research paper on stress

Stress and Health: A Review of Psychobiological Processes

Affiliations.

• 1 School of Psychology, University of Leeds, Leeds LS2 9JT, United Kingdom; email: [email protected].
• 2 Department of Psychological Science, School of Social Ecology, University of California, Irvine, California 92697, USA; email: [email protected].
• 3 Division of Primary Care, School of Medicine, University of Nottingham, Nottingham NG7 2UH, United Kingdom; email: [email protected].
• PMID: 32886587
• DOI: 10.1146/annurev-psych-062520-122331

The cumulative science linking stress to negative health outcomes is vast. Stress can affect health directly, through autonomic and neuroendocrine responses, but also indirectly, through changes in health behaviors. In this review, we present a brief overview of ( a ) why we should be interested in stress in the context of health; ( b ) the stress response and allostatic load; ( c ) some of the key biological mechanisms through which stress impacts health, such as by influencing hypothalamic-pituitary-adrenal axis regulation and cortisol dynamics, the autonomic nervous system, and gene expression; and ( d ) evidence of the clinical relevance of stress, exemplified through the risk of infectious diseases. The studies reviewed in this article confirm that stress has an impact on multiple biological systems. Future work ought to consider further the importance of early-life adversity and continue to explore how different biological systems interact in the context of stress and health processes.

Keywords: HPA axis; allostatic load; autonomic nervous system; cortisol; genomics.

Publication types

• Autonomic Nervous System / metabolism
• Hydrocortisone / metabolism
• Hypothalamo-Hypophyseal System / metabolism
• Pituitary-Adrenal System / metabolism
• Stress, Psychological / metabolism*
• Hydrocortisone

Recent developments in stress and anxiety research

• Published: 01 September 2021
• Volume 128 , pages 1265–1267, ( 2021 )

Cite this article

• Urs M. Nater 1 , 2  

5342 Accesses

2 Citations

Explore all metrics

Avoid common mistakes on your manuscript.

Stress and anxiety are virtually omnipresent in today´s society, pervading almost all aspects of our daily lives. While each and every one of us experiences “stress” and/or “anxiety” at least to some extent at times, the phenomena themselves are far from being completely understood. In stress research, scientists are particularly grappling with the conceptual issue of how to define stress, also with regard to delimiting stress from anxiety or negative affectivity in general. Interestingly, there is no unified theory of stress, despite many attempts at defining stress and its characteristics. Consequently, the available literature relies on a variety of different theoretical approaches, though the theories of Lazarus and Folkman ( 1984 ) or McEwen ( 1998 ) are relatively pervasive in the literature. One key issue in conceptualizing stress is that research has not always differentiated between the perception of a stimulus or a situation as a stressor and the subsequent biobehavioral response (often called the “stress response”). This is important, since, for example, psychological factors such as uncontrollability and social evaluation, i.e. factors that may influence how an individual perceives a potentially stressful stimulus or situation, have been identified as characteristics that elicit particularly powerful physiological stressful responses (Dickerson and Kemeny 2004 ). At the core of the physiological stress response is a complex physiological system, which is located in both the central nervous system (CNS) and the body´s periphery. The complexity of this system necessitates a multi-dimensional assessment approach involving variables that adequately reflect all relevant components. It is also important to consider that the experience of stress and its psychobiological correlates do not occur in a vacuum, but are being shaped by numerous contextual factors (e.g. societal and cultural context, work and leisure time, family and dyadic systems, environmental variables, physical fitness, nutritional status, etc.) and dispositional factors (e.g. genetics, personality, resilience, regulatory capacities, self-efficacy, etc.). Thus, a theoretical framework needs to incorporate these factors. In sum, as stress is considered a multi-faceted and inherently multi-dimensional construct, its conceptualization and operationalization needs to reflect this (Nater 2018 ).

The goal of the World Association for Stress Related and Anxiety Disorders (WASAD) is to promote and make available basic and clinical research on stress-related and anxiety disorders. Coinciding with WASAD’s 3rd International Congress held in September 2021 in Vienna, Austria, this journal publishes a Special Issue encompassing state-of-the art research in the field of stress and anxiety. This special issue collects answers to a number of important questions that need to be addressed in current and future research. Among the most relevant issues are (1) the multi-dimensional assessment that arises as a consequence of a multi-faceted consideration of stress and anxiety, with a particular focus on doing so under ecologically valid conditions. Skoluda et al. 2021 (in this issue) argue that hair as an important source of the stress hormone cortisol should not only be taken as a complementary stress biomarker by research staff, but that lay persons could be also trained to collect hair at the study participants’ homes, thus increasing the ecological validity of studies incorporating this important measure; (2) the incongruence between psychological and biological facets of stress and anxiety that has been observed both in laboratory and field research (Campbell and Ehlert 2012 ). Interestingly, there are behavioral constructs that do show relatively high congruence. As shown in the paper of Vatheuer et al. ( 2021 ), gaze behavior while exposed to an acute social stressor correlates with salivary cortisol, thus indicating common underlying mechanisms; (3) the complex dynamics of stress-related measures that may extend over shorter (seconds to minutes), medium (hours and diurnal/circadian fluctuations), and longer (months, seasonal) time periods. In particular, momentary assessment studies are highly qualified to examine short to medium term fluctuations and interactions. In their study employing such a design, Stoffel and colleagues (Stoffel et al. 2021 ) show ecologically valid evidence for direct attenuating effects of social interactions on psychobiological stress. Using an experimental approach, on the other hand, Denk et al. ( 2021 ) examined the phenomenon of physiological synchrony between study participants; they found both cortisol and alpha-amylase physiological synchrony in participants who were in the same group while being exposed to a stressor. Importantly, these processes also unfold over time in relation to other biological systems; al’Absi and colleagues showed in their study (al’Absi et al. 2021 ) the critical role of the endogenous opioid system and its relation to stress-related analgesia; (4) the influence of contextual and dispositional factors on the biological stress response in various target samples (e.g., humans, animals, minorities, children, employees, etc.) both under controlled laboratory conditions and in everyday life environments. In this issue, Sattler and colleagues show evidence that contextual information may only matter to a certain extent, as in their study (Sattler et al. 2021 ), the biological response to a gay-specific social stressor was equally pronounced as the one to a general social stressor in gay men. Genetic information is probably the most widely researched dispositional factor; Kuhn et al. show in their paper (Kuhn et al. 2021 ) that the low expression variant of the serotonin transporter gene serves as a risk factor for increased stress reactivity, thus clearly indicating the important role of dispositional factors in stress processing. An interesting factor combining both aspects of dispositional and contextual information is maternal care; Bentele et al. ( 2021 ) in their study are able to show that there was an effect of maternal care on the amylase stress response, while no such effect was observed for cortisol. In a similar vein, Keijser et al. ( 2021 ) showed in their gene-environment interaction study that the effects of FKBP5, a gene very closely related to HPA axis regulation, and early life stress on depressive symptoms among young adults was moderated by a positive parenting style; and (5) the role of stress and anxiety as transdiagnostic factors in mental disorders, be it as an etiological factor, a variable contributing to symptom maintenance, or as a consequence of the condition itself. Stress, e.g., as a common denominator for a broad variety of psychiatric diagnoses has been extensively discussed, and stress as an etiological factor holds specific significance in the context of transdiagnostic approaches to the conceptualization and treatment of mental disorders (Wilamowska et al. 2010 ). The HPA axis, specifically, is widely known to be dysregulated in various conditions. Fischer et al. ( 2021 ) discuss in their comprehensive review the role of this important stress system in the context of patients with post-traumatic disorder. Specifically focusing on the cortisol awakening response, Rausch and colleagues provide evidence for HPA axis dysregulation in patients diagnosed with borderline personality disorder (Rausch et al. 2021 ). As part of a longitudinal project on ADHD, Szep et al. ( 2021 ) investigated the possible impact of child and maternal ADHD symptoms on mothers’ perceived chronic stress and hair cortisol concentration; although there was no direct association, the findings underline the importance of taking stress-related assessments into consideration in ADHD studies. As the HPA axis is closely interacting with the immune system, Rhein et al. ( 2021 ) examined in their study the predicting role of the cytokine IL-6 on psychotherapy outcome in patients with PTSD, indicating that high reactivity of IL-6 to a stressor at the beginning of the therapy was associated with a negative therapy outcome. The review of Kyunghee Kim et al. ( 2021 ) also demonstrated the critical role of immune pathways in the molecular changes due to antidepressant treatment. As for the therapy, the important role of cognitive-behavioral therapy with its key elements to address both stress and anxiety reduction have been shown in two studies in this special issue, evidencing its successful application in obsessive–compulsive disorder (Ivarsson et al. 2021 ; Hollmann et al. 2021 ). Thus, both stress and anxiety are crucial transdiagnostic factors in various mental disorders, and future research needs elaborate further on their role in etiology, maintenance, and treatment.

In conclusion, a number of important questions are being asked in stress and anxiety research, as has become evident above. The Special Issue on “Recent developments in stress and anxiety research” attempts to answer at least some of the raised questions, and I want to invite you to inspect the individual papers briefly introduced above in more detail.

al’Absi M, Nakajima M, Bruehl S (2021) Stress and pain: modality-specific opioid mediation of stress-induced analgesia. J Neural Transm. https://doi.org/10.1007/s00702-021-02401-4

Article   PubMed   Google Scholar  

Bentele UU, Meier M, Benz ABE, Denk BF, Dimitroff SJ, Pruessner JC, Unternaehrer E (2021) The impact of maternal care and blood glucose availability on the cortisol stress response in fasted women. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02350-y

Article   Google Scholar  

Campbell J, Ehlert U (2012) Acute psychosocial stress: does the emotional stress response correspond with physiological responses? Psychoneuroendocrinology 37(8):1111–1134. https://doi.org/10.1016/j.psyneuen.2011.12.010

Denk B, Dimitroff SJ, Meier M, Benz ABE, Bentele UU, Unternaehrer E, Popovic NF, Gaissmaier W, Pruessner JC (2021) Influence of stress on physiological synchrony in a stressful versus non-stressful group setting. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02384-2

Dickerson SS, Kemeny ME (2004) Acute stressors and cortisol responses: a theoretical integration and synthesis of laboratory research. Psychol Bull 130(3):355–391

Fischer S, Schumacher T, Knaevelsrud C, Ehlert U, Schumacher S (2021) Genes and hormones of the hypothalamic-pituitary-adrenal axis in post-traumatic stress disorder. What is their role in symptom expression and treatment response? J Neural Transm (vienna). https://doi.org/10.1007/s00702-021-02330-2

Hollmann K, Allgaier K, Hohnecker CS, Lautenbacher H, Bizu V, Nickola M, Wewetzer G, Wewetzer C, Ivarsson T, Skokauskas N, Wolters LH, Skarphedinsson G, Weidle B, de Haan E, Torp NC, Compton SN, Calvo R, Lera-Miguel S, Haigis A, Renner TJ, Conzelmann A (2021) Internet-based cognitive behavioral therapy in children and adolescents with obsessive compulsive disorder: a feasibility study. J Neural Transm. https://doi.org/10.1007/s00702-021-02409-w

Ivarsson T, Melin K, Carlsson A, Ljungberg M, Forssell-Aronsson E, Starck G, Skarphedinsson G (2021) Neurochemical properties measured by 1 H magnetic resonance spectroscopy may predict cognitive behaviour therapy outcome in paediatric OCD: a pilot study. J Neural Transm. https://doi.org/10.1007/s00702-021-02407-y

Keijser R, Olofsdotter S, Nilsson WK, Åslund C (2021) Three-way interaction effects of early life stress, positive parenting and FKBP5 in the development of depressive symptoms in a general population. J Neural Transm. https://doi.org/10.1007/s00702-021-02405-0

Kuhn L, Noack H, Skoluda N, Wagels L, Rohr AK, Schulte C, Eisenkolb S, Nieratschker V, Derntl B, Habel U (2021) The association of the 5-HTTLPR polymorphism and the response to different stressors in healthy males. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02390-4

Kyunghee Kim H, Zai G, Hennings J, Müller DJ, Kloiber S (2021) Changes in RNA expression levels during antidepressant treatment: a systematic review. J Neural Transm. https://doi.org/10.1007/s00702-021-02394-0

Lazarus RS, Folkman S (1984) Stress, appraisal, and coping. Springer Publisher Company Inc, New York

Google Scholar  

McEwen BS (1998) Protective and damaging effects of stress mediators. N Engl J Med 338(3):171–179

Article   CAS   Google Scholar  

Nater UM (2018) The multidimensionality of stress and its assessment. Brain Behav Immun 73:159–160. https://doi.org/10.1016/j.bbi.2018.07.018

