The trusted source for
healthcare information and
Religious Involvement and Infectious Disease
Abstract & Commentary
By Howell Sasser, PhD. Dr. Sasser is Associate Professor of Epidemiology, New York Medical College, Valhalla, NY; he reports no financial relationship to this field of study.
Synopsis: Gillum and Holt assessed the prevalence of six infections by frequency of attendance at religious services. Although results varied by race/ethnicity and factors related to sexual and drug-use practices, there appeared to be lower prevalence among those attending more often. Although these results are weakened by certain methodological issues, they provide evidence that religious practice plays a role in limiting risky behavior.
Source: Gillum RF, Holt CL. Religious involvement and seroprevalence of six infectious diseases in US adults. South Med J 2010;103:403-408.
A key objective of continuing interest in the study of religion and health is the teasing out of the constituent parts of the construct "religiousness." The observation that those who are more religiously active are also healthier in some respects than those who are not begs the question: What aspect of religion is the "active ingredient?"1 A prime candidate in the context of some disease processes is the inhibitory effect of religion on risky behaviors. This appears particularly relevant where infectious, especially sexually transmitted, diseases are at issue.
Gillum and Holt used a large cross-sectional database to assess the co-prevalence of religious involvement and several infectious agents. They drew their population from the Third National Health and Nutrition Examination Survey (NHANES III), an ongoing health surveillance project of the U.S. Department of Health and Human Services. NHANES III collected data between 1988 and 1994, and used both questionnaires and collection of biological samples to assess the health states of respondents.
Religious activity was measured with a single question about frequency of attendance at religious services (categorized for analysis as Never, Less than weekly, Weekly, and More than weekly). Serologic data were available for hepatitis A (HAV), hepatitis B (HBV), hepatitis C (HCV), herpes simplex virus type 2 (HSV-2), Toxoplasma gondii, and Helicobacter pylori (for the period 1988-1991 only). Selected potentially confounding factors, including level of education, geographic region, marital status, drug use, number of sexual partners, and overall health, were also used in the analysis.
Of 33,994 NHANES III participants with baseline data, 11,507 met age and data completeness criteria and were included in Gillum and Holt's analysis. In the total population, seroprevalence of HSV-2 and HCV was statistically significantly higher in those who never attended religious services as compared with those who attended with any frequency. Seroprevalence of HBV was higher in never-attenders in the White, Black, and Mexican-American subpopulations. After adjustment for drug use and sexual behavior, the association of religious activity with HSV-2 and HBV became non-significant among the White and Black respondents, though not among the Mexican-Americans. The association with HCV was more robust even after adjustment. No significant association of religious activity with the enteric pathogens was observed.
Gillum and Holt offer a partial answer to the question of how religious activity influences a group of health conditions. There appears to be evidence that some element of religious behavior, most likely inhibition of risky behavior, results in lower prevalence of some infectious agents among religiously active people. This inference is strengthened by the apparent attenuation of the effect of attendance at religious services after controlling for factors associated with sexual and substance abuse-related risk. However, variations in the degree of protectiveness by race/ethnicity make clear that the effect is not monolithic, and that other cultural factors, such as strength of family structures and gender role norms, may represent unmeasured confounders. A number of other issues are worth noting.
The cross-sectional design of the study leaves doubt as to the sequence in time of events. The desired inference is that religious involvement preceded exposure (or the potential for exposure). However, it is also plausible that exposure came first, and perhaps even that it influenced subsequent religious behavior. It is also not possible to locate in time the other factors that were shown in this study to be associated with religious involvement, seropositivity, or both. This limits the strength of the study's findings to suggestion of a relationship that requires further study.
The use of frequency of attendance as the measure of religiousness was dictated by what was originally collected by NHANES, but is not reflective of current practice in this area of research. A number of more complex instruments measuring beliefs and individual spiritual practices, as well as participation in organized religious services, has been proposed.2-4 These measures avoid the potential bias created by a "healthy worshiper effect" the difficulty in sorting out whether those engaged in public activities are healthier as a result, or if those not so engaged are prevented from doing so by poor health.
A positive aspect of this study was the inclusion of both infections generally regarded as venereal (HSV-2, HBV) and others not usually so characterized (HAV, H. pylori, T. gondii). (HCV has, as it were, a foot in both camps.) The finding that the results were mixed for the venereal agents, but uniformly negative for the enteric agents, suggests that inhibition of certain high-risk behaviors may be a key component of the observed effect of religion on health. The timing of data collection for this study (1988-1992) may explain the otherwise odd absence of HIV, which at the time (pre-HAART) was treated as an acute, or at least rapidly progressing, infection. However, the stigma attached to HIV infection, and the frequency of HIV co-infection with other viral and bacterial pathogens, leaves the reader of the present study uneasy as to the possible unmeasured effect of HIV on its findings.
The study's focus on viral agents is understandable. They are generally persistent and allow for stable measurement of population prevalence. Because they are less easily cleared from the system than are bacterial pathogens, they also are more urgent targets for effective prevention strategies. Presumably, the inhibiting effect of religious belief on behavior means that exposure to all agents is reduced equally, but this cannot be inferred from the present study.
1. Strawbridge WJ, et al. Frequent attendance at religious services and mortality over 28 years. Am J Public Health 1997;87:957-961.
2. Underwood LG, Teresi J. The daily spiritual experience scale: Development, theoretical description, reliability, exploratory factor analysis, and preliminary construct validity using health-related data. Ann Behav Med 2002;24:22-33.
3. Koenig HG, et al. Religion index for psychiatric research. Am J Psychiatry 1997;154:885-886.
4. Fetzer Institute and National Institute on Aging working group. Multidimensional Measurement of Religiousness/Spirituality for Use in Health Research. Kalamazoo, MI: Fetzer Institute; 1999.