In the general population in Germany, with the onset of the pandemic and with the start of the second lockdown, people’s psychological wellbeing deceased significantly [5]. Younger people in particular had low wellbeing and perceived the restrictions more strongly compared to older people [5]. This also applies to medical students who volunteered in the context of the CoronAid initiative during the first phase of the pandemic who also report significantly lower levels of wellbeing at this time (13.5 ± 5.1) [26]. This score corresponds to the findings in our study of German medical students within the pandemic online semester. These perceptions were stronger in students in their preclinical phase compared to students in their clinical phases. It seems that even when the later students were restricted in their medical studies at university, having a focus on the clinical aspects of their studies may have buffered the decrease in their wellbeing to some extent. A further reason to explain this difference might be that the younger students (in their preclinical phase) require social contact when they are confronted with new surroundings, are not clear about all processes, examinations etc. These concrete social contacts were lacking in the digital semester, while their older counterparts in the clinical semesters have already established these contacts and can rely on these contacts to cope. Similar results were recorded in a qualitative study of wellbeing in distance learning [37]. Here, study-related barriers and enablers were described as clear influencing factors. These were higher for more experienced students than for beginners.

With respect to the stress perception, students had much higher stress before the pandemic than during the online and the hybrid semesters. This is related to all the challenges and stresses related to their study life, knowledge acquirement, internships and examinations, etc. When these study stressors were stopped as the university was in lockdown during the first months of the pandemic, students’ stress perceptions declined. Medical students voluntarily helping in the CoronAid initiative had stress cores in the moderate range (18.1 ± 7.6) [26]. Their stress perception as volunteers in hospitals was nevertheless higher as compared to the preclinical students during their online semester as observed in this study (probably because they are faced with the concrete problems of the pandemic in the hospitals), while their stress perception was nevertheless lower than that of students before the pandemic. Of course, they were not confronted with and stressed by examinations in the university, and were able to practically help. It might be that the volunteers have experienced that their current knowledge and their practical skills developed during the medical studies are already important enough to help their future colleagues and current COVID-19 patients. This experience may have enabled them to cope better with their fears and worries, and thus with their stressors. Interestingly, the stress perception of students from the preclinical and clinical phases were similar, whether they are studying online or in the hybrid semester; and both had lower stress than before. This is where the opportunities for free time management and reduced commute times seem to pay off. Since students were well equipped and prepared for digital teaching, no new stressors were added [38].

Several studies found that female students were more sensitive towards stress than male students [4,11,26]. However, this was not generally approved in this study, as the stress perception declined in women and men without significant gender related differences. During the hybrid semester, male students’ stress perception was lower compared to female students, while it was similar in the online semester and prior to the pandemic. The reason for this observation is unclear. This could be attributed to differences in the socio-emotional assessment of stress under digital learning conditions in terms of a basic openness to digital learning and the digital competence that goes along with it [39]. Nevertheless, female students’ wellbeing decreased stronger compared to male students. It seems that both perceived a decrease of stress but also of their wellbeing. Yet, it is important to underline that students’ stress perception was not significantly associated with their declining wellbeing. Some may feel stressed, but this does not necessarily impair their life satisfaction, while other may have low wellbeing but they are not stressed. Instead, their stress perception negatively affects their life satisfaction and it is positively associated with cool down perceptions (which both are inversely associated), which relates to emotional exhaustion and emotional distancing. Their stress might be more triggered by the medical courses in-person, digital or hybrid (and related internships and examinations), rather than by the pandemic which affected their psychological wellbeing because of general social distancing, as it was observed in other cohorts, too [5,6]. Whether the significant decrease of general life satisfaction and work engagement in women during the hybrid semester, which was not significant in men, is really due to the education form or rather a hint of ‘hope fatigue’ because of the long phases of social restrictions, is unclear. Their work engagement refers to their medical studies, which would indicate that the effect should be attributed to the hybrid format. Further, there were no gender-related differences in the support satisfaction which remained high in all three waves, while the support received from their university in particular declined in the hybrid semester, and this was, however, perceived stronger by women.