Rausch J, Flach E, Panizza A, Brunner R, Herpertz SC, Kaess M, Bertsch K (2021) Associations between age and cortisol awakening response in patients with borderline personality disorder. J Neural Transm. https://doi.org/10.1007/s00702-021-02402-3

Rhein C, Hepp T, Kraus O, von Majewski K, Lieb M, Rohleder N, Erim Y (2021) Interleukin-6 secretion upon acute psychosocial stress as a potential predictor of psychotherapy outcome in posttraumatic stress disorder. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02346-8

Sattler FA, Nater UM, Mewes R (2021) Gay men’s stress response to a general and a specific social stressor. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02380-6

Skoluda N, Piroth I, Gao W, Nater UM (2021) HOME vs. LAB hair samples for the determination of long-term steroid concentrations: a comparison between hair samples collected by laypersons and trained research staff. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02367-3

Stoffel M, Abbruzzese E, Rahn S, Bossmann U, Moessner M, Ditzen B (2021) Covariation of psychobiological stress regulation with valence and quantity of social interactions in everyday life: disentangling intra- and interindividual sources of variation. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02359-3

Szep A, Skoluda N, Schloss S, Becker K, Pauli-Pott U, Nater UM (2021) The impact of preschool child and maternal attention-deficit/hyperactivity disorder (ADHD) symptoms on mothers’ perceived chronic stress and hair cortisol. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02377-1

Vatheuer CC, Vehlen A, von Dawans B, Domes G (2021) Gaze behavior is associated with the cortisol response to acute psychosocial stress in the virtual TSST. J Neural Transm (Vienna). https://doi.org/10.1007/s00702-021-02344-w

Wilamowska ZA, Thompson-Hollands J, Fairholme CP, Ellard KK, Farchione TJ, Barlow DH (2010) Conceptual background, development, and preliminary data from the unified protocol for transdiagnostic treatment of emotional disorders. Depress Anxiety 27(10):882–890. https://doi.org/10.1002/da.20735

Download references

Author information

Authors and affiliations.

Department of Clinical and Health Psychology, Faculty of Psychology, University of Vienna, Vienna, Austria

Urs M. Nater

University Research Platform ‘The Stress of Life – Processes and Mechanisms Underlying Everyday Life Stress’, University of Vienna, Vienna, Austria

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Urs M. Nater .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Nater, U.M. Recent developments in stress and anxiety research. J Neural Transm 128 , 1265–1267 (2021). https://doi.org/10.1007/s00702-021-02410-3

Download citation

Accepted : 13 August 2021

Published : 01 September 2021

Issue Date : September 2021

DOI : https://doi.org/10.1007/s00702-021-02410-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Find a journal
• Publish with us
• Track your research

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 27 November 2021

Psychological and biological resilience modulates the effects of stress on epigenetic aging

• Zachary M. Harvanek   ORCID: orcid.org/0000-0003-3702-1051 1 ,
• Nia Fogelman 2 ,
• Ke Xu   ORCID: orcid.org/0000-0002-6472-7052 1 , 3 &
• Rajita Sinha   ORCID: orcid.org/0000-0003-3012-4349 1 , 2 , 4 , 5  

Translational Psychiatry volume  11 , Article number:  601 ( 2021 ) Cite this article

23k Accesses

39 Citations

417 Altmetric

Metrics details

Our society is experiencing more stress than ever before, leading to both negative psychiatric and physical outcomes. Chronic stress is linked to negative long-term health consequences, raising the possibility that stress is related to accelerated aging. In this study, we examine whether resilience factors affect stress-associated biological age acceleration. Recently developed “epigenetic clocks” such as GrimAge have shown utility in predicting biological age and mortality. Here, we assessed the impact of cumulative stress, stress physiology, and resilience on accelerated aging in a community sample ( N  = 444). Cumulative stress was associated with accelerated GrimAge ( P  = 0.0388) and stress-related physiologic measures of adrenal sensitivity (Cortisol/ACTH ratio) and insulin resistance (HOMA). After controlling for demographic and behavioral factors, HOMA correlated with accelerated GrimAge ( P  = 0.0186). Remarkably, psychological resilience factors of emotion regulation and self-control moderated these relationships. Emotion regulation moderated the association between stress and aging ( P  = 8.82e−4) such that with worse emotion regulation, there was greater stress-related age acceleration, while stronger emotion regulation prevented any significant effect of stress on GrimAge. Self-control moderated the relationship between stress and insulin resistance ( P  = 0.00732), with high self-control blunting this relationship. In the final model, in those with poor emotion regulation, cumulative stress continued to predict additional GrimAge Acceleration even while accounting for demographic, physiologic, and behavioral covariates. These results demonstrate that cumulative stress is associated with epigenetic aging in a healthy population, and these associations are modified by biobehavioral resilience factors.

Similar content being viewed by others

Associations of stress and stress-related psychiatric disorders with GrimAge acceleration: review and suggestions for future work

Epigenetic aging and perceived psychological stress in old age

The cortisol switch between vulnerability and resilience

Introduction.

Cumulative stress can have adverse psychiatric and physical effects, increasing risk for cardiometabolic diseases, mood disorders, post-traumatic stress disorder and addiction [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 ]. There are several potential psychological and biological mechanisms through which these effects may occur. For example, stress may reduce psychological resilience measures such as emotion regulation and self-control that are known to protect against psychiatric and physical health outcomes [ 1 , 12 , 13 , 14 ]. Notably, emotional stress exposure decreases cognitive and emotion regulation abilities [ 15 , 16 , 17 , 18 ], and this effect may be modulated by cortisol [ 15 ]. Furthermore, stress decreases self-control abilities [ 19 , 20 , 21 ] and impacts the likelihood of individuals engaging in healthy behaviors such as exercise or maintaining a healthy diet, and is associated with unhealthy behaviors such as smoking, alcohol, and drug use [ 22 , 23 , 24 , 25 ]. Recent evidence also suggests that stress effects on metabolic health may be affected by BMI-related changes in insulin resistance and other gut hormones [ 26 , 27 ]. Indeed, stress’s effects on physiology resulting in alterations in neuro-hormonal signaling pathways as well as increased inflammation are well documented [ 26 , 28 , 29 , 30 ]. Both stress and these physiologic changes may increase the risk of multiple physical and psychiatric illnesses, which in turn increase morbidity and mortality risk. This has often been described as an increased allostatic load, and notably many of these processes, such as metabolic and cardiovascular dysfunction, have been associated with human aging [ 31 ]. For example, insulin signaling might be linked to aging and aging-related diseases in humans [ 32 ], with recent data on metformin (a treatment for insulin resistance) suggesting it may be useful as an anti-aging drug [ 33 ].

There is growing evidence that cumulative stress may adversely impact health via accelerating the cellular aging process. For example, stress shortens telomere length and alters telomerase activity, and this interaction is modified by behavioral and psychological resilience factors [ 34 , 35 , 36 , 37 ]. However, recent studies have demonstrated mixed results on whether characteristics that contribute to resilience improve or worsen the impact of stress on health [ 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 ]. These data suggest that resiliency factors may modulate the relationship between chronic stress and aging, but to our knowledge this has not been tested in a healthy community sample. While there are many aspects of resilience, including cultural/societal, environmental, and personal which can decrease the negative consequences of stressors on individuals, herein we will focus on personal-level, psychological skills, including self-control and emotion regulation.

Recently developed DNA methylation-based epigenetic “clocks” appear to provide a more accurate measure of biological age than telomere length [ 48 , 49 , 50 , 51 ]. These clocks are built from a set of DNA methylation markers that correlate with chronologic age and serve as molecular estimators of biological age in cells, tissues, and individuals [ 52 ]. Epigenetic clocks have a significantly higher predictive value than previously used measures such as telomere length for frailty, [ 53 ] mortality risk [ 54 , 55 ], hazard ratios [ 56 ], and chronologic age [ 57 ]. The development of these biological aging markers promises to not only aid in identifying individuals at higher risk for aging-related illnesses, but potentially also developing interventions to prevent accelerated aging.

Previous studies (reviewed by Palma-Gudiel et al [ 58 ]) have utilized epigenetic clocks to demonstrate associations between trauma, early life adversity, or low socioeconomic status and accelerated epigenetic aging. Studies have often been focused upon selected populations, such military veterans [ 45 ], individuals with significant trauma histories [ 59 ], or specific cohorts at higher risk [ 60 , 61 , 62 ]. Notably, these studies did not exclude, and often explicitly included, individuals with significant mental and physical illnesses, including PTSD, MDD, and other disabilities [ 59 , 63 ]. These studies also primarily utilized epigenetic clocks trained upon chronologic age. However, a recently developed epigenetic clock, GrimAge, was trained using biomarkers of mortality and indicators of health, and has superior performance in predicting health outcomes when compared with other epigenetic clocks [ 51 , 64 ].

We utilized GrimAge Acceleration (“GAA”, defined as the residual of the regression of GrimAge to chronologic age, with a positive number indicating biological age greater than chronologic age) to conduct a cross-sectional study to answer three questions. First, is cumulative stress related to epigenetic markers of biological aging in a healthy young-to-middle-aged community population? Second, if stress is associated with epigenetic aging, does stress-related physiology contribute to stress-associated biological aging? And finally, how do psychological factors that contribute to resilience modulate these relationships? Based on previous research, we hypothesized that cumulative stress will be positively associated with GrimAge Acceleration (GAA), that stress effects on GrimAge will be related to changes in the hypothalamic-pituitary-adrenal axis (HPA) and insulin sensitivity, and that strong emotion regulation as measured by the Difficulties in Emotion Regulation Scale (DERS, [ 65 ]) and high self-control as measured by the Self Control Scale-Brief (SCS-B, [ 66 ]) will moderate the relationships between stress, physiology, and accelerated aging (See Fig. 1 for a model summarizing our hypotheses).

We hypothesize that stress is positively associated with accelerated biological aging, which we measure via GrimAge Acceleration (GAA), and that this relationship will be mediated by stress-related physiologic changes such as insulin and HPA signaling. We also hypothesize that strong psychological resilience factors will be protective against the negative consequences of stress on aging. Note that these relationships are predictive, not causative, as this study is cross-sectional and thus directionality of relationships cannot be conclusively examined.

Materials and methods

Cohort recruitment.

The participant cohort included 444 community adults between the ages of 18–50 in the greater New Haven, CT area who volunteered to participate in a study examining the role of stress and self-control at the Yale Stress Center as previously described [ 67 ]. Briefly, participants were recruited via advertisements online, in local newspapers, and at a community center between 2008 and 2012. Participants were excluded if they had a substance use disorder (not including nicotine) as assessed via the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (SCID-I for DSM-IVTR), were pregnant, had a chronic medical condition (e.g, hypertension, diabetes, hypothyroidism), or were unable to read English at or above the 6th grade level. Participants were also excluded if they had a concussion with loss of consciousness greater than 30 minutes, another head injury such as documented traumatic brain injury or another injury with documented lasting deficits, or were using any prescribed medications for any psychiatric or medical disorders. Breathalyzer and urine toxicology screens were conducted at each appointment to ensure the participants were drug-free. Of a total of 1000 potential participants who underwent initial screening for eligibility, epigenetic data combined with physiologic and behavioral data were available on 444, who comprised the current sample. All participants provided written and verbal informed consent to participate, and the research protocol was reviewed and approved by the Yale IRB.

Initial assessment and measurement of physiologic parameters

All eligible subjects met with a research assistant for two intake sessions to complete a physical health review with the Cornell Medical Index (CMI, [ 68 ]), structured clinical interview for diagnoses (SCID) of DSM-IVTR psychiatric illnesses, cumulative stress interview, self-report assessments and a separate morning biochemical evaluation after fasting overnight. The structured clinical interview was performed by masters’ or doctoral level clinical research staff. Fasting insulin and glucose were obtained and Cortisol was assessed at four time-points, spaced 15 min apart beginning at 7:30 AM after overnight fasting and collected while participants were in a quiet and comfortable laboratory setting at the Yale Stress Center. Participants were financially compensated for participating in the study.

