹ and reduced survival rates after hospital discharge among patients with religious struggles in North Carolina.²² Religion involves some of the most deeply and passionately held human beliefs, and it should not be surprising that these beliefs affect health.

Because the findings on religion and health may be relevant to patient care, physicians should have a basic knowledge of the existing research. Whether religion is good or bad for health, studies indicate that it is a powerful factor influencing adaptation to illness, medical decisions, health beliefs, and behaviors. Although we continue to struggle with how to apply information relating religion and health to clinical practice, sensitive and sensible applications do exist.²³