Health-risk Behaviors And Predictive Value Of Psychosocial Factors In College Students

Objectives The study aims to describe health-risk behaviors and cooccurrences in college students, and identify predictive value of coping strategy, social support and parental rearing style in these behaviors.Methods A questionnaire survey as baseline was conducted in June 2008 among freshman and sophomore from a medical university. They accomplished the closed-type questionaire intensively. Socio-demographic characteristics, self-perceived family income and learning performance, health-risk behaviors, including physical inactivity, unhealthy weight-control behaviors, binge eating, current smoking and alcohol use, heavy drinking, binge drinking, pathological internet use were collected by structured questionnaire. The participants were also asked to complete EMBU (Egma Minnen av Barndoms Uppfostran-My memories of upbringing), TAS-20 (Toronto Alexithymia Scale), GSES (General Self-Efficacy Scale), PSSS (Perceived Social Support Scale). Of the baseline students, a part of students were selected by cluster sampling as the cohort study participants.They were asked to report informations of physical inactivity, binge eating, current smoking and alcohol use, heavy drinking, binge drinking and pathological internet use every six months in 1.5 years.Then the variable patterns of these behaviors in the follow-up surveys were examined. These patterns were classified as: There was no health-risk behavior in either baseline or follow-up (pattern 1). There was no health-risk behavior in baseline survey but was at least one behavior in follow-up (pattern 2). There was health-risk behavior in baseline survey but was not in follow-up (pattern 3). There existed at least one health-risk behavior in both baseline and follow-up (pattern 4). Chi-square test, spearman correlation analysis, factor analysis, bivariate logistic regression model, multivariate logistic regression model, and one-way ANOVAs were used for statistical analysis. Results In the baseline survey, A questionaire survey was conducted among 3 777 students, a total of 3 718 students completed the questionnaire. Of the participants, 1 761 were male students, and 1 957 were female students. The mean age of the participants was 19.8 years. The prevalence rate of health-risk behaviors (at least one) was 84.1% in male students and 83.3% in female students. There existed significant differences in rates of the 8 health-risk behaviors and cooccurrences betweent different genders, grades majors and among different recruit levels. The 8 health-risk behaviors significantly related to each others in both genders. Bivariate and multivariate logistic regression model analyses show that the maternal rearing style, social support, general self-efficacy and defeature affection were the important factors of health-risk behaviors in college students.In the cohort study, 998 students were selected, the response rates of 3 follow-up investigations were 90.5%, 90%, and 90.4%. Only rate of current alcohol use and heavy drinking decreased significantly (P<0.01). In addition, rates of binge eating in both genders, current alcohol use in male students and physical inactivity in female students were at high levels in 4 surveys. The score of maternal excessive interference in the pattern 1 was the lowest while the score of social support was the highest.Conclusions There was a high prevalence of health-risk behaviors among medical college students, and they were exposed to two or more risk behaviors simultaneously. The distribution and cooccurrence characteristic of health-risk behaviors were distinct in different genders, grades, majors, and recruit levels. It was the low maternal excessive interference and high social support that proved to be the significant protective factors of health-risk behaviors among undergraduate students.