Psychological measures

Cumulative stress was assessed using the Cumulative Adversity Inventory (CAI, [ 69 ]), a 140-item multifaceted interview-based assessment of life events and subjective stress through which trained interviewers asked participants about specific stressful events that occurred during their lifetime, which comprised the subscales of major life events, life trauma events and recent life events. For purposes of scoring, a “yes” to the specific stressful event occurring led to a “1” and a sum of all the “yes” endorsements comprised the subscale score for these events subscale. The final subscale of chronic stress was the participant’s own sense of feeling overwhelmed and unable to manage the events for the other subscales listed. This was rated on a “not true”, “somewhat true”, or “very true” scale, with assigned scores of 0, 1, and 2, respectively. The final score is a sum of these values for the chronic stress subscale. The CAI-total score was a sum of each of the subscale score with a higher score indicating a higher overall level of lifetime cumulative stress. The CAI has been demonstrated to have excellent overall reliability as reported in previous research [ 12 , 26 , 70 , 71 , 72 ]. In our population for this study, the alpha reliability is 0.86. It has been previously shown to predict cumulative stress related brain volume reductions and sensitized stress functional responses as well as prediction of physical, metabolic and behavioral responses [ 26 , 70 , 71 , 72 ].

Emotion regulation was assessed using the Difficulties with Emotion Regulation Scale (DERS, [ 65 ]), which is a 41-item trait-level measure that assesses across domains of lack of emotional awareness, goals, clarity, strategies, acceptance, and impulse control in managing emotions. Higher scores on the DERS correspond to lower ability to regulate emotion. Alpha reliability has been reported to be >0.90 for the total score, and ≥0.80 for the sub-scores [ 65 ]. In this population, the alpha reliability is 0.92.

Self-control was assessed using the Self-Control Survey-Brief (SCS-B, [ 66 ]), which is a 13-item scale that assesses overall self-control. A higher score on the SCS-B suggests a stronger level of self-control. There are no sub-scores provided by the SCS-B, and the overall SCS-B has been reported to have an alpha reliability >0.80 [ 66 ]. The alpha reliability in this study is 0.85.

The Cornell Medical Index (CMI) was used to assess for participants’ current health. It is a 195-question interview that captures both physical and psychological health symptoms, and has been validated as an indicator for current general health in many studies [ 68 , 73 , 74 ]. A higher score on the CMI suggests more symptoms and worse overall health. The alpha reliability of the total CMI is 0.94. The psychological subscore has an alpha reliability of 0.92, and the biological subscore has a reliability of 0.90.

Cronbach alpha reliabilities for each of the scales described above were obtained using the alpha function in the R psych package [ 75 ].

DNA methylation and epigenetic clock analysis

DNA for epigenetic analysis was collected from whole blood samples as previously described [ 67 ]. Briefly, all samples were profiled using Illumina Infinium HumanMethylation450 Beadchips, which covers 96% of CpG islands and 99% of RefSeq genes. Quality control on these data are as previously published [ 67 ]. They are described in brief below:

Probe QC : To ensure high-quality data, we set a more stringent threshold of P  < 10 –12 . Intensity values showing P  > 10 −12 were set as zero. Additionally, we removed 11,648 probes on sex chromosomes and 36,535 probes within 10 base pairs of single-nucleotide polymorphisms. Finally, a total of 47,791 probes were removed and the remaining 437,722 probes were used for further analysis.

Sample QC : Using a detection P value < 10 –12 , one sample showing a call rate < 98% was excluded from analysis. Five samples showing sex discrepancy between the methylation predicted sex and self-reported sex were also excluded from analysis.

Data processing and normalization : Data processing and normalization were performed using the recently published protocol (Lehne et al., 2015). We first perform background correction and within-array normalization to the original green/red channel intensity data using the preprocessIllumina function in the minfi R package. The processed data were transformed to M/U methylation categories. Next, we separately performed between-array-normalization with the quantile method using the normalizeBetweenArrays function in the limma R package (version 3.26.2) after dividing the data matrix into 6 independent parts: Type I M Green, Type I M Red, Type I U Red, Type I U Green, Type II Red, Type II Green. The normalized data were merged and the beta value at each CpG site was determined.

After obtaining beta values, epigenetic clock analysis was performed as described in Lu et al. using the New Methylation Age Calculator at https://dnamage.genetics.ucla.edu/new [ 51 ]. Data were normalized as per their protocol, and the advanced analysis option was used. We focus on GrimAge acceleration (GAA), which is defined as the residuals of a linear correlation of GrimAge to chronologic age. No effects of array batch on GAA were observed (Supplementary Fig. 1 ).

The analyses herein were performed without accounting for individual variations in cell types. The Houseman method was used to determine cell type proportion [ 76 ], and the inclusion of cell fractions as covariates in a linear model does not impact the primary conclusions of this paper (see Supplementary material).

Statistical analysis

Data organization and analysis were conducted using R 3.6.3 [ 77 ] and RStudio. Linear regressions were first implemented to examine univariate associations between independent and dependent variables. Multivariable linear regressions adjust for demographic (sex, race, years of education, marital status, income) and behavioral (smoking, alcohol use, and BMI) covariates unless otherwise stated. These covariates were selected due to prior work demonstrating a relationship to epigenetic aging. Chronologic age is incorporated into the model as part of the calculation of GAA (the residual of GrimAge regressed upon chronologic age). There was no significant correlation between chronologic age and GAA. Analyses of the relationship between CAI, GAA, psychological and physiologic variables were performed in both the univariate unadjusted model and the multivariate adjusted model accounting for demographic and behavioral measures, but except when the conclusions differ, statistical values in the text represent the multivariate models for simplicity. CAI, DERS, and SCS were mean-centered to address issues of collinearity (particularly regarding individual regression coefficients) when assessing for moderation.

All tests were two-tailed with alpha set at 0.05. Statistical significance in both standard linear regressions and moderation analyses were assessed from t values. R 2 reported on plots represent the simple relationship between the stated variables, while adjusted R 2 values in the text represent the model. Partial η 2 values represent the effect size for the specific variable, with a value >= 0.01 typically indicating a small effect, >= 0.06 a medium effect, and >= 0.14 a large effect [ 78 ]. Wilcoxon signed-rank test was used to compare data between sexes. Mediation analysis was performed to determine if stress impacts GAA via behavioral and physiologic factors. Simple mediation effects were calculated via R using 10,000 simulations without bootstrapping using the mediation package [ 79 ]. Mediation was considered significant if the proportion mediated was greater than 0 with an alpha of 0.05. Serial mediation was calculated via R using the Lavaan package [ 71 ], with an indirect effect defined as the product of the coefficients of the effect of stress on BMI, of BMI on HOMA, and of HOMA on GAA. Assessment of the individual variables’ attributable GrimAge acceleration as well as confidence intervals were calculated using the Emmeans package using unadjusted pairwise comparisons.

Demographics and clinical characteristics

As shown in Table 1 , study participants were healthy and without evidence of medical or psychiatric diseases. The majority were non-smokers (79.6%), social drinkers with low risky alcohol intake screening scores (72.7% of participants have Alcohol Use Disorders Identification Test (AUDIT) < 8, and 91.7% < 15), and were not obese (74.5% of participants have a BMI < 30, 89.2% < 35). Both physical and psychological symptoms assessed on the Cornell Medical Index (CMI, [ 68 ]) were low, with 86% of participants scoring below the typical screening threshold of 30.

Cumulative stress predicts accelerated biological aging as measured by GrimAge

As expected, there was a high association between individuals’ chronologic age and GrimAge (Age: t  = 51.4, P  < 2e−16, adjusted R 2  = 0.856, Fig. 2A ). This relationship is not altered by inclusion of the covariates of smoking, alcohol use, BMI, race, sex, income, and years of education (Age: t  = 49.1, P  < 2e−16, partial η 2  = 0.848; model (GrimAge ~ Age + covariates) adjusted R 2  = 0.912), and this relationship remained significant accounting for cellular fractions (Supplementary Table 1 ). Also, using a univariate linear regression, greater cumulative stress as measured by the total Cumulative Adversity Index (CAI) score significantly predicted higher GAA (CAI: t  = 4.82  P  = 2.00e−6, η 2  = 0.050, adjusted R 2  = 0.0478, Fig. 2B ). While there were significant differences in GAA based on sex ( P  = 1.33e−7), both males (CAI: P  = 3.35e−4, adjusted R 2  = 0.0586) and females (CAI: P  = 3.12e−5, adjusted R 2  = 0.0652) demonstrated similar correlations between stress and GAA. Further analysis showed these results are consistent across CAI subscales, as well as with the Childhood Trauma Questionnaire and several of its subscales (Supplementary Table 2 ).

A Chronologic age significantly predicts GrimAge ( P  < 2e−16). B Cumulative stress total as measured by the CAI (CAI-Total) significantly predicts GAA before ( P  = 2.00e−6) and after accounting for covariates. C Higher insulin resistance (as measured by HOMA) shows a significant positive correlation with GAA before ( P  = 1.11e−8) and after accounting for covariates. D The Cortisol/ACTH ratio is negatively correlated with GAA before accounting for covariates ( P  = 2.39e−6), but not afterward. P and R 2 values in the figure represent simple univariate models (Y ~ X). In the main text, models are adjusted for covariates as stated.

After accounting for the covariates of smoking, alcohol use, BMI, race, sex, income, and years of education, the relationship between GAA and CAI remains significant (CAI: t  = 2.073, P  = 0.0388, partial η 2  = 0.010; model (GAA ~ CAI-total + covariates): adjusted R 2  = 0.3869); individual covariate effects shown in Supplementary Table 3 ). When considered as potential mediators of the relationship between stress and GAA, BMI (proportion mediated = 0.288, P  = 0.0042) and smoking (proportion mediated = 0.443, P  = 0.0030), but not alcohol use (proportion mediated = 0.001, P  = 0.931), show partial mediating effects (Supplementary Table 4 ).

Consistent with the underlying assumption that GAA is related to measures of health, GAA also predicted psychological and physical health symptoms as measured by the CMI (Supplementary Fig. 2A ; total CMI: t  = 3.449, P  = 6.18e−4, adjusted R 2  = 0.024).

Stress-related physiology is associated with GrimAge acceleration

Given the known relationship between cumulative stress and physiology, we assessed the relationship between the stress-related physiologic factors of insulin resistance and HPA-axis signaling and GAA. We found that higher HOMA (a measure of insulin resistance) significantly predicted GAA (Fig. 2C , HOMA: t  = 2.362, P  = 0.0186, partial η 2  = 0.013; model (GAA ~ HOMA + Covariates): adjusted R 2  = 0.389).

We then assessed whether cortisol/ACTH ratio changes impacted GAA. Indeed, low cortisol/ACTH ratio, a measure of adrenal sensitivity, was associated with GAA in a simple univariate model, (Fig. 2D , Cort/ACTH ratio: t  = −4.78, P  = 2.39e−6, η 2  = 0.049, adjusted R 2  = 0.0470), though this becomes non-significant when accounting for covariates (Cort/ACTH ratio: t  = −0.721, P  = 0.471, partial η 2  = 0.001; model (GAA ~ Cort/ACTH + Covariates): adjusted R 2  = 0.3816). We also find a significant association between stress and Cortisol/ACTH ratio (Supplementary Fig. 2B , CAI: t  = −2.146  P  = 0.0324; model (Cort/ACTH ratio ~ CAI + covariates): adjusted R 2  = 0.2197).

Emotion regulation moderates the relationship between stress and GrimAge acceleration directly

We then asked whether the relationship between cumulative stress and epigenetic aging was modulated by characteristics that contribute to an individual’s psychological resilience. We hypothesized that strong emotion regulation abilities would be protective against stress-related accelerated aging. We found that emotion regulation as assessed by the Difficulties in Emotion Regulation Scale (DERS, [ 65 ]) significantly moderated the relationship between GAA and CAI (Fig. 3A , CAI:DERS: F  = 11.22, P  = 8.82e−4, partial η 2  = 0.025; model (GAA ~ CAI X DERS + covariates): adjusted R 2  = 0.4004), such that poor emotion regulation significantly increased the effects of CAI on GAA. There was not a significant difference between males and females in emotion regulation ( P  = 0.0949).

A Individuals with stronger emotion regulation (as measured by lower DERS scores) suffer less GAA at high stress than individuals with poor emotion regulation before (GAA ~ CAI X DERS P  = 9.51e−5; GAA ~ CAI X DERS + Covariates: P  = 8.82e−4) and after accounting for covariates. For panel A, “good” represents the slope at the 25th percentile of DERS, “fair” at the 50th percentile, and “poor” the 75th percentile. B Better self-control (as measured by higher B-SCS scores) is protective against the effects of stress on GAA before accounting for covariates (GAA ~ CAI X SCS P  = 0.00226; GAA ~ CAI X SCS + Covariates: P  = 0.130), but not after including them in the model. C Stronger self-control moderates the relationship between stress and insulin resistance before (HOMA ~ CAI X SCS P  = 0.0115; HOMA ~ CAI X SCS + Covariates P  = 0.00732) and after accounting for covariates. For panels (B) and (C), “good” represents the slope at the 75th percentile of B-SCS, “fair” at the 50th percentile, and “poor” the 25th percentile.

