The period between late teens and early twenties, often referred to as emerging adulthood (1), is a critical phase in life where important developmental steps are taken (2), including the completion of education and the transition from school to work. This is especially true for university students who are in the final chapter of their educational journey and close to entering the labor market. In addition to their educational and occupational development, they are also in the process of establishing their own independent life away from the parental home (3). They have to adapt to a new and often challenging university environment during the beginning of their studies.

Overall, the physical health of university students, as most being in their late teens or early twenties, is reasonably good but a little bit lower than of their non-student peers (4). Nonetheless, there is some evidence for consistent health problems and indication of a correlation between health conditions and socio-economic status of students. In Germany, Diehl et al. (5, 6) found that subjective social status was associated with different health indicators, i.e., those with a lower status displaying greater health problems overall compared to those with a higher status. In a large health survey including 6,198 German university students, Grützmacher et al. also found that although students are overall quite healthy, they rated their health lower and seemed to consume more drugs and alcohol in comparison with their non-student peers (4, 7). Eisenberg et al. (8) found evidence for mental health problems among university students. Sixteen percent of undergraduate and 13 % of graduate students showed symptoms of depression or anxiety. They also noted that these two psychological disorders were more prevalent among students with a lower socio-economic status. Ibrahim et al. (9) found in a systematic review that the rates of depression of university students were indeed higher than in the general population. For students in the UK, the risk of depression was higher for those from a more disadvantaged background (10). Steptoe et al. (11) found similar results in a comparison of 23 countries. They also found that income inequality and the level of individualistic culture had an impact on depression among university students (11).

Social determinants are important factors for health differences (12) and are of course also found for university students (6). As university students have not yet developed a stable socio-economic status of their own, the socio-economic background of their parents is a major cause for health inequalities in young adults (13). The impact of the socio-economic position on health can therefore be best described as the impact of the highest socio-economic position of parents on health of their offspring. Additionally, Hagquist (14) proposes to distinguish between the socio-economic position students try to achieve and not to concentrate on the socio-economic position of their parents.

Most of these results show the importance of social inequality for students' health. The main pathways of social inequalities in health are behavioral, psychosocial and material factors (15–17). The harmful outcome of smoking, drinking and overweight is prominent in the literature (3) and well documented for university students (18). Research shows that these behavioral aspects can vary by socio-economic status (3, 19–22). Psychosocial aspects can be responsible for causing health problems (23) which are socially patterned as well (24). The harmful effects of psychosocial conditions, such as psychosocial stress on health, have been previously described in the effort-reward model of work stress (25) and were also found for university students (26). Thirdly, less advantageous material conditions are likely to cause unhealthy living conditions, such as an increased exposure to traffic, loud noise or inadequate housing (27).

Whether the socio-economic background of parents still has a significant impact on students' health independent or in combination with behavioral, psychosocial and material factors in German university students is still unclear. Our study therefore had three main objectives. First, we described students' health over the course of the time of their studies and analyzed the impact of parents' socio-economic status on self-rated health. Second, we analyzed the impact of behavioral, psychosocial and material factors on students' self-rated health, and third, we analyzed the possible impact of parents' socio-economic status in students' self-rated health via health behavior, psychosocial, and material resources.

We used data from the 2020 data release (15.0.0, May 2020) of the National Educational Panel Study [NEPS; see (28, 29)]. The NEPS is carried out by the Leibniz Institute for Educational Trajectories (LIfBi, Germany) in cooperation with a nationwide network. The NEPS study is conducted under the supervision of the German Federal Commissioner for Data Protection and Freedom of Information (BfDI) and in coordination with the German Standing Conference of the Ministers of Education and Cultural Affairs (KMK) and–in the case of surveys at schools–the Educational Ministries of the respective Federal States. All data collection procedures, instruments and documents were checked by the data protection unit of the Leibniz Institute for Educational Trajectories (LIfBi). The necessary steps were taken to protect participants' confidentiality according to national and international regulations of data security. Participation in the NEPS study was voluntary and based on the informed consent of participants. This consent to participate in the NEPS study can be revoked at any time.

NEPS provided data for six cohorts across the life course: newborns, children in kindergarten, children in the 5^th and 9^th grade, students at universities and adults in Germany. Our analyses were based on the starting cohort 5 consisting of university students. The data collection for this cohort started in the winter term of the academic year 2010/11 with a representative sample of 17,909 first-year university students. For most years, NEPS conducted two interviews per year, one computer-assisted telephone interview (CATI) and one computer-assisted web-interview (CAWI). The CATI-interview usually took place in the spring or summer and the CAWI in the autumn of the same year, with the exception of 2017, where both types of interviews were conducted simultaneously. Currently, there were 15 waves of data available: nine CATI waves and six CAWI waves. Starting cohort 5 started with 17,909 respondents in 2010/11. Due to panel mortality, the number of respondents was reduced to 6,531 after 15 waves (nine years later). The age structure was quite heterogeneous, with students being between the ages of 16 and 64 years with a mean age of 21 years in the first year of the survey. Since our focus was on young adults' health, we excluded all students who were older than 25 years in the first wave from the analysis. Furthermore, students who finished their studies were kept in the analysis as long as they did not drop out of NEPS. This left us with 24,394 observations of 8,898 students over a maximum period of 9 (CATI) waves for our main analysis.

As we described in the following sections, some variables in NEPS were assessed irregularly. Some were asked only once, some were asked with long intervals between them, some were asked only in later waves of NEPS. Therefore, we divided our analysis in two parts: a cross-sectional and a longitudinal part. Our cross-sectional analysis captured conditions at the beginning of the studies; the longitudinal analysis captured the development of students' health over time. To maximize our samples sizes for each analysis, we used the full available sample for each analysis resulting in slightly different sample sizes for each model. See Table 1 for an overview of the availability of key-independent variables in NEPS.

