1. Sociodemographic Characteristics of Study Subjects
<Table 2 > shows the results of frequency analysis to identify the socio-demographic characteristics of the study subjects. In terms of gender, more than half of the participants were women (57.2%) as compared to men (42.8%). 19.1% of the participants were in their 70s or older followed by 968 (18.6%) in their 50s. Most of the participants were married (83.8%). The highest academic background of the respondents was high schools (26.5%), followed by four-year universities (25.2%). In the family history of participants, 36.2% has hypertension, hyperlipidemia (6.3%), ischemic heart disease (5.6%), strokes (11%) and diabetes (18.4%).
2. Variations of Metabolic Syndrome by Sociodemographic Characteristics
A result of the independent sample t-test analysis for metabolic syndrome showed statistically significant differences in gender, marital status, and family history (high blood pressure, hyperlipidemia, ischemic heart disease, stroke, diabetes). Gender was higher for men than for women, marital status was higher for unmarried people than for married people, and in the family history, hypertension, hyperlipidemia, ischemic heart disease, stroke, and diabetes "NO" were higher than"YES". As a result of conducting a one-way distribution analysis (ANOVA), metabolic syndrome showed statistically significant differences in age, the number of families, and degree of education.
As a result of Duncan'spost-analysis, metabolic syndrome was the highest among those in their 30s ofage, followed by those in their 70s and older, 60s, and 50s, and those in their40s and 30s. The number of families was high in the order of four or more, three, two, and one, and the degree of education was the highest, followed by middle school, elementary school, high school, two-thirds-year university, and four-year university. <Table 3>
3. Correlation between study variables
Pearson's correlation analysis was conducted to confirm the correlation between the number of eating out, metabolic syndrome, and health behavior, which are the main variables of this study, and the analysis results are as shown in < Table 4>.
As a result, the number of times to eat out (r = .096, p < .01) was positively correlated with metabolic syndrome, and stress recognition rate (r=-.361, p < .01) and smoking (r=-.253, p < .01) and drinking (r=-.091, p < .01) had a negative correlation.<Table 4>
4. Effects of the Number of Eating Out on Metabolic Syndrome
Multiple linear regression analysis was performed to verify the effect of the number of eating out on metabolic syndrome, and demographic characteristics were controlled. As a result, the regression model was statistically significant (F = 286.370, p<).001), the explanatory power of the regression model was about 42.5% (adjR2=.425).
On the other hand, the Durbin-Watson statistic was 1.857, which was close to 2, and was evaluated as having no problem with the assumption of the independence of the residuals.
As a result of verifying the significance of the regression coefficient, the number of times to eat out had a significant positive (+) effect on metabolic syndrome (β = .036, p < .001). In other words, it was evaluated that the higher the number of eating out, the higher the metabolic syndrome.<Table 5>
5. The moderating effect of smoking on the effect of eating out on metabolic syndrome
Looking at the moderating effect of smoking, it was found that the interaction effect between the number of eating out and smoking did not have a statistically significant effect on metabolic syndrome(B=-.0154, p > .05). In other words, it was evaluated that smoking had no effect on the relationship in which metabolic syndrome increased as the number of eating out increased < Table 6>
6. The moderating effect of drinking on the effect of eating out on metabolic syndrome
According to the moderating effect of drinking, the interaction effect between the number of eating out and drinking was evaluated to have a negative (-) effect on metabolic syndrome (B=-.0358, p<).001) In other words, it was evaluated that drinking decreased its effect in the relationship where metabolic syndrome increases as the number of eating out increases.<Table 7>
7. The moderating effect of Stress recognition rate on the effect of eating out on metabolic syndrome
Looking at the moderating effect of the stress recognition rate, it can be seen that the interaction effect between the number of eating out and the stress recognition rate has a negative (-) effect on metabolic syndrome (B=-.0149, p < .001). In other words, it was evaluated that the stress recognition rate further reduces the effect in the relationship where metabolic syndrome increases as the number of eating out increases. <Table 8>