Insomnia, Daytime Sleepiness, and Quality of Life among 20,139 College Students in 60 Countries around the World-A 2016-2021 Study

The study addresses the growing global burden of sleep disorders among young adults, particularly college students, who are susceptible due to lifestyle changes (relocation, independence, stimulant use) and poor adherence to sleep hygiene. Insomnia (difficulty initiating or maintaining sleep) and daytime sleepiness (difficulty maintaining wakefulness) are closely related and have increased in prevalence, with the COVID-19 pandemic further disrupting sleep patterns through lockdowns, remote learning, and reduced social interaction. Sleep problems adversely affect physical and mental health, academic/work performance, and are linked bidirectionally with substance use. Quality of life (QoL) is multifactorial and increasingly used in health assessment but requires standardized tools. Evidence on socioeconomic influences (e.g., HDI, GDP per capita) on sleep is mixed and not standardized. Research gaps include global, cross-cultural, multi-year assessments spanning pre- and during-pandemic periods. The study aims to estimate global prevalence of insomnia and daytime sleepiness among students, examine associations with QoL, substances (alcohol, cannabinoids, psychostimulants, sedatives/hypnotics), and compare outcomes by socioeconomic context (HDI, GDP per capita) and pandemic period.

Prior studies report high and variable student insomnia prevalence by region and methodology: e.g., South Asia ~52% (range 35–70%), Ethiopia ~62%, Sudan up to 82.5%, and lower estimates in some higher-income settings (e.g., Germany, Poland, Norway, China). Socioeconomic disadvantage is linked to poorer sleep in both instrumented and self-report studies, and large US and Korean datasets associate lower income/education with worse sleep. COVID-19 research shows widespread sleep disruption among students (delayed bed/wake times, more naps, increased duration but poorer quality) and deteriorated mental health, both affecting sleep. Medical students often show high sleep problems and daytime sleepiness due to academic load and night duties, though findings vary. Substance use (alcohol, cannabinoids, psychostimulants) has complex bidirectional relationships with sleep; psychostimulants degrade sleep architecture and quality. Poor sleep is consistently associated with reduced QoL in student and population-based samples, with potential moderation by mental health status; higher national development relates to better QoL.

Design: Cross-sectional, global, online CAWI survey distributed via Facebook student groups using convenience sampling. Period: 31 January 2016 to 30 April 2021; respondents categorized as pre-pandemic (until 11 March 2020) and during-pandemic (from 11 March 2020) based on WHO declaration. Participants: Inclusion—college students ≥18 years providing informed consent; Exclusion—non-students, non-consent, age <18. Student status was self-confirmed; survey was anonymous and voluntary. Ethics approval: Bioethics Committee of Wroclaw Medical University (KB-234/2021); conducted per Declaration of Helsinki. Measures: - Sociodemographics: age, sex, country of residence, year of study, university profile (medical/non-medical). - Substance use (past 3 months): alcohol, cannabinoids, psychostimulants, sedatives/hypnotic drugs. - Insomnia: Athens Insomnia Scale (AIS; 8 items, 0–3 Likert; total 0–24; cutoff ≥6 for insomnia; sensitivity 93%, specificity 85%; Cronbach’s α=0.827). - Daytime sleepiness: Epworth Sleepiness Scale (ESS; 8 items, total 0–24; cutoff ≥11 for excessive daytime sleepiness; ≥16 suggests severe; reliability α≈0.742). - Quality of life: Manchester Short Assessment of Quality of Life (MANSA; 16 items; higher scores indicate better QoL; internal consistency α=0.764). Socioeconomic classification: Countries categorized by Human Development Index (HDI: very high, high, medium, low; UNDP) and by World Bank GDP per capita groups (high, upper-middle, lower-middle, low income). Statistical analysis: Descriptive statistics for quantitative variables; chi-square tests for categorical comparisons by pandemic period; t-test for age; Kendall’s tau for item–total correlations; backward stepwise logistic regression for insomnia (AIS≥6) and for daytime sleepiness (ESS≥11) with predictors: age, sex, year, university profile, pandemic period, HDI, GDP per capita, and substances. Backward stepwise linear regression for QoL (MANSA) with predictors above plus insomnia and daytime sleepiness. Significance p<0.05. Software: Statistica 13.3.

