Behavioral risk factor clusters among university students at nine universities in Libya

• Main Authors: Walid El Ansari, Khalid A Khalil, Derrick Ssewanyana, Christiane Stock
• Format: Article
• Language: English
• Published: AIMS Press 2018-08-01
• Series: AIMS Public Health
• Subjects: university students| gender| risk factors| health behaviors| cluster analysis
• Online Access: http://www.aimspress.com/article/10.3934/publichealth.2018.3.296/fulltext.html

<em>Objectives:</em> This study identifies and describes the clustering of 5 behavioral risk factors (BRFs) among university students. We also investigated whether cluster membership is associated with the students’ self-rated academic performance and self-rated health. <em>Material and methods:</em> A sample of 1300 undergraduates at 6 universities and 3 colleges in Libya completed a self-administered questionnaire that assessed BRFs (nutrition, physical activity, alcohol consumption, smoking, illicit drug use, inadequate sleep). A two-step cluster analysis generated student clusters with similar lifestyles. <em>Results:</em> Two contrasting clusters of almost even size emerged (after exclusion of alcohol and illicit drug use due to very low prevalence). Cluster 1 comprised students with higher engagement in all forms of physical activity, higher levels of health consciousness, greater daily fruit/vegetable intake and better sleep patterns than students in cluster 2. Only as regards the consumption of sweets, cluster 1 students had less favorable practices than cluster 2 students. The prevalence of smoking was equally low in both clusters. Students in cluster 2, depicting a less healthy lifestyle, were characterized by a higher proportion of women, of students with less income and of higher years of study. Belonging to cluster 2 was associated with lower self-rated health (OR: 0.46, <em>p</em> &lt; 0.001) and with lower self-rated academic performance (OR: 0.66, <em>p</em> &lt; 0.001). <em>Conclusion:</em> Preventive programs should not address BRFs in isolation and should particularly target students with clustering of BRFs using specifically tailored approaches.