The large sample of different academic curricula allowed us to highlight the effect of the COVID-19 outbreak and isolation measures on university students. We identified concerns about changes in teaching methods, not validating the academic year and increased workload. Specifically, the concern about not validating the academic year was associated with unfavorable changes in tobacco smoking and the concern about changes in teaching methods was associated with favorable changes in vigorous physical activity. To guarantee the effectiveness of online learning, the design principles of digital learning materials, learning goals and students’ preferences and characteristics should be rigorously evaluated [25]. Anxiety due to online learning needs to be relieved to ensure that students can actively and effectively engage in this online learning [26]. In the context of the COVID-19 outbreak, interactive pedagogic tools could be useful for education continuity and for maintaining human contact necessary in pedagogy [4]. Future strategies could combine on-site teaching with online courses and consider the role of social contact in students’ mental health and offer starting points to identify and support students at a higher risk of social isolation and negative psychological effects during the COVID-19 pandemic [12]. The current issue is to safely reopen universities with a combination of strategies that include containment (access control with contact tracing and quarantine) and mitigation (hygiene, sanitation, ventilation, and social distancing) practices [27]. A systematic review showed that medical students undergoing appropriate training could play an essential role in pandemic management and suggests a course and assessment structure for medical student COVID-19 training [28].

In this study, one third of students did not live with their parents and one half had less contact with friends. Bu et al. reported that being a student emerged as a higher risk factor for loneliness during lockdown than usual [29]. Students were worried that their relatives were infected and that medical supplies were lacking as also reported specifically in healthcare students [30]. Studies have suggested that preparation for healthcare students should be provided prior to a pandemic to build resilience, thereby reducing the impact of stress after exposure [31,32].

Unhealthy behaviors (tobacco smoking, binge drinking and cannabis use) significantly decreased during the COVID-19 period especially binge drinking as reported by Busse et al. among German students (ISWS 19 study) [23]. In the general population, studies reported that COVID-19 measures led to increased tobacco smoking [19,33] and that being a student was a protective factor for increased tobacco smoking [19]. The major proportion of favorable changes was for binge drinking, almost one third of students decreased their frequency of binge drinking. This decrease was also reported among Canadian adolescents during the COVID-19 outbreak [21]. Our results, showing that students who returned home to live with their parents during lockdown reduced their practices of binge drinking compared to students living in rented accommodation, support the non-epidemic study [34]. Binge drinking among students revolves around socializing (e.g., partying, and belonging to a social network of heavy drinkers) which may explain why this effect is not found in our study for tobacco and cannabis [34].This decrease of binge drinking was not reported in the general population in the 18–24 year age group [35]. Fear of being infected with COVID-19 was negatively associated with a decrease in binge drinking. Dumas also reported a fear of the infectivity of COVID-19, predicted using solitary substance use during the pandemic [21] and Nguyen et al. found that students with higher fear scores more likely had unhealthy lifestyles, such as smoking and drinking alcohol [36]. In our study, cannabis use decreased both in number of users and frequency. Daumas et al. reported among adolescents a decrease in cannabis use but with an increased frequency [21]. Among the general population, Rolland et al. showed similar rates between an increased use and a decreased use of cannabis and identified that being a student compared to a worker was a protective factor for increased cannabis consumption [32]. We observed that for binge drinking and cannabis use, men had behavioral changes, either better or worse. This can be explained by their mode of consuming before confinement, alone or in a group [21] and by the fact that the men perceived more drinking norms and influence of their peers than women and were therefore more sensitive to the decline of these influences during the lockdown [37].

Concerning positive health behaviors, as physical activity, there were more students with a decrease in physical activity (one third of students) than an increase (one quarter of students) as also reported in studies among students [17,18], and the general population [16]. The role of modifiable lifestyle factors like physical activity in maintaining health and wellness are fundamental [38]. The benefits of empowering students to actively preserve their own health should be underlined [8]. Epidemiological evidence has demonstrated a dose-response relationship between the physical activity performed before infection and a reduction in the incidence, duration, or severity of acute respiratory tract infections [39,40]. The decrease in physical activity due to mandatory stay-at-home is perhaps one of the most apparent consequences of complete lockdown, not only for active individuals habitually practicing recreational sports, but also for those who go to work by walking or cycling. These observations have potential implications that could aid the development of physical activity and nutritional recommendations to maintain health during the ongoing COVID-19 pandemic [15]. Universities have been identified as appropriate settings for the implementation of lifestyle interventions [41].

Depression was associated with an increase in the three unhealthy behaviors and a decrease in physical activity, as reported among adolescent and general populations [19,35]. De Man et al. highlighted in the Belgium C19 ISWS study that the duration of the exposure showed to have an effect on depressive symptoms severity, which was mediated by “academic stress” [42]. A possible mitigation strategy for improving mental health should include taking suitable amounts of daily physical activity and sleeping well [43]. Universities should provide psychological services oriented and adapted to these circumstances to mitigate the emotional impact on university members.

Caution is advised when generalizing these findings, for the following reasons: First, examining changes in health behaviors in the same survey with retrospective measurement of past behavior (before COVID-19 period) may lead to errors in self-reporting; second, it was a convenience sample, voluntary participation could have led to representativeness and self-selection bias as our sample had more women, also reported by an Italian study [18].