Self-control moderates the association between stress and insulin resistance, which is associated with GrimAge acceleration

We next assessed whether psychological resilience in the form of self-control (as measured via the SCS-B, [ 66 ]) alters the association between cumulative stress and GAA. We found higher self-control is protective against the effects of stress on GAA before accounting for covariates, but the interaction became non-significant when covariates were accounted for (Fig. 3B , CAI:SCS: F  = 2.303, P  = 0.130, partial η 2  = 0.005; model (GAA ~ CAI X SCS + Covariates: adjusted R 2  = 0.3874).

Given the potential interplay between self-control, insulin resistance, and stress, we next asked whether self-control moderated the relationship between stress and HOMA. We observed that, even when covariates are accounted for, self-control moderates the positive relationship between stress and HOMA, with stronger self-control blunting their relationship (Fig. 3C , CAI:SCS: F = 7.263, P  = 0.00732, partial η 2  = 0.017; model (HOMA ~ CAI X SCS + Covariates: adjusted R 2  = 0.2871). Notably, self-control does not moderate the relationship between CAI and BMI (CAI:SCS: F  = 0.679, P  = 0.41). Self-control did not significantly differ between males and females ( P  = 0.0550).

Exploratory mediation analyses suggest stress influences GrimAge via BMI and HOMA

While our ability to draw causative inferences are limited by the cross-sectional nature of our data, we used mediation analyses to explore potential relationships between weight, insulin resistance, and GAA. We hypothesized that the effects of BMI on GAA may be mediated through insulin resistance. Indeed, mediation analysis suggested that a significant portion of the effect of BMI on GAA may be mediated through HOMA (Supplementary Fig. 3A , proportion mediated = 0.247, P  = 0.02). Given these findings, we next asked whether BMI and insulin resistance act sequentially to mediate the effects of stress on GAA. We identified a significant indirect effect, suggesting that stress may affect GAA through increased BMI and elevated insulin resistance (Supplementary Fig. 3B , indirect effect = 0.003; P  = 0.030), though there continues to be a significant direct effect of stress on GAA as well (direct effect = 0.034, P  = 0.009).

Cumulative stress and estimated change in GrimAge

Finally, we sought to identify the comparative contributions of our significant variables to GAA. To do this, we constructed a linear regression model using all demographic covariates (sex, race, marital status, education, income), behavioral covariates (smoking, alcohol, BMI), physiologic factors (HOMA, Cortisol/ACTH ratio), and psychological factors. In this model, we continue to see a significant interaction between stress and emotion regulation in relation to GAA (CAI:DERS t  = 3.424, P  = 0.000677, partial η 2  = 0.027; model (GAA ~ CAI-total X DERS + HOMA + Cort/ACTH ratio + SCS + Covariates): adjusted R 2  = 0.4056). Notably in this model, HOMA ( t  = 2.308, P  = 0.0215, partial η 2  = 0.012), BMI ( t  = 2.641, P  = 0.00857, partial η 2  = 0.016), and smoking ( t  = 10.47, P  < 2e−16, partial η 2  = 0.204) also demonstrate significant effects on GAA. The impact of the cortisol/ACTH ratio on GAA is not significant ( t  = −0.668, P  = 0.504, partial η 2  = 0.001), and its removal from the model does not impact any of the above conclusions.

Using this final linear model, we estimated the changes in GrimAge for each significant variable (Table 2 ) using estimated marginal means [ 80 ]. When comparing the effects of high stress (CAI-total: 75th percentile) versus low stress (CAI-total: 25th percentile) in those with poor emotion regulation (DERS: 75th percentile), stress was associated with half a year of aging independent of all other covariates and physiologic factors. However, when emotion regulation was strong (DERS: 25th percentile), stress did not independently predict GAA. Again comparing 75th versus 25th percentiles, BMI independently was related to an increase of 0.46 years of GrimAge, and HOMA for ¼ of a year. We also identified daily smoking (3.8 years), male sex (1.2 years), self-identifying as Black (1 year), and never having married (0.71 years) as covariates that significantly predicted accelerated GrimAge. When accounting for cellular fractions we see similar results regarding the relationships between stress, emotion regulation, and GAA. However, when accounting for cellular fractions, the associations between GAA and both HOMA and marital status become non-significant (Supplementary Table 5 ). Prior literature [ 51 ] suggests that GrimAge predicts the hazard ratio exponentially (HR = 1.1 GAA ). Thus, each additional year of GAA would be expected to increase the relative risk of death by approximately 10%.

In this study, we report novel findings that cumulative stress is associated with accelerated epigenetic aging in a healthy, young-to-middle-aged community sample, even after adjusting for sex, race, BMI, smoking, alcohol use, income, marital status, and education. Epigenetic aging was measured by GrimAge, a marker which has previously been associated with increased morbidity and mortality and correlates with physical and psychological health symptoms in our study. The relationship between stress and age acceleration is most prominent in those with poor emotion regulation and was related to behavioral factors such as smoking and BMI. Both stress and GAA were associated with changes in insulin resistance, which was moderated via self-control. These results suggest a relationship between stress, physiology, and accelerated aging that is moderated by emotion regulation and self-control. Overall, these findings point to multiple potentially modifiable biobehavioral targets of intervention that may reduce or prevent the deleterious effects of stress on aging and long-term health outcomes.

This study included a generally healthy, young-to-middle-aged community population, yet we still identified a significant relationship between cumulative stress and age acceleration. The population was taking no prescription medications for any medical conditions, nor were they suffering from current mental illnesses, including major depressive disorder or generalized anxiety disorder. The study includes individuals with obesity, as well as a small number of individuals with risky drinking levels as determined by the AUDIT scores. The frequency of these individuals in the sample is generally in line with those in a community population, and thus we included alcohol use and BMI as covariates to account for the impact of these variables on the results. Prior work has demonstrated that GrimAge better predicts mortality than other epigenetic clocks, and GrimAge predicts lifespan more accurately than self-reporting smoking history, demonstrating that GrimAge is a biologically meaningful and potentially clinically useful biomarker for health [ 51 , 64 ]. Our findings are consistent with recent work showing that those with significant trauma histories [ 59 , 81 ] or with diagnoses of mental illnesses, such as Bipolar disorder or MDD, may experience accelerated aging as measured by epigenetic clocks [ 57 , 81 , 82 , 83 , 84 ]. In particular, this study builds on previous findings by Zannas et al that demonstrated a relationship between trauma and epigenetic aging using the Horvath clock. However, to the best of our knowledge this is the first study to investigate the impact of cumulative stress on epigenetic aging in a healthy community sample without significant physical or mental illness. Also it is the first to our knowledge to identify factors that contribute to psychological resilience as potential modulators of such an effect. This opens the possibility that the distinction between the effects of stress on pathologic and non-pathologic samples may be along a continuum. It would be interesting to examine resilience characteristics in the population studied by Zannas et al to determine if there is a limit to the protective effects of psychological resilience. Thus, preventive interventions that decrease stress and improve resilience may be useful for maintaining long-term mental and physical health.

The relationship between stress and epigenetic aging appears to be modulated via specific psychological traits, including emotion regulation and self-control. Those with better emotion regulation and higher levels of self-control were observed to have less age acceleration even at similar levels of stress. Indeed, based on their GAA, our estimates indicate that the relationship between stress and GrimAge is as powerful as BMI, but only for those with poor emotion regulation. As these are skills that may be developed through specific psychological interventions [ 85 ], these results raise the possibility that building emotion regulation skills could result in improvements in epigenetic aging, morbidity, and mortality [ 86 ] for these populations. As this is a cross-sectional study, we are not able to address whether these relationships are causal. These novel cross-sectional findings provide support for potential future research that may assess whether such an intervention could positively impact epigenetic aging and other indices of long-term health outcomes. Other studies could also examine different aspects of resilience, such as cultural or environmental factors that contribute to resilience to determine if they also are protective against the effects of stress on epigenetic age acceleration. Future studies could also explore other physiologic mechanisms through which psychological resilience may influence epigenetic aging. Based on prior work, inflammation could be particularly important for this relationship. In particular, prior studies have found C-reactive protein [ 87 ] and IL-6 [ 88 ] to be related to emotion regulation and measures of health. The work by Gianaros et al suggests that neurologic activity of the dorsal anterior cingulate cortex may be involved as well.

The relationship between cumulative stress, epigenetic aging, and insulin resistance is of particular note given the prominence of insulin signaling in aging-related pathways [ 89 , 90 ], as well as current trials investigating metformin as a potential anti-aging drug [ 33 ]. In association with this body of work, our study suggests insulin resistance as at least one factor through which stress is associated with accelerated aging, even in a healthy population not suffering from diabetes. As this study is limited by its cross-sectional nature, any causal hypotheses regarding interactions between stress, BMI, insulin resistance, and aging will require longitudinal data to draw specific inferences beyond correlative relationships. Longitudinal studies would also enable prospective assessments of stress, which may be less subject to recall bias based on their current context. This study also identifies the cortisol/ACTH ratio as a potential point of connection between stress and epigenetic aging. However, this measure is somewhat limited in that it reflects an acute measure of the HPA axis, and this relationship becomes non-significant with the inclusion of our covariates. Future studies could utilize other, longer-term measures of HPA axis function such as hair cortisol to better characterize the relationship between stress, epigenetic aging, and the HPA axis.

Nonetheless, this study is the first to identify a clear relationship between cumulative stress and GrimAge acceleration in a healthy population, which suggests stress may play a role in accelerated aging even prior to the onset of chronic diseases. Notably, this relationship was strongly moderated by resilience factors, including self-control and emotion regulation. We also identified smoking, BMI, insulin signaling, and potentially HPA signaling as mediators of this response. However, even when accounting for all these factors as well as demographic covariates such as race, cumulative stress continues to demonstrate a significant impact on GAA, suggesting other mechanisms relating stress to aging not identified herein are also present.

Code availability

R scripts utilized for data analysis are available by contacting the authors directly.

Roy B, Riley C, Sinha R. Emotion regulation moderates the association between chronic stress and cardiovascular disease risk in humans: a cross-sectional study. Stress. 2018:1-8, https://doi.org/10.1080/10253890.2018.1490724 .

Boehm JK, Kubzansky LD. The heart’s content: the association between positive psychological well-being and cardiovascular health. Psychol Bull. 2012:138;655-691.

Sampasa-Kanyinga H, Chaput J-P. Associations among self-perceived work and life stress, trouble sleeping, physical activity, and body weight among Canadian adults. Preventive Med. 2017;96:16–20. https://doi.org/10.1016/j.ypmed.2016.12.013 .

Article   Google Scholar  

Kelly SJ, Ismail M. Stress and type 2 diabetes: a review of how stress contributes to the development of type 2 diabetes. Annu Rev Public Health. 2015;36:441–62. https://doi.org/10.1146/annurev-publhealth-031914-122921 .

Article   PubMed   Google Scholar  

Liu MY, Li N, Li WA, Khan H. Association between psychosocial stress and hypertension: a systematic review and meta-analysis. Neurol Res. 2017;39:573–80. https://doi.org/10.1080/01616412.2017.1317904 .

Halaris A. Inflammation-associated co-morbidity between depression and cardiovascular disease. Curr Top Behav Neurosci. 2017;31:45–70. https://doi.org/10.1007/7854_2016_28 .

Article   CAS   PubMed   Google Scholar  

Joseph JJ, Golden SH. Cortisol dysregulation: the bidirectional link between stress, depression, and type 2 diabetes mellitus. Ann N. Y Acad Sci. 2017;1391:20–34. https://doi.org/10.1111/nyas.13217 .

Tsounis D, Bouras G, Giannopoulos G, Papadimitriou C, Alexopoulos D, Deftereos S. Inflammation markers in essential hypertension. Med Chem. 2014;10:672–81. https://doi.org/10.2174/1573406410666140318111328 .

Silverman MN, Sternberg EM. Glucocorticoid regulation of inflammation and its functional correlates: from HPA axis to glucocorticoid receptor dysfunction. Ann N. Y Acad Sci. 2012;1261:55–63. https://doi.org/10.1111/j.1749-6632.2012.06633.x .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Miller R, Kirschbaum C. Cultures under stress: A cross-national meta-analysis of cortisol responses to the Trier Social Stress Test and their association with anxiety-related value orientations and internalizing mental disorders. Psychoneuroendocrinology. 2019;105:147–54. https://doi.org/10.1016/j.psyneuen.2018.12.236 .

Giacco D, Laxhman N, Priebe S. Prevalence of and risk factors for mental disorders in refugees. Semin Cell Dev Biol. 2018;77:144–52. https://doi.org/10.1016/j.semcdb.2017.11.030 .

