| OCR Text |
Show and social-network variables than does BRFSS. Therefore, UHSS enables researchers to analyze physical activity in a broad social context. However, the UHSS' physical activity measure is not used for official prevalence estimates by the Utah Department of Health. In Study IV, the effects of family or household were measured by whether or not any other adult member of the household participated in regular vigorous exercise. For the analysis of multiple roles' effect, a combined measure of an individual's marital, employment, and parental status is created with eight categories: (a) married, full-time employed, being a parent; (b) married, full-time, not a parent; (c) married, not full-time, a parent; (d) married, not full-time, not a parent; (e) not married, full-time, a parent; (f) not married, full-time, not a parent; (g) not married, not full-time, a parent; and (h) not married, not full-time, and not a parent. The first category represents having three time-demanding roles/obligations; whereas the last category is the comparison group for this analysis, indicating those without any of these roles/obligations. Religious network is measured by religious affiliation and involvement. Since the Church of Jesus Christ of Latter-Day Saints (LDS) is the predominant religion in Utah and approximately 68% of the respondents were self-reported LDS members in this analysis, religious affiliation was coded as LDS, other religions, or no religion. Respondents, who reported religious affiliation and attended services once a week or more, were considered "active," whereas respondents reporting an affiliation but less attendance were considered "less active." The variable of religious affiliation and involvement had four categories: (a) active LDS, (b) less active LDS, (c) active other religion, and (d) less active other religion or no religion. Health insurance status was measured by a dichotomous variable. In order to control the effect of gender and education, all analyses were conducted within each of the six gender-education subgroups. The results of bivariate analysis are reported in Table 2, that is, the number and percentage of adults who engaged in regular vigorous exercise by family and social network variables according to gender-education subgroups. The results of multivariate analysis (odds ratios from a multiple logistic analysis, conducted by the statistical software SAS and SUDAAN) are presented in Table 3. The important findings from the logistic regression analysis are highlighted below. Effects of Family or Household There was a significant positive effect on an individual's exercise behavior regardless of gender or education when another adult member of the household participated in vigorous exercise. Individuals living in a household with adults who did not engage in vigorous exercise had approximately an one-fifth to one-third likelihood of participating in vigorous exercise themselves compared to those who lived with physically active members. Effects of Multiple Roles The analysis demonstrated significant opposite effects on regular vigorous exercise by multiple roles for men and women at the highest education level. Among college-educated women, those who were married, full-time employed, and not a parent were less likely to engage in physical activity than those who were not married, not full-time employed, and not a parent. However, male college graduates who were married, full-time employed, and not a parent were five times more likely to engage in vigorous exercise than those male college graduates without those obligations. Different results for men and women were also observed at the lowest education level. Multiple roles/obligations had no significant effect on engaging in regular vigorous exercise among women reporting high school or less education; whereas men with the same level of education, being full-time employed regardless of their marital and parental status, were significantly less likely to engage in regular vigorous exercise than their comparison group. These findings indicate an interactive effect of gender, education, and multiple roles on physical activity. Effects of Religious Affiliation and Involvement Approximately 49% of LDS members, age 18 to 64, engaged in regular vigorous exercise at least three times per week; so did 56% of non-LDS or nonreligious persons in the same age group. (The data are not reported in Table 2.) Among male college graduates, active LDS persons were significantly less likely to engage in vigorous exercise than all other groups. Among women with some college education, there was no difference in participation in vigorous exercise between LDS groups. However, women of other or no religion with some college education were more likely to participate in vigorous exercise than their LDS counterparts. The religious effect was not statistically significant with regard to the likelihood of vigorous exercise among men with 39 |