How can these observations be explained? During the Summer Semester 2020 (W4) there was no vaccine available and thus students may have feared the risk of own infection, kept social distancing and were participating their online courses. In contrast, in the Winter Semester 2021/2022 (W7) the more infectious Delta virus variant was predominating, but students had the opportunity to receive a vaccination. During this phase of stronger increases of infected persons, the university courses were in hybrid format, online and in-person. Students’ fears of own infection might be lower (because of their vaccination), but social distancing was nevertheless required and this may have affected their psychological wellbeing. In fact, as shown in our study, students’ psychological wellbeing, life satisfaction and work engagement were still low also in the W7 sample, while the stress level slightly increased, probably because of the teaching format and protection requirements in the classes. Medical students’ life satisfaction was declining during their online and hybrid semesters of preclinical semester students, while there were no significant changes in students within their clinical semesters. The reason that preclinical semester students seem to be more challenged than clinical semester students might be explained by the new university and social environment with new duties and with examination tests and thus stress; yet, also the older students may be stressed due to emotional exhaustion because of the overload of knowledge and major examinations.

Are medical students really more anxious and depressed because of the pandemic? A systematic review by Lasheras et al. reported a 28% prevalence of anxiety in medical students during the pandemic, and this rate was “similar to that prior to the pandemic but correlates with several specific COVID-related stressors“ [40]. Findings from Nepal indicate that the rates of depression among medical students might be lower as compared to prior data [41]. In that study, the prevalence of depression was higher in students within their preclinical years as compared to students in their clinical years [41]. Yet, there are obvious cultural and geographic differences with higher prevalence in Western societies compared to Asian societies [13]. Medical students might be generally at larger risks for developing anxieties and emotional exhaustion [42], probably because they understand the underlying mechanisms of disease and related risks, but also because of an often (perfectionistic) personality [43], and academic pressure and financial difficulties [44]. However, particularly in the early phases of the pandemic, students’ insecurities how they may proceed with their medical studies, their clinical courses and finally their examinations might have been higher than in the later phases when universities have learned to deal with the restrictions and found ways to offer online courses and later on hybrid formats. This would fit to our findings of higher satisfaction with university support in the hybrid semester. Nevertheless, in our study, students from W4 and W7 perceived their stress similar, and also their wellbeing, life satisfaction and work engagement was similar—indicating that the format of education was not of crucial relevance for all outcome variables.

Our regression analyses to predict medical students’ stress perception, psychological wellbeing, and work engagement indicated that students´ stress perception can be explained only to some extent by their general dissatisfaction with their medical studies, whether they are studying in-person, online, or in a hybrid format. As their satisfaction with their studies was stably high also during the pandemic semester, while their satisfaction with the university support was at its lowest during the hybrid semesters, their stress decline obviously has other reasons. Regardless of the teaching mode, students´ wellbeing decreased significantly because of the pandemic, while their stress levels can be explained by the change from in-person to online to hybrid semesters. Similarly, their work engagement (which did not change because of the pandemic related teaching formats) was predicted best by their satisfaction with the studies (which also remained stable for the three teaching formats) throughout all three waves of recruitment. To explain their wellbeing, satisfaction with their studies and the support received by their university was of some relevance in the regression model, but cannot explain too much of the wellbeing variance. Thus, the decline is related to the pandemic related general negative outcomes, not by the teaching formats. However, during their hybrid semester, their wellbeing could be explained much better by satisfaction with the medical studies (which was indeed high), by being in the clinical phase of the studies, by satisfaction with university support, by support received from by other students, and by being male. The hybrid semester was obviously different in its outcomes on students, and it was applied in a later phase of the pandemic with high increases of infection incidences by the more aggressive Delta variant of the virus, but with the chance to be vaccinated. Thus, more chances for social contacts and less restricted teaching formats might explain some of the described observations, too.