The global outbreak of the coronavirus disease 2019 (COVID-19) has brought unprecedented and significant stress to daily life of nearly all people around the globe (1). Governments worldwide mandated several preventive restrictions, also affecting the universities and their students. Students were impacted not just in their academic, but also in their interpersonal occupational functioning.

Previous studies indicated that young adults show a lower risk for serious somatic complications due to COVID-19 (2). However, little is known about the mental health consequences. In comparison to chronic physical disorders, mental disorders have their peak onset mostly during young adulthood (e. g., major depressive episode with a median age-of-onset in the early to mid 20s) (3), which makes university students a vulnerable population. Also, other stressors like academic pressure and increased responsibilities due to separation and individuation from their family may play an important role in vulnerability for mental health problems (4). A recent meta-analysis investigating the prevalence of depressive symptoms among university students in China showed that over one-fourth of students experienced depressive symptoms during the COVID-19 pandemic (5). A cross-sectional survey among university students in France revealed that ~50% of the respondents reported moderate to severe depressive symptoms (6). Several risk factors for depressive symptoms such as loss of employment, older age, more loneliness, higher perceived (study) stress as well as lagging academically behind others were observed in previous cross-sectional studies (7–10).

The prevalence of suicide, a leading cause of death among young adults in high-income countries (11) remained overall relatively unchanged during the pandemic (12–14). However, the literature reporting suicide rates in university students is scarce. O'Connor and colleagues demonstrated that younger adults (18–29 years) reported higher levels of suicidal ideation than older adults during the pandemic in the UK (15). Also, a Centers for Disease Control and Prevention (CDC) survey showed similar results, with young adults having higher rates of suicidal ideation than the general population during the pandemic (16).

Heterogeneous results were reported regarding alcohol consumption among students during the COVID-19 pandemic: While some studies demonstrated an increased frequency of drinking (17, 18), others indicated a decreased frequency (19). On the other hand, a both decreased and increased quantity in alcohol consumption among university students was reported (20, 21). A survey among Israeli and Russian students showed that substance use (tobacco, alcohol, and cannabis) was increased during the pandemic (22). On the contrary, Busse et al. observed unchanged cannabis use during the pandemic (23). It has also been shown that substances were used by some students as a coping mechanism with stress related to the pandemic (24).

The literature regarding change in eating disorder symptoms among university students during the COVID-19 pandemic is scarce. Zhou and Wade (25) showed a significant increase in eating disorder symptoms during the COVID-19 pandemic among female university students. A strong association between problematic eating behaviors and lockdown related stress was also demonstrated by another study among university students (26).

A recent cross-sectional study among college students highlighted that the majority of the participants (61%) reported that the management of preexisting mental health problems was disrupted (27). It is of note that pre-existing psychiatric disorders have been found to be associated with greater increases in stress as well as anxiety during COVID-19 mandatory lockdown (28).

Overall, the majority of the studies examining mental health in university students during the pandemic is cross-sectional in nature. Therefore, information on the magnitude of changes in mental health due to the pandemic as well as causality is scarce. There are some studies comparing the mental health status of university students before and during the pandemic which overall indicate a worsening of mental health conditions. For instance, the prevalence of university students experiencing moderate to severe depression and anxiety increased after the onset of the pandemic (29). In addition, an increase in psychological stress, depression, and anxiety symptoms was demonstrated in another study (30).

One year after the onset of the pandemic many people reported noticing high exhaustion and helplessness, pointed out the uncertainty and not being able to see an end in sight. Nevertheless, there is only little known about the changes in mental health during the pandemic, especially in university students. A comparative analysis of cross-sectional surveys during spring and autumn of 2020, respectively, in university students in Switzerland showed that students in the autumn cohort (October 2020) reported more binge drinking events and less physical activity than those in the spring cohort (April 2020) (31). Furthermore, students in the autumn cohort reported lower self-efficacy and lower social support in comparison to students in the spring cohort, which indicates the negative impacts of a long-lasting pandemic in several dimensions. On the other hand, Reznik et al. demonstrated no significant increase in binge drinking among university students from Russia and Israel surveyed in May and October/November of 2020, respectively (32). In a study from Italy, a significant increase in depressive symptoms during the COVID-19 lockdown was shown in comparison with the results before the lockdown (33). However, a subsequent reduction in depressive and anxiety symptoms was observed after the lockdown.

During the first wave of the pandemic (March–April 2020), German government had mandated several preventive measures, one of them being a lockdown. During the first lockdown period, we conducted a cross-sectional study in July and August 2020 which showed that more than one third of the students reported clinically relevant depressive symptoms; higher perceived stress, lower social support and lack of indirect social contact as well as not being a parent were significant predictors for increased depressive symptoms (10).

