Stress is a term frequently used in a variety of social, academic and employment settings. Everyone needs a certain amount of pressure to perform at their best. However, when pressure exceeds a person’s ability to cope, this results in stress. Moreover, stress can set up a cycle of distress and reduce ability to cope even in ordinary situations. In today’s ultra competitive environment, students face more stress than ever, be it related to studying, examinations, or peer, teacher or parental pressure 1 . University students, often experience undue amounts of stress, which can have negative academic, emotional and health outcomes 2 . They face multiple stressors such as academic overload, constant pressure to succeed, competition with peers and in some countries financial burden as well as concerns about the future 3 . As all this may lead to psychopathology, the health of university students, especially healthcare students, has been the subject of increasing focus in recent years 4 5 . Voltmer et al. reported a decrease in quality of health and an increase in risk patterns, indicating a need for prevention and health promotion focussing on individual behavior 6 .

In addition, common stressful life events are associated with both mental health symptoms and substance use in young adolescents 7 . Therefore, stress should not be considered on its own, but rather should be associated with potential risk behaviors. Moreover, the risk of onset of substance use and related problems is heightened during the university period. The most common substances used by young adults are alcohol, tobacco and cannabis 8 9 10 . The use of two or more of these substances simultaneously, known as polydrug use, has received growing attention in the literature due to an increase in prevalence in early adulthood 11 . Onset of smoking and alcohol drinking during early adulthood is a well-documented and significant public health risk factor, and is linked to a high risk of chronic diseases at older age 12 13 14 . Furthermore, substance use is associated with immediate health problems such as academic difficulties, injuries, interpersonal violence 15 , high-risk sexual behavior 16 , depression, and mental disorders 17 . Alcohol misuse in students has led to specific behavior known as binge drinking 18 19 20 .

Stress in young adults is also often associated with the occurrence of substance-free addictions (also called behavioral addictions), such as cyberaddiction, addiction to sex, workaholism, and addiction to shopping 21 22 23 . Problematic internet use and addiction is characterized by excessive urges or behaviors regarding internet use, leading to impairment or distress among adolescents and young adults 24 . Risk behavior and associated stress mainly related to tobacco, alcohol and cannabis use 7 25 , have already been widely reported in a population of healthcare students 1 2 26 27 . However, eating disorders and risk of cyber addiction have never been associated with stress in university students. In especially stressful situations, students often have an inaccurate perception of body weight, which can trigger abnormal eating behavior and eating disorders in predisposed individuals. Sassaroly et al. reported that isolated stressful situations might induce dieting, binging and compensatory behaviors in female students with no previous clinical eating disorders, but who were dissatisfied with their bodily aspect 28 .

Hence, the purpose of this study was to determine the prevalence of main substance use and behavioral addictions among university students in France and to investigate the relationship between substance use, behavioral addictions, and stress in order to implement appropriate preventive interventions to foster public health benefits.

To the best of our knowledge, this is the first study conducted in a large population of university students, in France, with its main focus on the relationship between perceived stress, substance use and related disorders.

Perceived stress among university students is gender related. Female students are more exposed to stress associated with risk of cyber addiction, disorders related to alcohol use, eating disorders, and regular smoking than their male counterparts. Stress levels in our study were higher among female students than male students. Whilst some previous studies found similar results to ours 39 , others reported that gender was not a risk factor for stress 40 .

Our results also revealed a high prevalence of students with risk of eating disorder, since almost a quarter of our population had positive SCOFF score. Previously, Costarelli reported similar prevalence 41 . Eating disorders are the main behaviors associated with perceived stress and increased risk in each stress quartile.

Significant results were found for alcohol use and perceived stress. Whereas stress scores were positively associated with some alcohol disorders, they were not associated with excess alcohol use (drunkenness and binge drinking). It is important to highlight that the rate of students with alcohol use disorders (positive ADOSPA test) increased with level of perceived stress, even after adjustment. These results suggest that high stress levels lead students, not only to hazardous drinking patterns, such as dependence or getting high, but also to harmful behaviors towards alcohol, regardless of quantity or frequency of consumption.

Cigarette smoking in students is often described as social smoking and most students smoke for reasons other than stress relief, especially to facilitate social interactions, to avoid feeling alone, and behaving as a member of a group 42 . During stressful events, cigarettes not only help stress management, but also express non-verbally to others that the student is stressed 43 . Student tobacco consumption in response to stress can be associated with increased risk of nicotine dependence in later life.

