Online data collection was conducted between February and May 2014, using a battery of questionnaires posted on Googles docs. Information regarding this survey was distributed to all registered students at the University Paris Ouest Nanterre La Defense and other universities in the Parisian area.

The study protocol was approved by the ethics committee of the Psychological Science and Learning Science department at the University of Paris Ouest Nanterre La Défense, UFR SPE (Department of Psychology and Education) and by the CNIL (Commission nationale de l'informatique et des libertés). In accordance with the Helsinki declaration, written informed consent was obtained from each student before inclusion.

Out of 630 replies, 147 were incomplete and/or useless. Thus, the final sample consisted of 483 college students (355 women, 128 men), aged between 18 and 24 years (M: 20.23; SD: 1.99) from the Parisian area (11.6% were not registered at the University Paris Ouest Nanterre La Defense).

The questionnaire of the study was divided in two parts. In the first part, students were asked to give socio-demographic data concerning their gender, age, place of residence, current year of studies, study program and university of origin. They also had to report, using a visual analog scale, whether they were satisfied with the discipline they chose at university and whether they had ever repeated an academic year. The second part included six questionnaires:

Self-esteem was measured using the 10-item version of the Rosenberg Self-Esteem Scale (RSES), developed by Rosenberg (1965) and translated and validated in French by Vallieres and his team. All items were answered using a 4-point Likert scale format ranging from strongly agree to strongly disagree (Vallieres and Vallerand, 1990). A score lower than 30 indicates a low self-esteem (Chabrol et al., 2004). Cronbach's alpha coefficient in this study was: (0.86).

Perceived self-efficacy was evaluated using the General Self-Efficacy Scale (GSE), developed by Matthias Jerusalmen and Ralf Schwarse (Luszczynska et al., 2005). It was translated in French and validated by Dumont and his team and consisted of 10 items that are answered using a 4-point Likert scale ranging from “not at all true” to “exactly true” (Dumont et al., 2000). Cronbach's alpha coefficient in this study was: (0.84).

Optimism was assessed using the 10-item revised version of the Life Orientation Test (LOT-R) developed by Scheier and Carver (Trottier et al., 2008). It was translated in French and validated by Sultan and Bureau (Shankland and Martin-Krumm, 2012). A higher score indicates higher optimism. Cronbach's alpha coefficient in this study was: (0.83).

Student's well-being was measured using the Satisfaction With Life Scale (SWLS) developed by Diener et al. (1985); Diener (2006). It was translated in French and validated by Blais and his team and consisted of five items that are answered on a scale from 7 (strongly agree) to 1 (strongly disagree) (Blais et al., 1989). A higher indicates a higher satisfaction. Cronbach's alpha coefficient in this study was: (0.82).

Perceived stress was assessed using the 10-item Perceived Stress Scale (PSS-10). It was developed by Cohen et al. (1983) and translated and validated in French by Bellinghausen et al. (2009). Each item is rated on a 5 points scale from “1” (never) to “5” (very often). It includes two factors: perceived helplessness and perceived self-efficacy (Bellinghausen et al., 2009), and two scores thresholds: a score superior to 24 for anxiety and to 26 for depression (Collange et al., 2013). In this study Cronbach's alpha coefficient in this study was: (0.82).

Psychological distress was evaluated using the 28-item General Health Questionnaire (GHQ-28) scale. This measure was initially described by Goldberg in 1972 (Goldberg and Hillier, 1979) then translated and validated in French by Bolognini et al. (1989). It is divided into 4 subscales which measure somatic symptoms, anxiety/insomnia, social dysfunction and severe depression (Goldberg and Hillier, 1979), each consisting of 7 items that are answered using a 4-point Likert scale (Bolognini et al., 1989). A score greater than or equal to 5 indicates psychological distress (Guelfi, 1993). Cronbach's alpha coefficient in this study was: (0.84).

Statistical analyses were performed using the software Statistica (version 12). First, descriptive analyses (such as percentages, means and standard deviations) were produced. Then, bivariate analyses (Mann–Whitney U-test, Spearman's correlation) were performed in order to investigate the possible links between all variables of interests (gender, age, year of studies, UFRclassification at the University of Paris Ouest Nanterre, self-esteem, optimism, generalized self-efficacy and psychological distress (somatic symptoms, anxiety/insomnia, social dysfunction, severe depression). Finally, multivariate analyses (multiple regressions) were run to test the link between perceived stress as the dependent variable and including, as predictors, the year of studies, the academic sector, self-esteem, optimism, the sense of self-efficacy and psychological distress. The significance limit was set at p < 0.05.

The final sample consisted of 483 students from the Parisian area (88.41% from the University of Nanterre), of whom the majority were women (73.5%). The average age was 20.23 years (±1.99). Most students lived with their parents (68.7%), did not have kids (88.4%) and reported being somewhat satisfied with their studies (38.5%). This sample included students from all years of studies (first year of license to PhD) and from different academic sectors. The majority of students did not repeat any academic year (77.6%).

Low self-esteem was reported by 57.6% of students while 25% of them were slightly satisfied by their well-being and 56.7% had little optimism. The majority of them also reported having a low sense of self-efficacy (62.7%), anxiety (86.3%) and depression (79.3%). Finally, according to their GHQ-28 scores, 72.9% of students declared suffering from psychological distress.