After the first wave of the pandemic, a debate about a possible second wave was started by July 2020. First, a partial lockdown was initiated in November 2020. Due to the high infection rates, harder measures were imposed in mid-December, which were extended due to the third wave of the pandemic. In the beginning of May 2021 most of the restrictions were alleviated. The state of Saxony was one of the most affected states in Germany with the most cases per 100.000 inhabitants (5.634/100.000 on April 8, 2021) (34). Given these circumstances, a second cross-sectional study among the students of University Leipzig was planned. The University of Leipzig with ~30,000 students and 14 faculties, ranging from medicine to theology, can be considered representative for a full-scale University in Germany and Europe (10).

The aims of this follow-up study were: (a) to estimate the severity of depressive and eating disorder symptoms as well as alcohol and drug consumption during the third wave of the pandemic in German university students; (b) to compare the severities of the aforementioned mental health symptoms with the severities in the sample of a congruent cross-sectional study from 2020 (10); and (c) to investigate risk and protective factors related to mental health. Based on existing literature we hypothesized that the severity of depressive symptoms, eating disorder symptoms, and alcohol and drug consumption would be increased in comparison to a sample of university students assessed in 2020. It was further hypothesized that a decrease in income, previous mental illnesses, female gender, and social isolation would increase the severity of these mental health issues.

The present study examined the differences in depressive symptoms, eating disorder symptoms as well as alcohol and drug consumption between two online surveys from 2020 and 2021 among university students. Further, the predictive factors of mental health symptoms were investigated. Findings suggest an increase in the severity of depressive symptoms as well as alcohol and drug consumption. Apprehensively, significantly more respondents reported suicidal ideation in the survey from 2021. No significant difference in bulimia nervosa symptoms was observed between the two samples. Lower self-efficacy, less social support and lower resilience as well as higher perceived stress and more loneliness were reported by the participants of the survey from 2021 in comparison to the participants from the survey 2020.

Regarding factors predicting mental health symptoms being female was a positive predictive factor for hazardous alcohol use and anorexia nervosa in comparison to being male, while being diverse was a positive predictor for all mental health outcomes, including depressive symptoms. Also, younger age, higher perceived stress and loneliness were positive predictors for all mental health outcomes. Decrease in income was only a positive predictor for higher depressive symptoms unlike than our hypothesis. Furthermore, having a previous mental illness significantly predicted only the eating disorder symptoms and depressive symptoms. Finally, social isolation (no direct contacts during the pandemic) significantly predicted only lower levels of harmful alcohol use.

Overall, there is a lack of studies focusing on the mental health of university students during the COVID-19 pandemic. Regarding longitudinal data, the majority of the existing literature focused on the pre- to in-pandemic changes of mental health symptoms. An increase in depressive symptoms after the outburst of the pandemic was shown in several studies among university students (29, 30), but also in the general population (44, 45). On the other hand, a recent longitudinal study conducted in Germany on the general population showed that depressive and anxiety symptoms decreased over time (from March/April to June 2020) (46). The result may be explained by the time of last assessment, which was ~2 months later than the first assessment and before the second lockdown. However, our second online survey was conducted during the second lockdown, which was with many restrictions stricter than the first one. Further, the study designs showed differences regarding study sample (study sample consisted of the same participants in the first and last assessment in the study of Bendau et al. (46)).

The factors predicting more depressive symptoms, such as a history of mental illnesses, younger age, lower self-efficacy and social support, were mainly in line with previous studies (31, 46, 47). In the first study, we have found that not being a parent as well as having indirect contacts 1–2 times a week, higher perceived stress, lower social support, and lower self-efficacy predicted higher levels of depressive symptoms (10). However, in the present study not being a parent and having indirect contacts did not significantly predict depressive symptoms.

Contrary to our results, Efstathiou et al. (48) reported no significant change in suicidal ideation of respondents between the first and second lockdown period in Greece (48). Further, another study reported a decrease in suicidal ideation among adolescents before (September 2019) and during the COVID-19 pandemic (June 2020) in Hong Kong (25). In a representative sample of Chines college students, survey during initial breakout of the COVID-19 pandemic and after decline in the second wave revealed that psychological distress significantly reduced after the peak of pandemic, but persisted in some students (49). These differences may be due to the study sample as well as the country in which the study was conducted, considering the different approaches of the governments as well as the impact of the pandemic on citizens.