Risk of cyber addiction was 27.5%, which might seem high but not all students become cyberaddicted. The prevalence of internet addiction disorder in an Italian student population was 5% and 1% respectively for moderate and serious users 44 . While studies have indicated that internet addicts are mostly young males with introverted personalities, the rate is currently rising in the female population 45 . Our study revealed a positive trend towards increased stress levels in students at risk of cyber addiction. Excessive behavior in internet addicts may serve as a stress coping strategy, albeit inadequate. The Internet may also be used as a forum for expanding social networks and consequently improving the chance of meaningful relationships, self-confidence, social abilities, and social support. Those using internet primarily for online chatting believed that it was psychologically beneficial. They also believed that frequent internet users were lonely and that the Internet could be addictive. “Chat” users, who are socially fearful, may be using the Internet as a form of low-risk social approach and an opportunity to rehearse social behavior and communication skills, which may help them improve interactions in face-to-face social environments 46 . Nevertheless, addicts may use the internet for extended periods, isolating themselves from other forms of social contact, and focussing almost entirely on the Internet rather than broader life events 47 .

Physical activity is also considered to have beneficial effects on mental health and stress coping capacity. Indeed, exercise as a moderator of stress levels was previously investigated in 48 university student populations 49 . Moreover, these results suggested an appraisal of physical activities in order to reduce psychological distress among students.

Although curriculum was not a risk factor for stress in our study, previous studies have shown that medical students are highly stressed and have proposed coping strategies 50 . To our knowledge, three recent nationally-representative surveys have studied the prevalence of substance use among young adults in France and allow comparisons 33 51 52 . Our results were similar to those reported by Riou Franca et al., for tobacco and cannabis, and Melchior et al. for alcohol, binge drinking and polydrug use. Common definitions and criteria concerning substance use need to be validated and used internationally to allow comparisons.

There is an attitude among students of turning a blind eye towards existing stress, which is a serious problem and may be a harbinger of serious mental and psychosocial problems. Based on the findings from our research, it is important for future prevention and intervention efforts to consider social setting and heightened stress among students as potential risk factors for engaging in risk behaviors.

Baker et al. found evidence of moderate rates of Internet use in the healthcare sector among adult internet users, as well as moderate effects of the Internet on the knowledge of users. Internet is clearly an important tool with the potential to improve information dissemination. The Internet is a means of transmitting information on health. The rise in popularity and possibilities of the Internet has led to a revolution in the provision of health-related information and treatment. While the health sector has primarily employed the Internet as a psycho-educational portal, advances in interactive technology have increased the potential of the medium to deliver targeted health interventions and other behavior change programs 49 53 . An interactive website: MyStudentBody.com offers a brief, tailored intervention to help heavy-drinking college students reduce their alcohol use. Acquiring and improving knowledge on student populations is a crucial factor in the development of health promotion programs in order to meet students’ needs and to help them cope. Enhanced understanding of the personality profile of university students can be helpful in academic and career choices. Developing efficient coping strategies in students and improving academic environments could also contribute to preventing the potential deleterious consequences of stress 50 .

Our study has certain limitations in as much as students were invited to participate in the study, and therefore the sample was not randomized. However, since several study disciplines were represented, the sample might be considered as representative of the student population. In addition, this cross-sectional study was based on self-reported information provided by students and collected throughout the year with very few answers returned during the stressful examination period. There is also some potential for reporting bias which may have occurred because of the respondents' interpretation of the question or desire to report their emotions in a certain way or simply due to inaccuracies in responses. Despite these limitations, we consider that our study provides important new information on the multiple health risks related to stress in students.

French-Normandy higher education students, especially female students, represent a stressed population with unhealthy risk behavior regarding nutrition, alcohol and Internet. Our findings stress the need to develop investigations in student populations, for better understanding of these new phenomena and to assess interactions with other risk behaviors. Prospective design studies are requiredto identify substance use and disorder changes in stressful situations. As a result, we are currently conducting the ELSE (Etude Longitudinale sur la Santé des Etudiants) longitudinal study with a cohort of undergraduate students to investigate both their levels of stress and factors associated with stress. Furthermore, new research areas are necessaryto investigate the impact of these risk behaviors on students and on their future professional careers.

The authors declare that they have no competing interests.

All authors participated in the study conception and design. LR and MPT participated in data acquisition and extraction. MPT and JL performed statistical analysis. MPT, JL and PD performed interpretation of data. MPT drafted the manuscript. All authors contributed to critical revision and have read and approved the final manuscript.