Results showed that men had a higher sense of self-efficacy than women. The latter had significantly higher scores on perceived stress and both its factors (perceived helplessness and perceived self-efficacy) and presented more psychological distress, including more somatic symptoms and anxiety/insomnia.

When simultaneously taking into account all studied factors (gender, age, year of studies, UFR classification at the University of Paris Ouest Nanterre, self-esteem, optimism, generalized self-efficacy and psychological distress (somatic symptoms, anxiety/insomnia, social dysfunction, severe depression), we found a significant link between stress and most of these factors. In fact, on the one hand, there was a negative correlation between perceived stress and factors that could be considered as “positive” ones such as self-esteem, optimism and sense of self-efficacy. On the other hand, there was a positive correlation between perceived stress and factors that could be considered as “negative” ones such as psychological distress and its four factors on the GHQ-28 scale including somatic symptoms, anxiety and insomnia, social dysfunction and severe depression.

Results showed a significant link between gender and scores on the perceived stress scale, specifically in women who were more stressed on most of the evaluated factors. Analyses did not show any association between perceived stress score (and its factors) and age, year of studies or UFR classification.

Additional studies on perceived helplessness and perceived self-efficacy, two subscales of the PSS-10 scale, showed similar results concerning the negative and positive associations found with perceived stress.

A linear multiple regression analysis was performed using perceived stress as the dependent variable and including as predictors the year of studies, the academic sector, self-esteem, optimism, the sense of self-efficacy and psychological distress. The total variance accounted for by the model was 57% [F_{(7, 475)} = 93.269; p < 0.0001]. Life satisfaction (Beta = −0.11; p = 0.002), self-esteem (Beta = −0.20; p = 0.000002), optimism (beta = −0.12; p = 0.002), self-efficacy (beta = −0.19; p = 0.00) and psychological distress (beta = 0.38; p = 0.00) independently and significantly predicted perceived stress. Furthermore, among those variables, self-esteem, optimism and self-efficacy negatively predicted perceived stress and were considered to be “positive” variables while psychological distress positively predicted perceived stress and was considered to be a “negative” variable.

In line with several previous research stating the importance of psychological problems among college students (Lafay et al., 2003; Lassarre et al., 2003; Boujut et al., 2009; Strenna et al., 2009), our results show that the students included in our sample (N = 483) have high levels of anxiety (86.3%), depression (79.3%), psychological distress (72.9%) and have a low self-esteem (57.6%). The level of stress in our sample is slightly higher than the ones found in the literature(Vandentorren et al., 2005; Strenna et al., 2009; Dachew et al., 2015; Deasy et al., 2015; Larcombe et al., 2016; Weier and Lee, 2016).

Overall, the psychological difficulties were significantly higher in women than men when it came to perceived stress, perceived helplessness, perceived self-efficacy, global psychological distress, somatic symptoms, anxiety and insomnia. They also have a lower sense of self-efficacy than men. However, we did not find significant differences between women and men concerning self-esteem, life satisfaction, and optimism.

On the one hand, these results are in accordance with many previous research performed in different countries (Backović et al., 2012; Cruz et al., 2013; Shamsuddin et al., 2013) that show that levels of stress and psychological distress are higher in female than male college students (Spitz et al., 2007; Backović et al., 2012; Deasy et al., 2015). On the other hand, our results concerning associations between gender and the sense of self-efficacy are in opposition to other literature data (Follenfant and Meyer, 2003; Ayle and Nagels, 2014). This was also the case for our results concerning the links between gender and self-esteem (Dozot et al., 2009).

When it comes to the link between stress and gender, our results confirm the significant link previously find in other studies. Female students are usually found to be more stressed than male students (Fornés–Vives et al., 2012; Cruz et al., 2013; Shamsuddin et al., 2013). However, Koochaki et al. (2011) did not find any significant differences according to gender and oppositely, two studies found that male students reported higher stress levels than females (Acharya, 2003; Ahern and Norris, 2011).

When it comes to the link between stress and age, we did not find any associations between these two variables, independently of the studied stress factor. This result is in line with some studies (Koochaki et al., 2011) while others show a negative association between perceived stress and age (Fornés–Vives et al., 2012; Voltmer et al., 2012).

Both the year of studies and the academic sector did not have a significant impact on perceived stress in the regression analyses. Nonetheless, and unlike other protective factors (i.e., life satisfaction, self-esteem, optimism and generalized self-efficacy) that negatively predicted perceived stress, we found that psychological distress significantly contributed the most to the variance of perceived stress. The total variance of perceived stress accounted for by the model including all studied factors was 57%. These factors could be considered stress vulnerability factors among students.

Interestingly, previous research found significant associations between perceived stress and: (1) psychological distress (La Rosa et al., 2000; Strenna et al., 2009); (2) self-esteem (Boujut, 2007) optimism (Mazé and Verlhiac, 2013) self-efficacy (Han, 2005). In fact, according to the literature, the most important predictor of stress symptoms in university students was the sense of self-efficacy (Han, 2005). Nevertheless, we found that self-esteem (beta = −0.20; p = 0.000002) and self-efficacy (beta = −0.19; p = 0.00) negatively predicted it. Regression analysis also showed that psychological distress (beta = 0.38; p = 0.00) was the most powerful positive predictor of stress symptoms.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.