Sample: 20,139 college students from 60 countries; mean age 22.6±3.6 years; 78.2% women; 77.6% non-medical; 27.3% first-year; 90.4% from very high HDI countries; 87.9% from high-income countries; 50.6% completed pre-pandemic. Alcohol use in past 3 months: 82%. Insomnia: Mean AIS 8.26±4.35; 11,597 students (57.6%) screened positive (AIS≥6). High proportions reported insufficient sleep duration (73.6%) and unsatisfactory sleep quality (71.2%); difficulty falling asleep (78.9%) and nocturnal awakenings (78.5%) were common. Sleep quality item correlated most with total AIS (r=−0.486, p<0.001; duration r=0.417, p<0.001). Daytime sleepiness: Mean ESS 7.90±4.23; 5,442 (27.0%) had ESS≥11. Sleep propensity was greatest during afternoon rest (90.4%), on the bus (73.4%), and watching TV (72.5%); 10.4% reported some likelihood of dozing while driving. Risk factors (logistic regression): - Pandemic period increased insomnia risk (OR 2.17 [2.05–2.31], p<0.001) but decreased daytime sleepiness risk (OR 0.78 [0.74–0.84], p<0.001). - Female sex: higher odds for insomnia (OR 1.25 [1.67, 1.35], p<0.001; as reported) and for daytime sleepiness (OR 1.44 [1.33–1.56], p<0.001). - Year of study: higher years associated with lower odds of both insomnia and daytime sleepiness (e.g., Year V insomnia OR 0.82 [0.73–0.91], p<0.001; ESS OR 0.77 [0.68–0.87], p<0.001). - University profile: medical students had lower insomnia risk (OR 0.90 [0.84–0.97], p=0.004) but higher daytime sleepiness risk (OR 1.20 [1.12–1.29], p<0.001). - Socioeconomics: High GDP per capita was protective against insomnia (OR 0.41 [0.24–0.69], p<0.001) and daytime sleepiness (OR 0.74 [0.59–0.93], p=0.008). HDI differences were not consistently significant; medium HDI showed higher insomnia odds (OR 1.70 [1.04–2.76], p=0.033). - Substances: Sedatives/hypnotics associated with higher odds of insomnia (OR 2.60 [2.36–2.87], p<0.001) and daytime sleepiness (OR 1.27 [1.16–1.39], p<0.001). Psychostimulants increased insomnia (OR 1.51 [0.63–1.89], p<0.001) and daytime sleepiness (OR 1.14 [1.04–1.26], p=0.003). Cannabinoid use slightly increased insomnia odds (OR 1.10 [1.00–1.20], p=0.034). Alcohol showed no independent effect in the final models. Quality of life: Mean MANSA 60.9±11.46. QoL scores were higher during the pandemic (p<0.001). Very high HDI residents reported higher QoL (B=2.001, p=0.002); no significant differences by GDP per capita. Hypnotic drugs and psychostimulant use were negatively associated with QoL. Both insomnia (B≈−3.142, p<0.001) and daytime sleepiness (B≈−1.331, p<0.001) were associated with lower QoL. Correlations: Higher AIS and ESS scores correlated with lower QoL (rAIS=−0.355; rESS=−0.155; both p<0.001); AIS and ESS were positively correlated (r=0.153, p<0.001).

The study demonstrates a high global prevalence of insomnia (58%) and excessive daytime sleepiness (27%) among college students and confirms their interrelationship. Findings align with prior evidence that pandemic-related disruptions to daily routines increased insomnia while altering daytime sleepiness patterns. Socioeconomic context matters: residence in high GDP per capita countries reduced risks of insomnia and daytime sleepiness, likely reflecting differences in living conditions, access to healthcare, stressors, and health behaviors (including sleep hygiene). Academic factors were relevant: non-medical students were more insomnia-prone, whereas medical students had higher daytime sleepiness, plausibly due to workload and clinical duties. Substance use, particularly sedatives/hypnotics and psychostimulants, was associated with worse sleep and lower QoL, underscoring the bidirectional relationship between sleep and psychoactive substances. Importantly, both insomnia and daytime sleepiness were linked to substantially lower QoL across multiple life domains, highlighting sleep’s centrality to student well-being. Together, these results address the research aims by quantifying global prevalence, identifying modifiable and contextual risk factors, and demonstrating significant impacts on QoL across diverse socioeconomic settings spanning pre- and during-pandemic periods.

Insomnia and excessive daytime sleepiness are common and interrelated among students worldwide. The COVID-19 pandemic significantly disrupted diurnal rhythms, contributing to increased insomnia. Students in high GDP per capita countries had lower risks of sleep disorders. Sleep problems substantially reduce students’ quality of life. Interventions should promote sleep hygiene and educate about the impacts of stimulant and sedative use, and further representative, cross-country research—including meta-analyses—is warranted.

Convenience sampling via anonymous online survey introduces selection bias and prevents verification of responses and response rate estimation; potential for bots could not be excluded. Reliance on self-report psychometric instruments may not capture clinical diagnoses; no psychiatric or clinical validation was performed. Data on chronic physical/mental conditions and chronic medications were not collected, limiting confounding control. Sample imbalances existed across HDI/GDP categories and sex (predominantly women). Despite limitations, strengths include a very large, multinational sample and coverage of both pre- and during-pandemic periods.