ObjectivesTo estimate the effect of student status on mental health condition during COVID-19 general lockdown in France.DesignCross-sectional analysis comparing students and non-students recruited in the same study.SettingParticipants of the web-based CONFINS study implemented during the general lockdown in France in spring 2020.Participants2260 participants (78% women) including 1335 students (59%).Primary and secondary outcome measuresThrough an online questionnaire, participants declared if they have experienced suicidal thoughts, coded their perceived stress on a 10-points scale and completed validated mental health scales (Patient Health Questionnaire-9 for depressive symptoms, Generalised Anxiety Disorder-7 for anxiety symptoms) during the last 7 days. The effect of college student status on each mental health condition was estimated using multivariate logistic regression analyses. Stratified models for students and non-students were performed to identify population-specific factors.ResultsStudent status was associated with a higher frequency of depressive symptoms (adjusted OR (aOR)=1.58; 95% CI 1.17 to 2.14), anxiety symptoms (aOR=1.51; 95% CI 1.10 to 2.07), perceived stress (n=1919, aOR=1.70, 95% CI 1.26 to 2.29) and suicidal thoughts (n=1919, aOR=1.57, 95% CI 0.97 to 2.53). Lockdown conditions that could be potentially aggravating on mental health like isolation had a higher impact on students than on non-students.ConclusionsCollege students were at higher risk of mental health disturbances during lockdown than non-students, even after taking into account several potential confounding factors. A close follow-up and monitoring of students’ mental health status is warranted during lockdown periods in this vulnerable population.

Background Due to their impact on premature mortality and long-term disabilities, a better understanding of health risk behavior (HRB) determinants among college students is crucial in order to build the most appropriate prevention tools. Although selfesteem appears to be a relevant candidate, a clear picture summarizing its multiple links with HRB is lacking to guide clinicians and researchers. This study aims to provide a systematic review of the associations between health risk behavior and self-esteem among college students. Methods This search was performed in several databases on 02/02/17. Study eligibility criteria were original articles in peerreviewed journals, in English; observational quantitative studies; among college students; and investigated the association between self-esteem and HRB. The PRISMA statements were complied with. Results One hundred fifteen articles were included: 46 on substance use, 35 on sexual behavior, 11 on nutritional habits, 27 on physical activity, and 5 on other HRB. Most studies reported an association between higher self-esteem and healthier behavior. For alcohol consumption and number of sexual partners, both negative and positive associations have been reported. Directionality was investigated in a few studies suggesting potential bidirectional effects. Conclusions This review points out the need for consensus for the definition of self-esteem and HRB. There was high heterogeneity in (1) the measurement of self-esteem either in the concept measured (global vs. domain) or in the way to implement validated tools; and (2) the definition of HRB. Self-esteem seems to be a relevant target to intervene on HRB, especially alcohol abuse and physical activity. Trial Registration Registration number: PROSPERO (ID = CRD42017056599).

No abstract