Abravanel BT, Sinha R. Emotion dysregulation mediates the relationship between lifetime cumulative adversity and depressive symptomatology. J Psychiatr Res. 2015;61:89–96. https://doi.org/10.1016/j.jpsychires.2014.11.012 .

Wolff, M, Enge, S, Kräplin, A, Krönke, KM, Bühringer, G, Smolka, MN et al. Chronic stress, executive functioning, and real-life self-control: an experience sampling study. J Pers. 2020, https://doi.org/10.1111/jopy.12587 .

Duckworth AL, Kim B, Tsukayama E. Life stress impairs self-control in early adolescence. Front Psychol. 2012;3:608 https://doi.org/10.3389/fpsyg.2012.00608 .

Lewis EJ, Yoon KL, Joormann J. Emotion regulation and biological stress responding: associations with worry, rumination, and reappraisal. Cogn Emot. 2018;32:1487–98. https://doi.org/10.1080/02699931.2017.1310088 .

Raio CM, Orederu TA, Palazzolo L, Shurick AA, Phelps EA. Cognitive emotion regulation fails the stress test. Proc Natl Acad Sci USA. 2013;110:15139–44. https://doi.org/10.1073/pnas.1305706110 .

Article   PubMed   PubMed Central   Google Scholar  

Sinha R. How does stress increase risk of drug abuse and relapse? Psychopharmacol (Berl). 2001;158:343–59. https://doi.org/10.1007/s002130100917 .

Article   CAS   Google Scholar  

Sinha R. Chronic stress, drug use, and vulnerability to addiction. Ann N. Y Acad Sci. 2008;1141:105–30. https://doi.org/10.1196/annals.1441.030 .

Baumeister RF, Bratslavsky E, Muraven M, Tice DM. Ego depletion: is the active self a limited resource? J Pers Soc Psychol. 1998;74:1252–65. https://doi.org/10.1037//0022-3514.74.5.1252 .

Maier SU, Makwana AB, Hare TA. Acute stress impairs self-control in goal-directed choice by altering multiple functional connections within the brain’s decision circuits. Neuron. 2015;87:621–31. https://doi.org/10.1016/j.neuron.2015.07.005 .

Muraven M, Baumeister RF. Self-regulation and depletion of limited resources: does self-control resemble a muscle? Psychol Bull. 2000;126:247–59. https://doi.org/10.1037/0033-2909.126.2.247 .

Beutel TF, Zwerenz R, Michal M. Psychosocial stress impairs health behavior in patients with mental disorders. BMC Psychiatry. 2018;18:375 https://doi.org/10.1186/s12888-018-1956-8 .

Wemm SE, Sinha R. Drug-induced stress responses and addiction risk and relapse. Neurobiol Stress. 2019;10:100148 https://doi.org/10.1016/j.ynstr.2019.100148 .

Kwarteng JL, Schulz AJ, Mentz GB, Israel BA, Perkins DW. Independent effects of neighborhood poverty and psychosocial stress on obesity over time. J Urban Health. 2017;94:791–802. https://doi.org/10.1007/s11524-017-0193-7 .

Stults-Kolehmainen MA, Sinha R. The effects of stress on physical activity and exercise. Sports Med. 2014;44:81–121. https://doi.org/10.1007/s40279-013-0090-5 .

Chao AM, Jastreboff AM, White MA, Grilo CM, Sinha R. Stress, cortisol, and other appetite-related hormones: Prospective prediction of 6-month changes in food cravings and weight. Obes (Silver Spring). 2017;25:713–20. https://doi.org/10.1002/oby.21790 .

Sinha R, Jastreboff AM. Stress as a common risk factor for obesity and addiction. Biol Psychiatry. 2013;73:827–35. https://doi.org/10.1016/j.biopsych.2013.01.032 .

Sinha R. Role of addiction and stress neurobiology on food intake and obesity. Biol Psychol. 2018;131:5–13. https://doi.org/10.1016/j.biopsycho.2017.05.001 .

Wirtz PH, von Känel R. Psychological stress, inflammation, and coronary heart disease. Curr Cardiol Rep. 2017;19:111 https://doi.org/10.1007/s11886-017-0919-x .

Lavretsky H, Newhouse PA. Stress, inflammation, and aging. Am J Geriatr Psychiatry. 2012;20:729–33. https://doi.org/10.1097/JGP.0b013e31826573cf .

Edes AN, Crews DE. Allostatic load and biological anthropology. Am J Phys Anthropol. 2017;162:44–70. https://doi.org/10.1002/ajpa.23146 . Suppl 63 .

Costantino S, Paneni F, Cosentino F. Ageing, metabolism and cardiovascular disease. J Physiol. 2016;594:2061–73. https://doi.org/10.1113/JP270538 .

Barzilai N, Crandall JP, Kritchevsky SB, Espeland MA. Metformin as a tool to target aging. Cell Metab. 2016;23:1060–5. https://doi.org/10.1016/j.cmet.2016.05.011 .

Mason AE, Hecht FM, Daubenmier JJ, Sbarra DA, Lin J, Moran PJ, et al. Weight loss maintenance and cellular aging in the supporting health through nutrition and exercise study. Psychosom Med. 2018;80:609–19. https://doi.org/10.1097/psy.0000000000000616 .

Puterman E, Lin J, Blackburn E, O’Donovan A, Adler N, Epel E. The power of exercise: buffering the effect of chronic stress on telomere length. PLoS ONE. 2010;5:e10837 https://doi.org/10.1371/journal.pone.0010837 .

Ornish D, Lin J, Chan JM, Epel E, Kemp C, Weidner G, et al. Effect of comprehensive lifestyle changes on telomerase activity and telomere length in men with biopsy-proven low-risk prostate cancer: 5-year follow-up of a descriptive pilot study. Lancet Oncol. 2013;14:1112–20. https://doi.org/10.1016/s1470-2045(13)70366-8 .

Puterman E, Epel ES, Lin J, Blackburn EH, Gross JJ, Whooley MA, et al. Multisystem resiliency moderates the major depression-telomere length association: findings from the Heart and Soul Study. Brain Behav Immun. 2013;33:65–73. https://doi.org/10.1016/j.bbi.2013.05.008 .

Osório C, Probert T, Jones E, Young AH, Robbins I. Adapting to stress: understanding the neurobiology of resilience. Behav Med. 2017;43:307–22. https://doi.org/10.1080/08964289.2016.1170661 .

Sandifer PA, Walker AH. Enhancing disaster resilience by reducing stress-associated health impacts. Front Public Health. 2018;6:373 https://doi.org/10.3389/fpubh.2018.00373 .

Kennedy B, Fang F, Valdimarsdóttir U, Udumyan R, Montgomery S, Fall K. Stress resilience and cancer risk: a nationwide cohort study. J Epidemiol Community Health. 2017;71:947–53. https://doi.org/10.1136/jech-2016-208706 .

Bergh C, Udumyan R, Fall K, Almroth H, Montgomery S. Stress resilience and physical fitness in adolescence and risk of coronary heart disease in middle age. Heart. 2015;101:623–9. https://doi.org/10.1136/heartjnl-2014-306703 .

Bergh C, Udumyan R, Fall K, Nilsagård Y, Appelros P, Montgomery S. Stress resilience in male adolescents and subsequent stroke risk: cohort study. J Neurol Neurosurg Psychiatry. 2014;85:1331–6. https://doi.org/10.1136/jnnp-2013-307485 .

Felix AS, Lehman A, Nolan TS, Sealy-Jefferson S, Breathett K, Hood DB, et al. Stress, resilience, and cardiovascular disease risk among black women. Circ Cardiovasc Qual Outcomes. 2019;12:e005284 https://doi.org/10.1161/circoutcomes.118.005284 .

Mehta D, Bruenig D, Lawford B, Harvey W, Carrillo-Roa T, Morris CP, et al. Accelerated DNA methylation aging and increased resilience in veterans: the biological cost for soldiering on. Neurobiol Stress. 2018;8:112–9. https://doi.org/10.1016/j.ynstr.2018.04.001 .

Boks MP, van Mierlo HC, Rutten BP, Radstake TR, De Witte L, Geuze E, et al. Longitudinal changes of telomere length and epigenetic age related to traumatic stress and post-traumatic stress disorder. Psychoneuroendocrinology. 2015;51:506–12. https://doi.org/10.1016/j.psyneuen.2014.07.011 .

James SA. John Henryism and the health of African-Americans. Cult Med Psychiatry. 1994;18:163–82. https://doi.org/10.1007/BF01379448 .

Gupta S, Belanger E, Phillips SP. Low socioeconomic status but resilient: panacea or double trouble? John Henryism in the International IMIAS Study of Older Adults. J Cross Cult Gerontol. 2019;34:15–24. https://doi.org/10.1007/s10823-018-9362-8 .

Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14:R115 https://doi.org/10.1186/gb-2013-14-10-r115 .

Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nat Rev Genet. 2018;19:371–84. https://doi.org/10.1038/s41576-018-0004-3 .

Levine ME, Lu AT, Quach A, Chen BH, Assimes TL, Bandinelli S, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018;10:573–91. https://doi.org/10.18632/aging.101414 .

Lu AT, Quach A, Wilson JG, Reiner AP, Aviv A, Raj K, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019;11:303–27. https://doi.org/10.18632/aging.101684 .

Bell CG, Lowe R, Adams PD, Baccarelli AA, Beck S, Bell JT, et al. DNA methylation aging clocks: challenges and recommendations. Genome Biol. 2019;20:249 https://doi.org/10.1186/s13059-019-1824-y .

Breitling LP, Saum KU, Perna L, Schöttker B, Holleczek B, Brenner H. Frailty is associated with the epigenetic clock but not with telomere length in a German cohort. Clin Epigenetics. 2016;8:21 https://doi.org/10.1186/s13148-016-0186-5 .

Gao X, Zhang Y, Mons U, Brenner H. Leukocyte telomere length and epigenetic-based mortality risk score: associations with all-cause mortality among older adults. Epigenetics. 2018;13:846–57. https://doi.org/10.1080/15592294.2018.1514853 .

Marioni RE, Harris SE, Shah S, McRae AF, von Zglinicki T, Martin-Ruiz C, et al. The epigenetic clock and telomere length are independently associated with chronological age and mortality. Int J Epidemiol. 2018;45:424–32. https://doi.org/10.1093/ije/dyw041 .

Jylhävä J, Pedersen NL, Hägg S. Biological age predictors. EBioMedicine. 2017;21:29–36. https://doi.org/10.1016/j.ebiom.2017.03.046 .

Fries GR, Bauer IE, Scaini G, Wu MJ, Kazimi IF, Valvassori SS, et al. Accelerated epigenetic aging and mitochondrial DNA copy number in bipolar disorder. Transl Psychiatry. 2017;7:1283 https://doi.org/10.1038/s41398-017-0048-8 .

Palma-Gudiel H, Fananas L, Horvath S, Zannas AS. Psychosocial stress and epigenetic aging. Int Rev Neurobiol. 2020;150:107–28. https://doi.org/10.1016/bs.irn.2019.10.020 .

Zannas AS, Arloth J, Carrillo-Roa T, Iurato S, Röh S, Ressler KJ, et al. Lifetime stress accelerates epigenetic aging in an urban, African American cohort: relevance of glucocorticoid signaling. Genome Biol. 2015;16:266 https://doi.org/10.1186/s13059-015-0828-5 .

Simons RL, Lei MK, Beach SR, Philibert RA, Cutrona CE, Gibbons FX, et al. Economic hardship and biological weathering: The epigenetics of aging in a U.S. sample of black women. Soc Sci Med. 2016;150:192–200. https://doi.org/10.1016/j.socscimed.2015.12.001 .

Chen E, Miller GE, Yu T, Brody GH. The great recession and health risks in African American youth. Brain Behav Immun. 2016;53:234–41. https://doi.org/10.1016/j.bbi.2015.12.015 .

Brody GH, Miller GE, Yu T, Beach SR, Chen E. Supportive family environments ameliorate the link between racial discrimination and epigenetic aging: a replication across two longitudinal cohorts. Psychol Sci. 2016;27:530–41. https://doi.org/10.1177/0956797615626703 .

Wolf EJ, Maniates H, Nugent N, Maihofer AX, Armstrong D, Ratanatharathorn A, et al. Traumatic stress and accelerated DNA methylation age: A meta-analysis. Psychoneuroendocrinology. 2018;92:123–34. https://doi.org/10.1016/j.psyneuen.2017.12.007 .

McCrory C, Fiorito G, Hernandez B, Polidoro S, O’Halloran AM, Hever A, et al. GrimAge outperforms other epigenetic clocks in the prediction of age-related clinical phenotypes and all-cause mortality. J Gerontol A Biol Sci Med Sci. 2020. https://doi.org/10.1093/gerona/glaa286 .

