Loneliness may link campus relationships to student mental distress, German panel study finds

University students who felt better connected to their fellow students reported fewer mentally unhealthy days three years later, and a new analysis suggests that loneliness may account for a substantial part of that association. The study, published in SSM – Population Health, drew on three waves of the German National Educational Panel Study and followed 3014 students who entered higher education in Germany in 2010. The researchers, Gunnar Voss and Philipp Pohlenz, examined whether two distinct kinds of campus relationships—ties with peers and perceived relationships with teaching staff—were linked to later psychological distress through the experience of loneliness.

The evidence came from a secondary analysis of Starting Cohort 5 of the German National Educational Panel Study, using a simplified panel version of the data. The full panel contains 17,909 participants and 19 survey waves through 2022. For the present analysis, 3550 participants were documented as enrolled at both wave 6 and wave 10, and 536 of them, or 15.1 percent, lacked at least one core analysis variable, leaving 3014 students with complete information on the relationship scores, loneliness, mentally unhealthy days, and demographic covariates. The sample size followed from these eligibility and data-availability criteria rather than an a priori power calculation, and because the analyses were unweighted, they do not provide nationally representative prevalence estimates.

The measures were spread across the panel’s schedule in a way that gave the analysis its temporal structure. Relationship measures were collected at wave 4 in 2012, loneliness at wave 6 in 2013/2014, and mentally unhealthy days at wave 10 in 2016. Peer integration was measured with three items asking whether students had successfully built contacts with other students, discussed field-specific questions with fellow students, and had many contacts with students, rated on a four-point scale; the scale had a Cronbach’s alpha of .84. Student–faculty relationship quality was captured with four items on receiving recognition from teaching staff, getting along well with them, being treated fairly, and whether staff showed interest, with an alpha of .79. Loneliness was assessed with a single statement, “I often feel lonely,” rated from 1 for “does not apply at all” to 5 for “applies completely.” The outcome was the number of days in the previous 30 on which respondents felt unwell because of their mental state, a measure from the CDC Healthy Days approach that captures perceived mentally unhealthy days rather than a diagnosis of depression or anxiety.

The descriptive picture showed a sample in which psychological distress was relatively low on average but unevenly distributed. Students reported a mean of 3.34 mentally unhealthy days, with a standard deviation of 4.97, and the distribution was right-skewed as is common for Healthy Days data in non-clinical populations. Some 37.3 percent reported no mentally unhealthy day in the previous month, 47.0 percent reported between one and six days, 9.4 percent reported 7 to 13 days, and 6.4 percent reported at least 14 days, a threshold frequently used to indicate frequent mental distress. On the loneliness statement, 19.8 percent selected “does not apply at all,” 44.9 percent “tends not to apply,” 20.7 percent “partly applies,” 11.4 percent “tends to apply,” and 3.2 percent “applies completely.” The authors stress that these are agreement categories rather than frequencies of loneliness, and that the single item does not establish a prevalence of frequent loneliness.

The statistical approach used ordinary least squares mediation models, estimated separately for each relationship measure and adjusted for age and gender. For peer integration, the total association with later distress was negative and statistically significant, with an unstandardized coefficient of −0.611. Higher peer integration was associated with lower loneliness at the subsequent wave, and loneliness in turn was positively associated with distress. The indirect association through loneliness was −0.216, with a 95 percent bootstrap confidence interval from −0.295 to −0.145 based on 5000 resamples, corresponding to 35.3 percent of the total association. The direct association between peer integration and distress remained statistically significant after accounting for loneliness. The authors emphasize that this proportion describes the fitted statistical decomposition, not the share of distress causally explained by loneliness.

The pattern for student–faculty relationship quality was descriptively different. Its total association with distress was not statistically significant, and its direct association was close to zero, yet the measure was negatively associated with loneliness, with a standardized coefficient of −0.092 compared with −0.206 for peer integration. The bootstrapped indirect association through loneliness excluded zero, at −0.146 with a confidence interval from −0.225 to −0.082. In a joint outcome model that included both relationship measures and loneliness, peer integration remained negatively associated with mentally unhealthy days while the faculty coefficient was not statistically significant, and loneliness and the female indicator had the largest standardized coefficients. The authors caution that the difference between the peer and faculty estimates has not been formally tested, and that the uncertain direct and total faculty associations should not be read as evidence that these relationships are unimportant.

