Assessing the influence of religion on health behavior

Abstract

A primary aim of this study was to confirm the factor structure of the Health and Religious Congruency Scale (HARCS), a measure previously developed by the same research team. The HARCS questions directly link religious beliefs/activities to health behaviors. Confirmatory factor analysis (CFA) showed that the current data fit poorly to the factor structure found in the pilot study. Because the current sample was more religiously diverse than the pilot study sample, and could potentially provide a factor structure that better reflects the views of individuals from different religious affiliations, a principal components analysis was conducted on the current data. CFA was then performed on combined data from the pilot study and current investigation. The resulting factor structure had acceptable Goodness of Fit Indices. After eliminating one scale because of highly skewed data and limited utility, and two other scales because of poor test-retest reliability, validity tests were performed on the four remaining HARCS subscales and the total score of the HARCS. Subscale 1 is related to the general influence of religion on drinking and smoking. Hierarchical regression showed that religious variables, drinking and smoking behaviors, age, and an interaction between religion and drinking/smoking accounted for approximately 50% of the variance in the subscale. Subscales 2-4 were related to the impact of religion on eating, physical activity, and weight. Religious variables, health behaviors, and age accounted for small amounts of variance in these scales. In subscales 2-4 few participants endorsed that religion impacted the health behaviors of interest. Overall, results provide further support that from a religious perspective drinking and smoking behaviors are viewed differently than eating and physical activity. The four subscales have adequate reliability, however, the three subscales pertaining to exercise, eating, and weight appear to have little relevancy to the general population. In contrast, subscale 1 appears to have utility with individuals of various religious orientations. For this reason, future studies should consider using subscale 1 without the other subscales. Future studies will be required to determine if the other subscales are valid and useful among select groups such as those participating in faith-based health programs.