Regarding alcohol consumption, the evidence is heterogeneous to date. An increase in total AUDIT-C scores during the pandemic compared to pre-COVID-19 was shown in a sample of vulnerable groups (people with self-reported mental health problems) (50). While an increased alcohol consumption during the pandemic, as found in our study, was reported by several other studies (51), a decreased or unchanged consumption was also observed (52, 53). The inconsistent results may be explained by the differences in study population as well as measurement time (beginning of the pandemic etc.) and different methods of assessment. Surprisingly, more direct contacts, higher self-efficacy, social support as well as resilience were positive predictors of higher scores for AUDIT-C, although no direct contacts as well as lower social support and self-efficacy predicted the higher levels of alcohol consumption in the first survey (10). Multiple social factors could account for these findings: During the first lockdown most of the people showed a high determination and motivation to follow the restrictions mandated by the government hoping that the pandemic will not last long. However, with the second wave many people noticed high exhaustion and hereby started not protecting themselves as carefully as in the first pandemic. Voluntary protective behavior was found to be less common especially among younger adults (< 30 years) (52). For this reason, it is possible that social gatherings took place more often, which students showing more resilience and having more social support and self-efficacy attended and consumed more alcohol there. Statistically significantly more participants reported consuming drugs in the second study. However, more participants in the second study indicated less drug consumption since the pandemic. Many studies highlight increased substance use since the onset of the pandemic (22, 54, 55).

This study also revealed an unchanged symptom severity of bulimia nervosa. In the existing literature there are inconsistent results regarding eating disorders: Breiner et al. demonstrated no significant changes in global scores of EDE-Q (Eating Disorder Examination Questionnaire) as well as self-induced vomiting and laxative use from prior to during the pandemic (56). On the other hand, some studies reported an increase in eating disorder symptoms (25, 57).

Because of higher mortality rates and several somatic comorbidities, the older adults were more in focus than the younger age groups at the beginning of the pandemic. However, findings showed that younger adults are experiencing more mental health symptoms, including depressive and anxiety symptoms, during the pandemic than older adults (58). Also, the present study highlights an increase in mental health symptoms among university students after ~one year of the pandemic. Overall, the existing evidence to date provides only a limited picture on changes in mental health of university students during the pandemic. Therefore, further studies investigating changes in mental health outcomes and associated factors are needed. Moreover, public health measures concerning the mental health of young adults should be promoted and adapted during crises like the pandemic. Through the expansion of online intervention possibilities barriers such as contacting medical professionals can be alleviated (59, 60).

Strengths of this study include the large sample size and the availability of a large comparison sample from 2020. Thus, changes in mental health outcomes during the course of the pandemic could be evaluated. The survey was conducted in a large German university including a wide range of faculties and with a high response rate, allowing for generalizability of the results to German students. Further, validated and well-established instruments were used.

This study has several limitations. First, change in depressive symptoms and symptoms of eating disorders as well as alcohol and drug consumption were assessed using cross-sectional student cohorts in 2020 and 2021, respectively. Because of the anonymity condition, it was not possible to identify the participants who participated in the online survey in 2020. Of note is that only 34.8% (n = 1,965) of the final sample stated participating in the first survey from 2020. Second, the results are limited to a single German university, therefore generalizability may be limited, also given the fact that solely self-report measures had been applied. Third, it is not possible to draw causal conclusions because of the observational nature of the study.

The present study reveals an increase in severities of depressive symptoms, including suicidal ideation, as well as drug and alcohol consumption among students of University Leipzig. Being diverse and of younger age, higher perceived stress and loneliness were risk factors for higher depressive and eating disorder symptoms, respectively, as well as alcohol consumption. Universities and health care policy should recognize and address mental health issues of young adults especially during times of crisis. Targeted preventive programs as well as easy-to-access measures to support students with mental health problems should be prepared and performed accordingly. Against this background, a return to pre-pandemic face-to-face seminars and course in universities, social activities and other events and “campus life”—covered by a hygiene concept and taking into account that a vaccination is broadly available nowadays, should be favored, whenever possible.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Ethics Statement. The Ethics Committee of the Medical Faculty, University of Leipzig, waived ethical approval for this study because of anonymity of the survey (08.03.2021). The participants provided informed consent to participate in this study. The patients/participants provided their written informed consent to participate in this study.

EK and ED-S equally contributed to the current paper and share first authorship. EK, ED-S, and CR-K conceptualized the study and constructed the questionnaires. ED-S and EK performed data analyses and drafted the initial version of the manuscript. SB and CR-K critically reviewed and revised the manuscript. All authors have read and approved the content of the final manuscript.

We also acknowledge support from Leipzig University for Open Access Publishing.

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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.