Exploring the Association of Fear, Mental Health, and Lifestyle with Insomnia: A Machine Learning Approach among University Students.
4 affiliations
- 1Department of Pediatrics, Student Research Committee, Birjand University of Medical Sciences, BirjandIran.
- 2Department of Anatomy, Birjand University of Medical Sciences, BirjandIran.
- 3Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Birjand, BirjandIran.
- 4Department of Family and Community Medicine, School of Medicine, Cardiovascular Diseases Research Center, Birjand University of Medical Sciences, BirjandIran.
Abstract
The aim of this study was to identify key predictors of insomnia among university students, focusing on psychological, behavioral, and lifestyle factors. The study also aimed to explore the role of machine learning in identifying insomnia-related patterns and informing personalized intervention strategies. A cross-sectional study was conducted with 576 students from universities in Birjand during the 2023-2024 academic year. Machine learning methods, including Random Forest, Decision Tree, and SHAP analysis, were used for feature importance analysis and knowledge extraction. Self-reported questionnaires revealed that fear of not falling asleep particularly in the late hours and stress were the most significant predictors of insomnia. Anxiety, poor sleep hygiene, and depression were also strongly associated with insomnia. Other associated factors included financial and family concerns, bedtime, wake-up time, sleep duration, and social media usage. Machine learning models further highlighted these factors, with those showing poor sleep hygiene and high anxiety levels being more likely to experience insomnia. The findings underscore the complex relationship between mental health issues and sleep disorders, particularly insomnia, in university students. The results also suggest a bidirectional relationship between insomnia, anxiety, depression, and stress. The use of machine learning showed potential for identifying insomnia-related patterns and informing personalized intervention strategies. These results highlight the need for effective strategies to address insomnia in students, particularly in light of the growing stressors faced in academic and personal life. However, due to the cross-sectional design, causal inferences cannot be drawn.
Related
- Problematic Social Media Use and Insomnia Severity Among Saudi University Students: Cross-Sectional Indirect-Association Analyses Involving Fatigue and Anxiety
- Exploring the Association Between Parenting Styles, Dental Anxiety, and Dental Caries Experience Among Children Attending a University Dental Clinic: A Cross-Sectional Study.
- Do machine learning models really encode cross-cultural lifestyle-health differences? A leakage-free reanalysis of two university cohorts in India and Japan
- Exploring the association between Zhejiang University index and osteoporosis risk in middle-aged and older adults based on the National Health and Nutrition Examination Survey database
- Exploring the structural relationships between soft skills, emotional intelligence, cognitive flexibility, and self-directed learning in the digital age: A cross-sectional PLS-SEM study of female students of king Faisal University, Saudi Arabia
BibTeX
@article{khademi2026exploring,
title = {{Exploring the Association of Fear, Mental Health, and Lifestyle with Insomnia: A Machine Learning Approach among University Students.}},
author = {Sara Khademi and Mohammad Reza Saebipour and Hamid Saadatfar and Seyed Mohammad Riahi},
journal = {Sleep science (Sao Paulo, Brazil)},
year = {2026},
month = sep,
doi = {10.1055/s-0046-1829077},
url = {https://europepmc.org/article/MED/42997738},
note = {CC BY},
}
RIS
TY - JOUR
TI - Exploring the Association of Fear, Mental Health, and Lifestyle with Insomnia: A Machine Learning Approach among University Students.
AU - Sara Khademi
AU - Mohammad Reza Saebipour
AU - Hamid Saadatfar
AU - Seyed Mohammad Riahi
PY - 2026
DA - 2026/09/01
JO - Sleep science (Sao Paulo, Brazil)
DO - 10.1055/s-0046-1829077
UR - https://europepmc.org/article/MED/42997738
AB - The aim of this study was to identify key predictors of insomnia among university students, focusing on psychological, behavioral, and lifestyle factors. The study also aimed to explore the role of machine learning in identifying insomnia-related patterns and informing personalized intervention strategies. A cross-sectional study was conducted with 576 students from universities in Birjand during the 2023-2024 academic year. Machine learning methods, including Random Forest, Decision Tree, and SHAP analysis, were used for feature importance analysis and knowledge extraction. Self-reported questionnaires revealed that fear of not falling asleep particularly in the late hours and stress were the most significant predictors of insomnia. Anxiety, poor sleep hygiene, and depression were also strongly associated with insomnia. Other associated factors included financial and family concerns, bedtime, wake-up time, sleep duration, and social media usage. Machine learning models further highlighted these factors, with those showing poor sleep hygiene and high anxiety levels being more likely to experience insomnia. The findings underscore the complex relationship between mental health issues and sleep disorders, particularly insomnia, in university students. The results also suggest a bidirectional relationship between insomnia, anxiety, depression, and stress. The use of machine learning showed potential for identifying insomnia-related patterns and informing personalized intervention strategies. These results highlight the need for effective strategies to address insomnia in students, particularly in light of the growing stressors faced in academic and personal life. However, due to the cross-sectional design, causal inferences cannot be drawn.
SN - 1984-0659
SN - 1984-0063
ER -