Gratz KL, Roemer L. Multidimensional assessment of emotion regulation and dysregulation: development, factor structure, and initial validation of the difficulties in Emotion Regulation Scale. J Psychopathol Behav Assess. 2004;26:41–54. https://doi.org/10.1023/B:JOBA.0000007455.08539.94 .

Tangney JP, Baumeister RF, Boone AL. High self-control predicts good adjustment, less pathology, better grades, and interpersonal success. J Personal. 2004;72:271–324. https://doi.org/10.1111/j.0022-3506.2004.00263.x .

Xu K, Zhang X, Wang Z, Hu Y, Sinha R. Epigenome-wide association analysis revealed that SOCS3 methylation influences the effect of cumulative stress on obesity. Biol Psychol. 2018;131:63–71. https://doi.org/10.1016/j.biopsycho.2016.11.001 .

Brodman K, Erdmann AJ Jr, Lorge I, Wolff HG, Broadbent TH. The Cornell medical index; a adjunct to medical interview. J Am Med Assoc. 1949;140:530–4. https://doi.org/10.1001/jama.1949.02900410026007 .

Turner RJ, Wheaton B, Lloyd DA. The epidemiology of social stress. Am Sociological Rev. 1995;60:104–25. https://doi.org/10.2307/2096348 .

Ansell EB, Gu P, Tuit K, Sinha R. Effects of cumulative stress and impulsivity on smoking status. Hum Psychopharmacol. 2012;27:200–8. https://doi.org/10.1002/hup.1269 .

Rosseel Y. lavaan: an R package for structural equation modeling. 2012. 2012;48:36 https://doi.org/10.18637/jss.v048.i02 .

Seo D, Tsou KA, Ansell EB, Potenza MN, Sinha R. Cumulative adversity sensitizes neural response to acute stress: association with health symptoms. Neuropsychopharmacology. 2014;39:670–80. https://doi.org/10.1038/npp.2013.250 .

Perlmutter M, Nyquist L. Relationships between self-reported physical and mental health and intelligence performance across adulthood. J Gerontol. 1990;45:P145–155. https://doi.org/10.1093/geronj/45.4.p145 .

Abramson JH. The cornell medical index as an epidemiological tool. Am J Public Health Nations Health. 1966;56:287–98. https://doi.org/10.2105/ajph.56.2.287 .

Revelle W. psych: Procedures for Psychological, Psychometric, and Personality Research, 2021. https://CRAN.R-project.org/package=psych .

Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ, Nelson HH, et al. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinform. 2012;13:86 https://doi.org/10.1186/1471-2105-13-86 .

R Foundation for Statistical Computing. R: a language and environment for statistical computing. R Foundation for Statistical Computing; 2020.

Richardson JTE. Eta squared and partial eta squared as measures of effect size in educational research. Educ Res Rev. 2011;6:135–47. https://doi.org/10.1016/j.edurev.2010.12.001 .

Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation: R package for causal mediation analysis. 2014. 2014;59:38 https://doi.org/10.18637/jss.v059.i05 .

Searle SR, Speed FM, Milliken GA. Population marginal means in the linear model: an alternative to least squares means. Am Statistician. 1980;34:216–21. https://doi.org/10.1080/00031305.1980.10483031 .

Wolf EJ, Morrison FG. Traumatic stress and accelerated cellular aging: from epigenetics to cardiometabolic disease. Curr Psychiatry Rep. 2017;19:75 https://doi.org/10.1007/s11920-017-0823-5 .

Yang, R, Wu, GWY, Verhoeven, JE, Gautam, A, Reus, VI, Kang, JI et al. A DNA methylation clock associated with age-related illnesses and mortality is accelerated in men with combat PTSD. Mol Psychiatry. 2020. https://doi.org/10.1038/s41380-020-0755-z .

Squassina, A, Pisanu, C & Vanni, R mood disorders, accelerated aging, and inflammation: is the link hidden in telomeres? Cells. 2019:8, https://doi.org/10.3390/cells8010052 .

Higgins-Chen AT, Boks MP, Vinkers CH, Kahn RS, Levine ME. Schizophrenia and epigenetic aging biomarkers: increased mortality, reduced cancer risk, and unique clozapine effects. Biol Psychiatry. 2020, https://doi.org/10.1016/j.biopsych.2020.01.025 .

Guendelman S, Medeiros S, Rampes H. Mindfulness and emotion regulation: insights from neurobiological, psychological, and clinical studies. Front Psychol. 2017;8:220–220. https://doi.org/10.3389/fpsyg.2017.00220 .

Roy B, Riley C, Sinha R. Emotion regulation moderates the association between chronic stress and cardiovascular disease risk in humans: a cross-sectional study. Stress (Amst, Neth). 2018;21:548–55. https://doi.org/10.1080/10253890.2018.1490724 .

Appleton AA, Buka SL, Loucks EB, Gilman SE, Kubzansky LD. Divergent associations of adaptive and maladaptive emotion regulation strategies with inflammation. Health Psychol: Off J Div Health Psychol, Am Psychological Assoc. 2013;32:748–56. https://doi.org/10.1037/a0030068 .

Gianaros PJ, Marsland AL, Kuan DCH, Schirda BL, Jennings JR, Sheu LK, et al. An inflammatory pathway links atherosclerotic cardiovascular disease risk to neural activity evoked by the cognitive regulation of emotion. Biol psychiatry. 2014;75:738–45. https://doi.org/10.1016/j.biopsych.2013.10.012 .

Cabreiro F, Au C, Leung KY, Vergara-Irigaray N, Cocheme HM, Noori T, et al. Metformin retards aging in C. elegans by altering microbial folate and methionine metabolism. Cell. 2013;153:228–39. https://doi.org/10.1016/j.cell.2013.02.035 .

Martin-Montalvo A, Mercken EM, Mitchell SJ, Palacios HH, Mote PL, Scheibye-Knudsen M, et al. Metformin improves healthspan and lifespan in mice. Nat Commun. 2013;4:2192 https://doi.org/10.1038/ncomms3192 .

Download references

Acknowledgements

The authors would like to acknowledge the Yale Center of Genome Analysis for DNA methylation profiling. Funding for this study is from NIH Common Fund UL1-DE019586 (R.S.), PL1-DA24859 (R.S.), R01-AA013892 (R.S.), NIH R01DA047063 (K.X.), NIH T32MH019961 (Z.M.H.), NIH R25MH071584 (Z.M.H.). These data were presented at the SOBP virtual conference in April 2021 as a poster.

Author information

Authors and affiliations.

Department of Psychiatry, Yale University, New Haven, CT, USA

Zachary M. Harvanek, Ke Xu & Rajita Sinha

Yale Stress Center, Yale University, New Haven, CT, USA

Nia Fogelman & Rajita Sinha

Department of Psychiatry, Connecticut Veteran Healthcare System, West Haven, CT, USA

Department of Neuroscience, Yale University, New Haven, CT, USA

Rajita Sinha

Child Study Center, Yale University, New Haven, CT, USA

You can also search for this author in PubMed   Google Scholar

Contributions

Z.M.H., K.X., and R.S. conceptualized the project. Z.M.H. and N.F. performed the data analysis, with recommendations from K.X. and R.S. Z.M.H. produced the figures and tables. Z.M.H. wrote the manuscript, and all authors contributed to and edited the manuscript.

Corresponding author

Correspondence to Rajita Sinha .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary material, supplementary figure 1, supplementary figure 2, supplementary figure 3, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Harvanek, Z.M., Fogelman, N., Xu, K. et al. Psychological and biological resilience modulates the effects of stress on epigenetic aging. Transl Psychiatry 11 , 601 (2021). https://doi.org/10.1038/s41398-021-01735-7

Download citation

Received : 28 June 2021

Revised : 31 October 2021

Accepted : 10 November 2021

Published : 27 November 2021

DOI : https://doi.org/10.1038/s41398-021-01735-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Self-control is associated with health-relevant disparities in buccal dna-methylation measures of biological aging in older adults.

• Y. E. Willems
• A. deSteiguer
• Laurel Raffington

Clinical Epigenetics (2024)

Mind body medicine: a modern bio-psycho-social model forty-five years after Engel

• Gregory Fricchione

BioPsychoSocial Medicine (2023)

• Ekaterina Protsenko
• Owen M. Wolkowitz
• Kristine Yaffe

Translational Psychiatry (2023)

Associations between psychological resilience and epigenetic clocks in the health and retirement study

• Aijie Zhang
• Yiqiang Zhan

GeroScience (2023)

Dimensions of childhood adversity differentially affect biological aging in major depression

• Ryan Rampersaud

Translational Psychiatry (2022)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

• Research Article
• Open access
• Published: 06 April 2021

Health anxiety, perceived stress, and coping styles in the shadow of the COVID-19

• Szabolcs Garbóczy 1 , 2 ,
• Anita Szemán-Nagy 3 ,
• Mohamed S. Ahmad 4 ,
• Szilvia Harsányi 1 ,
• Dorottya Ocsenás 5 , 6 ,
• Viktor Rekenyi 4 ,
• Ala’a B. Al-Tammemi 1 , 7 &
• László Róbert Kolozsvári   ORCID: orcid.org/0000-0001-9426-0898 1 , 7  

BMC Psychology volume  9 , Article number:  53 ( 2021 ) Cite this article

16k Accesses

41 Citations

Metrics details

In the case of people who carry an increased number of anxiety traits and maladaptive coping strategies, psychosocial stressors may further increase the level of perceived stress they experience. In our research study, we aimed to examine the levels of perceived stress and health anxiety as well as coping styles among university students amid the COVID-19 pandemic.

A cross-sectional study was conducted using an online-based survey at the University of Debrecen during the official lockdown in Hungary when dormitories were closed, and teaching was conducted remotely. Our questionnaire solicited data using three assessment tools, namely, the Perceived Stress Scale (PSS), the Ways of Coping Questionnaire (WCQ), and the Short Health Anxiety Inventory (SHAI).

A total of 1320 students have participated in our study and 31 non-eligible responses were excluded. Among the remaining 1289 participants, 948 (73.5%) and 341 (26.5%) were Hungarian and international students, respectively. Female students predominated the overall sample with 920 participants (71.4%). In general, there was a statistically significant positive relationship between perceived stress and health anxiety. Health anxiety and perceived stress levels were significantly higher among international students compared to domestic ones. Regarding coping, wishful thinking was associated with higher levels of stress and anxiety among international students, while being a goal-oriented person acted the opposite way. Among the domestic students, cognitive restructuring as a coping strategy was associated with lower levels of stress and anxiety. Concerning health anxiety, female students (domestic and international) had significantly higher levels of health anxiety compared to males. Moreover, female students had significantly higher levels of perceived stress compared to males in the international group, however, there was no significant difference in perceived stress between males and females in the domestic group.

The elevated perceived stress levels during major life events can be further deepened by disengagement from home (being away/abroad from country or family) and by using inadequate coping strategies. By following and adhering to the international recommendations, adopting proper coping methods, and equipping oneself with the required coping and stress management skills, the associated high levels of perceived stress and anxiety could be mitigated.

Peer Review reports

Introduction

On March 4, 2020, the first cases of coronavirus disease were declared in Hungary. One week later, the World Health Organization (WHO) declared COVID-19 as a global pandemic [ 1 ]. The Hungarian government ordered a ban on outdoor public events with more than 500 people and indoor events with more than 100 participants to reduce contact between people [ 2 ]. On March 27, the government imposed a nationwide lockdown for two weeks effective from March 28, to mitigate the spread of the pandemic. Except for food stores, drug stores, pharmacies, and petrol stations, all other shops and educational institutions remained closed. On April 16, a week-long extension was further announced [ 3 ].

The COVID-19 pandemic with its high morbidity and mortality has already afflicted the psychological and physical wellbeing of humans worldwide [ 4 , 5 , 6 , 7 , 8 , 9 ]. During major life events, people may have to deal with more stress. Stress can negatively affect the population’s well-being or function when they construe the situation as stressful and they cannot handle the environmental stimuli [ 10 ]. Various inter-related and inter-linked concepts are present in such situations including stress, anxiety, and coping. According to the literature, perceived stress can lead to higher levels of anxiety and lower levels of health-related quality of life [ 11 ]. Another study found significant and consistent associations between coping strategies and the dimensions of health anxiety [ 12 ].