Sensitivity analyses tested whether the findings depended on measurement and model choices. After adjustment for wave-1 general health, the indirect estimates were smaller but remained negative: −0.174 days for peer integration and −0.110 for faculty relationship quality. Requiring all seven relationship items, which reduced the sample to 2954, also preserved the pattern. The researchers then replaced the linear assumptions with models suited to the data’s actual shapes, using multinomial models for the five loneliness categories and negative-binomial models for the skewed count of unhealthy days, combined with 5000 paired person-level bootstrap resamples. These nonlinear models also yielded negative indirect differences for both relationship measures, with peer estimates of −0.236 and −0.216 days and faculty estimates of −0.129 and −0.132 days depending on the contrast. The direction of the indirect associations thus persisted under alternative functional assumptions, though these contrasts do not establish causal mediation.

The theoretical framing distinguishes two forms of social capital. Peer integration is interpreted as an approximation to bonding social capital, the resources available through ties within groups sharing a social identity, while faculty relationship quality is considered in light of linking social capital, a concept that concerns relationships across institutional differences in power and authority. Teaching staff occupy roles involving assessment, guidance, and institutional authority, so relationships with them have a linking dimension that a bridging label alone would obscure. The authors are careful to note that neither measure establishes which resources students actually obtain through their contacts, and that the items describe perceived relationships rather than network structure, reciprocity, or resources mobilized. Loneliness, in turn, is defined as the subjective discrepancy between desired and experienced social connectedness, and proposed mechanisms linking it to mental health include heightened sensitivity to social threat, impaired self-regulation, and physiological stress responses.

The study’s limitations are substantial and shape what the findings can claim. The observational design leaves causal explanations unresolved; temporal separation in measurement does not eliminate confounding, selection, or reciprocal influence, and prior mental health may influence later social integration and loneliness. Loneliness was measured with a single item at one wave, which does not distinguish social from emotional loneliness, and the outcome is a single Healthy Days item that does not distinguish depression, anxiety, or other conditions. Unmeasured confounding remains possible, with parental education, financial resources, paid work alongside study, and institutional sector not included in the models. Comparisons with the 536 excluded participants were limited by extensive missingness in their relationship scores and cannot rule out selection bias. The proposed organizational level—programs, courses, cohorts, teaching arrangements, timetables, campus spaces, and digital infrastructures that organize whether and how students repeatedly encounter one another—was not measured, so the institutional propositions remain a research agenda rather than an empirical result of the analysis.

The data were collected between 2012 and 2016, which allowed the associations to be examined without pandemic-related disruption of university life, but their transferability to contemporary higher education requires examination. Within those constraints, the findings identify relationships worth examining in research on student mental health: peer integration showed negative direct and indirect associations with later distress, while faculty relationship quality showed a negative indirect association through loneliness alongside uncertain direct and total associations. The authors suggest that universities can provide occasions for connection without being able to prescribe friendship, pointing to cohort structures, facilitated study groups, structured advising, and reliable feedback as arrangements that future studies should evaluate alongside both participation and health outcomes. The relevance of these pre-pandemic observational findings to current university settings, they conclude, requires replication with recent data.

Subject of Research: Longitudinal associations between peer integration, student–faculty relationship quality, loneliness, and psychological distress among German university students

Article Title: The social roots of student mental health: Organizational opportunity structures, social capital, and loneliness

Article References: Voss, G., & Pohlenz, P. (2026). The social roots of student mental health: Organizational opportunity structures, social capital, and loneliness. SSM – Population Health, 36, Article 101992. https://doi.org/10.1016/j.ssmph.2026.101992

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101992

Keywords: student mental health, loneliness, social capital, peer integration, faculty relationships, psychological distress, German National Educational Panel Study, mediation analysis, higher education, Healthy Days, observational study, organizational opportunity structures