In the first year of the NEPS-student-cohort, students in our analytic sample had a mean age of 20 years and were between 16 and 25 years old (see Table 2). Roughly, 40% of first year students in NEPS were male. Eighty Two percent had no migration background, nearly 3% were migrants themselves and 7% had parents or grandparents who had migrated to Germany. Twenty-Two percent of the students started their studies at a University of Applied Sciences. The majority of students studied Linguistics or Cultural Studies, Mathematics and Sciences were on second place, followed by Engineering, Economics, Social Sciences, Medicine, Law, Arts and Fine Arts, and Agricultural-, Forest- and Nutrition Sciences and Veterinary Medicine.

This study explored the impact of the parental socio-economic status on the health status of university students based on cohort 5 of the National Educational Panel Study. Our study contributed to the literature in several ways. Surprisingly, little was known about the general health of university students and even less research on students' health was based on a large scale representative survey. Most studies either focused on specific health behavior of students (39), on specific groups, such as medical students (40), or on mental health only (8). We found that, overall, university students were in good health with over 90 % of the students reported to be of good and very good health in their first study year. Students in NEPS were therefore healthier than in other surveys and they also showed less symptoms of potentially harmful health-related behavior (4, 8).

Our study was especially important with regard to health inequality. Most students are in a privileged situation. They are in the process of acquiring an education that will most likely give them access to high-ranking jobs and a good income. Although the expansion of education in the last decades had improved the access to higher education across all social classes, there was still a bias toward children from a higher socio-economic background being more likely to be enrolled at universities. We found that the importance of the socio-economic status of parents for the health of their offspring was still persistent among university students, a connection that was already known for adolescents (41) and young children (42). This showed that the “social class” of origin had more persistence than the aspired social class as suggested by Hagquist (14). However, we found no impact of parental education on students' health. This was in line with other studies who also found that education of parents had no correlation with health, once education of children was accounted for (43). Yet, in our analyses, the occupational status of parents was related to children's health. An interesting observation in this context was that once material conditions of students were introduced to the model, the coefficient of ISEI was reduced. This was an important finding as it showed that material conditions might be a mediator between occupation of parents and health of students. These differences between measures of socio-economic position of families and the interfering role of the own SEP of the student are an important finding for future research on the influence of parental background on children's health.

As it is often found that inequality has a strong indirect effect on health (15, 16), one main aspect of our analysis was the analysis of indirect effects of family background via material resources, psychosocial factors, and health-related behavior on students' health. Our analysis reproduced previous findings and confirmed that the socio-economic conditions of parents had an indirect impact on health of students via behavioral, psychosocial, and material factors. Students' health behavior, their psychosocial and material resources showed clear patterns depending on their socio-economic background (described above). Education and the occupational status of parents mattered, more or less, equally for the indirect effect. This suggests that different possible mediating pathways should be taken into account while analyzing life-course related health inequalities, otherwise the importance of parental education would have not been noticed, due to a non-significant direct effect.

Our study had some shortcomings. NEPS only contained self-rated health as a health measure. Therefore, we were not able to extend our analysis with a measure of psychological health like depression, which is an important indicator for students' wellbeing (10). It would have been also important to control for measures of stress as an important co-founder for health (25). Some measures had also not been available across all waves and had to be forwarded between waves.

Another caveat was the missing of a direct measure of financial conditions of students or their parents. Especially a measure of students' income would have been an important information. As with all panel data, we might had issues with panel mortality and selection: It is likely that especially students who drop out are the ones less likely to continue with their studies. This potential selection of students who did better in their studies should have resulted in a more homogeneous group of students with less impact of their social background, since students with higher educated parents typically do better in university. That health inequality still exists speaks to the importance of inequality for health research.

One final limitation is our approach to mediation analysis. As was pointed out by Richiardi et al. (44) and VanderWeele (45), mediation analyses can be biased and other methods, such as counterfactual analysis would be a better alternative do disentangle mediation. Therefore, our specific mediation analysis only gave us only an impression of the indirect effect of parental socio-economic status on student health. Due to the fact that the mediator variables were categorical variables with multiple categories we also couldn't apply a standard mediation analysis and were therefore not able to calculate total effects and determine the magnitude and significance of the indirect effects.

The importance of the life course for health outcomes later on in life is widely known (46). Especially the theory of cumulative inequality and the critical period model for health inequality are important in this regard. The time between adolescence and adulthood marks an important period (1) and can be seen as a “critical period.” For many, this is the phase where they leave their parental homes and start their own life away from home. It is also the time before the so called “rush-hour” of life starts, where people have to start their career, find a partner and start a family, build a home, etc. (47). The health inequalities we found in university students are likely to accumulate over time and differences between social groups may even increase over time, resulting in greater inequality and worsened health outcomes later on in life.

We believe our study is an important contribution to the understudied subject of students' health. We see the impact of social inequality on health among such a privileged group like university students as an important sign of the importance of health inequality. Given that more and more students in Germany come from a non-academic background, health inequality is likely to become more important in the future. And health promotion of students should be of special interest for universities and for public health.

Publicly available datasets were analyzed in this study. This data can be found here: https://www.neps-data.de/.

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

CD wrote the manuscript and conducted the data analysis. KD, JS, MR, WS, PR, ND, and CRP contributed in writing the manuscript and participated in the final review of the paper. All authors contributed to the article and approved the submitted version.