Health anxiety is one of the most common types of anxiety and it describes how people think and behave toward their health and how they perceive any health-related concerns or threats. Health anxiety is increasingly conceptualized as existing on a spectrum [ 13 , 14 ], and as an adaptive signal that helps to develop survival-oriented behaviors. It also occurs in almost everyone’s life to a certain degree and can be rather deleterious when it is excessive [ 13 , 14 ]. Illness anxiety or hypochondriasis is on the high end of the spectrum and it affects someone’s life when it interferes with daily life by making people misinterpret the somatic sensations, leading them to think that they have an underlying condition [ 14 ].

According to the American Psychiatric Association—Diagnostic and Statistical Manual of Mental Disorders (fifth edition), Illness anxiety disorder is described as a preoccupation with acquiring or having a serious illness, and it reflects the high spectrum of health anxiety [ 15 ]. Somatic symptoms are not present or if they are, then only mild in intensity. The preoccupation is disproportionate or excessive if there is a high risk of developing a medical condition (e.g., family history) or the patient has another medical condition. Excessive health-related behaviors can be observed (e.g., checking body for signs of illness) and individuals can show maladaptive avoidance as well by avoiding hospitals and doctor appointments [ 15 ].

Health anxiety is indeed an important topic as both its increase and decrease can progress to problems [ 14 ]. Looking at health anxiety as a wide spectrum, it can be high or low [ 16 ]. While people with a higher degree of worry and checking behaviors may cause some burden on healthcare facilities by visiting them too many times (e.g., frequent unnecessary visits), other individuals may not seek medical help at healthcare units to avoid catching up infections for instance. A lower degree of health anxiety can lead to low compliance with imposed regulations made to control a pandemic [ 17 ].

The COVID-19 pandemic as a major event in almost everyone’s life has posed a great impact on the population’s perceived stress level. Several studies about the relation between coping and response to epidemics in recent and previous outbreaks found higher perceived stress levels among people [ 18 , 19 , 20 , 21 ]. Being a woman, low income, and living with other people all were associated with higher stress levels [ 18 ]. Protective factors like being emotionally more stable, having self-control, adaptive coping strategies, and internal locus of control were also addressed [ 19 , 20 ]. The findings indicated that the COVID-19 crisis is perceived as a stressful event. The perceived stress was higher amongst people than it was in situations with no emergency. Nervousness, stress, and loss of control of one’s life are the factors that are most connected to perceived stress levels which leads to the suggestion that unpredictability and uncontrollability take an important part in perceived stress during a crisis [ 19 , 20 ].

Moreover, certain coping styles (e.g., having a positive attitude) were associated with less psychological distress experiences but avoidance strategies were more likely to cause higher levels of stress [ 21 ]. According to Lazarus (1999), individuals differ in their perception of stress if the stress response is viewed as the interaction between the environment and humans [ 22 ]. An Individual can experience two kinds of evaluation processes, one to appraise the external stressors and personal stake, and the other one to appraise personal resources that can be used to cope with stressors [ 22 , 23 ]. If there is an imbalance between these two evaluation processes, then stress occurs, because the personal resources are not enough to cope with the stressor’s demands [ 23 ].

During stressful life events, it is important to pay attention to the increasing levels of health anxiety and to the kind of coping mechanisms that are potential factors to mitigate the effects of high anxiety. The transactional model of stress by Lazarus and Folkman (1987) provides an insight into these kinds of factors [ 24 ]. Lazarus and Folkman theorized two types of coping responses: emotion-focused coping, and problem-focused coping. Emotion-focused coping strategies (e.g., distancing, acceptance of responsibility, positive reappraisal) might be used when the source of stress is not embedded in the person’s control and these strategies aim to manage the individual’s emotional response to a threat. Also, emotion-focused coping strategies are directed at managing emotional distress [ 24 ]. On the other hand, problem-focused coping strategies (e.g., confrontive coping, seeking social support, planful problem-solving) help an individual to be able to endure and/or minimize the threat, targeting the causes of stress in practical ways [ 24 ]. It was also addressed that emotion-focused coping mechanisms were used more in situations appraised as requiring acceptance, whereas problem-focused forms of coping were used more in encounters assessed as changeable [ 24 ].

A recent study in Hunan province in China found that the most effective factor in coping with stress among medical staff was the knowledge of their family’s well-being [ 25 ]. Although there have been several studies about the mental health of hospital workers during the COVID-19 pandemic or other epidemics (e.g., SARS, MERS) [ 26 , 27 , 28 , 29 ], only a few studies from recent literature assessed the general population’s coping strategies. According to Gerhold (2020) [ 30 ], older people perceived a lower risk of COVID-19 than younger people. Also, women have expressed more worries about the disease than men did. Coping strategies were highly problem-focused and most of the participants reported that they listen to professionals’ advice and tried to remain calm [ 30 ]. In the same study, most responders perceived the COVID-19 pandemic as a global catastrophe that will severely affect a lot of people. On the other hand, they perceived the pandemic as a controllable risk that can be reduced. Dealing with macrosocial stressors takes faith in politics and in those people, who work with COVID-19 on the frontline.

Mental disorders are found prevalent among college students and their onset occurs mostly before entry to college [ 31 ]. The diagnosis and timely interventions at an early stage of illness are essential to improve psychosocial functioning and treatment outcomes [ 31 ]. According to research that was conducted at the University of Debrecen in Hungary a few years ago, the students were found to have high levels of stress and the rate of the participants with impacted mental health was alarming [ 32 ]. With an unprecedented stressful event like the COVID-19 crisis, changes to the mental health status of people, including students, are expected.

Aims of the study

In our present study, we aimed at assessing the levels of health anxiety, perceived stress, and coping styles among university students amidst the COVID-19 lockdown in Hungary, using three validated assessment tools for each domain.

Methods and materials

Study design and setting.

This study utilized a cross-sectional design, using online self-administered questionnaires that were created and designed in Google Forms® (A web-based survey tool). Data collection was carried out in the period April 30, 2020, and May 15, 2020, which represents one of the most stressful periods during the early stage of the COVID-19 pandemic in Hungary when the official curfew/lockdown was declared along with the closure of dormitories and shifting to online remote teaching. The first cases of COVID-19 were declared in Hungary on March 4, 2020. On April 30, 2020, there were 2775 confirmed cases, 312 deaths, and 581 recoveries. As of May 15, 2020, the number of confirmed cases, deaths, and recovered persons was 3417, 442, and 1287, respectively.

Our study was conducted at the University of Debrecen, which is one of the largest higher education institutions in Hungary. The University is located in the city of Debrecen, the second-largest city in Hungary. Debrecen city is considered the educational and cultural hub of Eastern Hungary. As of October 2019, around 28,593 students were enrolled in various study programs at the University of Debrecen, of whom, 6,297 were international students [ 33 ]. The university offers various degree courses in Hungarian and English languages.

Study participants and sampling

The target population of our study was students at the University of Debrecen. Students were approached through social media platforms (e.g., Facebook®) and the official student administration system at the University of Debrecen (Neptun). The invitation link to our survey was sent to students on the web-based platforms described earlier. By using the Neptun system, we theoretically assumed that our survey questionnaire has reached all students at the University. The students who were interested and willing to participate in the study could fill out our questionnaire anonymously during the determined study period; thus, employing a convenience sampling approach. All students at the University of Debrecen whose age was 18 years or older and who were in Hungary during the outbreak had the eligibility to participate in our study whether undergraduates or postgraduates.

Study instruments

In our present study, the survey has solicited information about the sociodemographic profile of participants including age (in years), gender (female vs male), study program (health-related vs non-health related), and whether the student stayed in Hungary or traveled abroad during the period of conducting our survey in the outbreak. Our survey has also adopted three international scales to collect data about health anxiety, coping styles, and perceived stress during the pandemic crisis. As the language of instruction for international students at the University of Debrecen is English, and English fluency is one of the criteria for international students’ admission at the University of Debrecen, the international students were asked to fill out the English version of the survey and the scales. On the other hand, the Hungarian students were asked to fill out the Hungarian version of the survey and the validated Hungarian scales. Also, we provided contact information for psychological support when needed. Students who felt that they needed some help and psychological counseling could use the contact information of our peer supporters. Four International students have used this opportunity and were referred to a higher level of care. The original scales and their validated Hungarian versions are described in the following sections.

Perceived Stress Scale (PSS)

The Perceived Stress Scale (PSS) measures the level of stress in the general population who have at least completed a junior high school [ 34 ]. In the PSS, the respondents had to report how often certain things occurred like nervousness; loss of control; feeling of upset; piling up difficulties that cannot be handled; or on the contrary how often the students felt they were able to handle situations; and were on top of things. For the International students, we used the 10-item PSS (English version). The statements’ responses were scored on a 5-point Likert scale (from 0 = never to 4 = very often) as per the scale’s guide. Also, in the 10-item PSS, four positive items were reversely scored (e.g. felt confident about someone’s ability to handle personal problems) [ 34 ]. The PSS has satisfactory psychometric properties with a Cronbach’s alpha of 0.78, and this English version was used for international students in our study.

For the Hungarian students, we used the Hungarian version of the PSS, which has 14 statements that cover the same aspects of stress described earlier. In this version of the PSS, the responses were evaluated on a 5-point Likert scale (0–4) to mark how typical a particular behavior was for a respondent in the last month [ 35 ]. The Hungarian version of the PSS was psychometrically validated in 2006. In the validation study, the Hungarian 14-item PSS has shown satisfactory internal consistency with a Cronbach’s alpha of 0.88 [ 35 ].

Ways of Coping Questionnaire (WCQ)

The second scale we used was the 26-Item Ways of Coping Questionnaire (WCQ) which was developed by Sørlie and Sexton [ 36 ]. For the international students, we used the validated English version of the 26-Item WCQ that distinguished five different factors, including Wishful thinking (hoped for a miracle, day-dreamed for a better time), Goal-oriented (tried to analyze the problem, concentrated on what to do), Seeking support (talked to someone, got professional help), Thinking it over (drew on past experiences, realized other solutions), and Avoidance (refused to think about it, minimized seriousness of it). The WCQ examined how often the respondents used certain coping mechanisms, eg: hoped for a miracle, fantasized, prepared for the worst, analyzed the problem, talked to someone, or on the opposite did not talk to anyone, drew conclusions from past things, came up with several solutions for a problem or contained their feelings. As per the 26-item WCQ, responses were scored on a 4-point Likert scale (from 0 = “does not apply and/or not used” to 3 = “used a great deal”). This scale has satisfactory psychometric properties with Cronbach's alpha for the factors ranged from 0.74 to 0.81[ 36 ].

For the Hungarian students, we used the Hungarian 16-Item WCQ, which was validated in 2008 [ 37 ]. In the Hungarian WCQ, four dimensions were identified, which were cognitive restructuring/adaptation (every cloud has a silver lining), Stress reduction (by eating; drinking; smoking), Problem analysis (I tried to analyze the problem), and Helplessness/Passive coping (I prayed; used drugs) [ 37 ]. The Cronbach’s alpha values for the Hungarian WCQ’s dimensions were in the range of 0.30–0.74 [ 37 ].

Short Health Anxiety Inventory (SHAI)

The third scale adopted was the 18-Items Short Health Anxiety Inventory (SHAI). Overall, the SHAI has two subscales. The first subscale comprised of 14 items that examined to what degree the respondents were worried about their health in the past six months; how often they noticed physical pain/ache or sensations; how worried they were about a serious illness; how much they felt at risk for a serious illness; how much attention was drawn to bodily sensations; what their environment said, how much they deal with their health. The second subscale of SHAI comprised of 4 items (negative consequences if the illness occurs) that enquired how the respondents would feel if they were diagnosed with a serious illness, whether they would be able to enjoy things; would they trust modern medicine to heal them; how many aspects of their life it would affect; how much they could preserve their dignity despite the illness [ 38 ]. One of four possible statements (scored from 0 to 3) must be chosen. Alberts et al. (2013) [ 39 ] found the mean SHAI value to be 12.41 (± 6.81) in a non-clinical sample. The original 18-item SHAI has Cronbach’s alpha values in the range of 0.74–0.96 [ 39 ]. For the Hungarian students, the Hungarian version of the SHAI was used. The Hungarian version of SHAI was validated in 2011 [ 40 ]. The scoring differs from the English version in that the four statements were scored from 1 to 4, but the statements themselves were the same. In the Hungarian validation study, it was found that the SHAI mean score in a non-clinical sample (university students) was 33.02 points (± 6.28) and the Cronbach's alpha of the test was 0.83 [ 40 ].

Data analyses

Data were extracted from Google Forms® as an Excel sheet for quality check and coding then we used SPSS® (v.25) and RStudio statistical software packages to analyze the data. Descriptive and summary statistics were presented as appropriate. To assess the difference between groups/categories of anxiety, stress, and coping styles, we used the non-parametric Kruskal–Wallis test, since the variables did not have a normal distribution and for post hoc tests, we used the Mann–Whitney test. Also, we used Spearman’s rank correlation to assess the relationship between health anxiety and perceived stress within the international group and the Hungarian group. Comparison between international and domestic groups and different genders in terms of health anxiety and perceived stress levels were also conducted using the Mann–Whitney test. Binary logistic regression analysis was also employed to examine the associations between different coping styles/ strategies (treated as independent variables) and both, health anxiety level and perceived stress level (treated as outcome variables) using median splits. A p-value less than 5% was implemented for statistical significance.

Ethical considerations

Ethical permission was obtained from the Hungarian Ethical Review Committee for Research in Psychology (Reference number: 2020-45). All methods were carried out following the institutional guidelines and conforming to the ethical standards of the declaration of Helsinki. All participants were informed about the study and written informed consent was obtained before completing the survey. There were no rewards/incentives for completing the survey.

Sociodemographic characteristics of respondents

A total of 1320 students have responded to our survey. Six responses were eliminated due to incompleteness and an additional 25 responses were also excluded as the students filled out the survey from abroad (International students who were outside Hungary during the period of conducting our study). After exclusion of the described non-eligible responses (a total of 31 responses), the remaining 1289 valid responses were included in our analysis. Out of 1289 participants (100%), 73.5% were Hungarian students and around 26.5% were international students. Overall, female students have predominated the sample (n = 920, 71.4%). The median age (Interquartile range) among Hungarian students was 22 years (5) and for the international students was 22 years (4). Out of the total sample, most of the Hungarian students were enrolled in non-health-related programs (n = 690, 53.5%), while most of the international students were enrolled in health-related programs (n = 213, 16.5%). Table 1 demonstrates the sociodemographic profile of participants (Hungarian vs International).

Perceived stress, anxiety, and coping styles

For greater clarity of statistical analysis and interpretation, we created preferences regarding coping mechanisms. That is, we made the categories based on which coping factor (in the international sample) or dimension (in the Hungarian sample) the given person reached the highest scores, so it can be said that it is the person's preferred coping strategy. The four coping strategies among international students were goal-oriented, thinking it over, wishful thinking, and avoidance, while among the Hungarian students were cognitive restructuring, problem analysis, stress reduction, and passive coping.

The 26-item WCQ [ 31 ] contains a seeking support subscale which is missing from the Hungarian 16-item WCQ [ 32 ]; therefore, the seeking support subscale was excluded from our analysis. Moreover, because the PSS contained a different number of items in English and Hungarian versions (10 items vs 14 items), we looked at the average score of the answers so that we could compare international and domestic students.

In the evaluation of SHAI, the scoring of the two questionnaires are different. For the sake of comparability between the two samples, the international points were corrected to the Hungarian, adding plus one to the value of each answer. This may be the reason why we obtained higher results compared to international standards.

Among the international students, the mean score (± standard deviation) of perceived stress among male students was 2.11(± 0.86) compared to female students 2.51 (± 0.78), while the mean score (± standard deviation) of health anxiety was 34.12 (± 7.88) and 36.31 (± 7.75) among males and females, respectively. Table 2 shows more details regarding the perceived stress scores and health anxiety scores stratified by coping strategies among international students.

In the Hungarian sample, the mean score (± standard deviation) of perceived stress among male students was 2.06 (± 0.84) compared to female students 2.18 (± 0.83), while the mean score (± standard deviation) of health anxiety was 33.40 (± 7.63) and 35.05 (± 7.39) among males and females, respectively. Table 3 shows more details regarding the perceived stress scores and health anxiety scores stratified by coping strategies among Hungarian students.

Concerning coping styles among international students, the statements with the highest-ranked responses were “wished the situation would go away or somehow be finished” and “Had fantasies or wishes about how things might turn out” and both fall into the wishful thinking coping. Among the Hungarian students, the statements with the highest-ranked responses were “I tried to analyze the problem to understand better” (falls into problem analysis coping) and “I thought every cloud has a silver lining, I tried to perceive things cheerfully” (falls into cognitive restructuring coping).

On the other hand, the statements with the least-ranked responses among the international students belonged to the Avoidance coping. Among the Hungarians, it was Passive coping “I tried to take sedatives or medications” and Stress reduction “I staked everything upon a single cast, I started to do something risky” to have the lowest-ranked responses. Table 4 shows a comparison of different coping strategies among international and Hungarian students.

To test the difference between coping strategies, we used the non-parametric Kruskal–Wallis test, since the variables did not have a normal distribution. For post hoc tests, we used Mann–Whitney tests with lowered significance levels ( p  = 0.0083). Among Hungarian students, there were significant differences between the groups in stress ( χ 2 (3) = 212.01; p < 0.001) and health anxiety ( χ 2 (3) = 80.32; p  < 0.001). In the post hoc tests, there were significant differences everywhere ( p  < 0.001) except between stress reduction and passive coping ( p  = 0.089) and between problem analysis and passive coping ( p  = 0.034). Considering the health anxiety, the results were very similar. There were significant differences between all groups ( p  < 0.001), except between stress reduction and passive coping ( p  = 0.347) and between problem analysis and passive coping ( p  = 0.205). See Figs.  1 and 2 for the Hungarian students.

Perceived stress differences between coping strategies among the Hungarian students

Health anxiety differences between coping strategies among the Hungarian students

Among the international students, the results were similar. According to the Kruskal–Wallis test, there were significant differences in stress ( χ 2 (3) = 73.26; p  < 0.001) and health anxiety ( χ 2 (3) = 42.60; p  < 0.001) between various coping strategies. The post hoc tests showed that there were differences between the perceived stress level and coping strategies everywhere ( p  < 0.005) except and between avoidance and thinking it over ( p  = 0.640). Concerning health anxiety, there were significant differences between wishful thinking and goal-oriented ( p  < 0.001), between wishful thinking and avoidance ( p  = 0.001), and between goal-oriented and avoidance ( p  = 0.285). There were no significant differences between wishful thinking and thinking it over ( p  = 0.069), between goal-oriented and thinking it over ( p  = 0.069), and between avoidance and thinking it over ( p  = 0.131). See Figs.  3 and 4 .

Perceived stress differences between coping strategies among the international students

Health anxiety differences between coping strategies among the international students

The relationship between coping strategies with health anxiety and perceived stress levels among the international students

We applied logistic regression analyses for the variables to see which of the coping strategies has a significant effect on SHAI and PSS results. In the first model (model a), with the health anxiety as an outcome dummy variable (with median split; median: 35), only two coping strategies had a statistically significant relationship with health anxiety level, including wishful thinking (as a risk factor) and goal-oriented (as a protective factor).

In the second model (model b), with the perceived stress as an outcome dummy variable (with median split; median: 2.40), three coping strategies were found to have a statistically significant association with the level of perceived stress, including wishful thinking (as a risk factor), while goal-oriented and thinking it over as protective factors. See Table 5 .

The relationship between coping strategies with health anxiety and perceived stress levels among domestic students

By employing logistic regression analysis, with the health anxiety as an outcome dummy variable (with median split; median: 33.5) (model a), three coping strategies had a statistically significant relationship with health anxiety level among domestic students, including stress reduction and problem analysis (as risk factors), while cognitive restructuring (as a protective factor).

Similarly, with the perceived stress as an outcome dummy variable (with median split; median: 2.1429) (model b), three coping strategies had a statistically significant relationship with perceived stress level, including stress reduction and problem analysis (as risk factors), while cognitive restructuring (as a protective factor). See Table 6 .

Comparisons between domestic and international students

We compared health anxiety and perceived stress levels of the Hungarian and international students’ groups using the Mann–Whitney test. In the case of health anxiety, the results showed that there were significant differences between the two groups ( W  = 149,431; p  = 0.038) and international students’ levels were higher. Also, there was a significant difference in the perceived stress level between the two groups ( W  = 141,024; p  < 0.001), and the international students have increased stress levels compared to the Hungarian ones.

Comparisons between genders within students’ groups (International vs Hungarian)

Firstly, we compared the international men’s and women’s health anxiety and stress levels using the Mann–Whitney test. The results showed that the international women’s health anxiety ( W  = 11,810; p  = 0.012) and perceived stress ( W  = 10,371; p  < 0.001) levels were both significantly higher than international men’s values. However, in the Hungarian sample, women’s health anxiety was significantly higher than men’s ( W  = 69,643; p  < 0.001), but there was no significant difference in perceived stress levels among between Hungarian women and men ( W  = 75,644.5; p  = 0.064).

Relationship between health anxiety and perceived stress

We correlated the general health anxiety and perceived stress using Spearman’s rank correlation. There was a significant moderate positive relationship between the two variables ( p  < 0.001; ρ  = 0.446). Within the Hungarian students, there was a significant correlation between health anxiety and perceived stress ( p  < 0.001; ρ  = 0.433), similarly among international students as well ( p  < 0.001; ρ  = 0.465).

In our study, we found that individuals who were characterized by a preference for certain coping strategies reported significantly higher perceived stress and/or health anxiety than those who used other coping methods. These correlations can be found in both the Hungarian and international students. In the light of our results, we can say that 48.4% of the international students used wishful thinking as their preferred coping method while around 43% of the Hungarian students used primarily cognitive restructuring to overcome their problems.

Regulation of emotion refers to “the processes whereby individuals monitor, evaluate, and modify their emotions in an effort to control which emotions they have, when they have them, and how they experience and express those emotions” [ 41 ]. There is an overlap between emotion-focused coping and emotion regulation strategies, but there are also differences. The overlap between the two concepts can be noticed in the fact that emotion-focused coping strategies have an emotional regulatory role, and emotion regulation strategies may “tax the individual’s resources” as the emotion-focused coping strategies do [ 23 , 42 ]. However, in emotion-focused coping strategies, non-emotional tools can also be used to achieve non-emotional goals, while emotion regulation strategies may be used for maintaining or reinforcing positive emotions [ 42 ].

Based on the cognitive-behavioral model of health anxiety, emotion-regulating strategies can regulate the physiological, cognitive, and behavioral consequences of a fear response to some degree, even when the person encounters the conditioned stimulus again [ 12 , 43 ]. In the long run, regular use of these dysfunctional emotion control strategies may manifest as functional impairment, which may be associated with anxiety disorders. A detailed study that examined health anxiety in the view of the cognitive-behavioral model found that, regardless of the effect of depression, there are significant and consistent correlations between certain dimensions of health anxiety and dysfunctional coping and emotional regulation strategies [ 12 ].

Similar to our current study, other studies have found that health anxiety was positively correlated with maladaptive emotion regulation and negatively with adaptive emotion regulation [ 44 ], and in the case of state anxiety that emotion-focused coping strategies proved to be less effective in reducing stress, while active coping leads to a sense of subjective well-being [ 17 , 27 , 45 , 46 , 47 ]

SHAI values were found to be high in other studies during the pandemic, and the SHAI results of the international students in our study were found to be even slightly higher compared to those studies [ 44 , 48 ]. Besides, anxiety values for women were found to be higher than for men in several studies [ 44 , 48 , 49 , 50 ]. This was similar to what we found among the international students but not among the Hungarian ones. We can speculate that the ability to contact someone, the closeness of family and beloved ones, familiarity with the living environment, and maybe less online search about the coronavirus news could be factors counting towards that finding among Hungarian students. Also, most international students were enrolled in health-related study programs and his might have affected how they perceived stress/anxiety and their preferred coping strategies as well. Literature found that students of medical disciplines could have obstacles in achieving a healthy coping strategy to deal with stress and anxiety despite their profound medical knowledge compared to non-health-related students [ 51 , 52 ]. Literature also stressed the immense need for training programs to help students of medical disciplines in adopting coping skills and stress-reducing strategies [ 51 ].

The findings of our study may be a starting point for the exploration of the linkage between perceived stress, health anxiety, and coping strategies when people are not in their domestic context. People who are away from their home and friends in a relatively alien environment may tend to use coping mechanisms other than the adequate ones, which in turn can lead to increased levels of perceived stress.

Furthermore, our results seem to support the knowledge that deep-rooted health anxiety is difficult to change because it is closely related to certain coping mechanisms. It was also addressed in the literature that personality traits may have a significant influence on the coping strategy used by a person [ 53 ], revealing sophisticated and challenging links to be considered especially during training programs on effective coping and management skills. On the other hand, perceived stress which has risen significantly above the average level in the current pandemic, can be most effectively targeted by the well-formulated recommendations and advice of major international health organizations if people successfully adhere to them (e.g. physical activity; proper and adequate sleep; healthy eating; avoiding alcohol; meditation; caring for others; relationships maintenance, and using credible information resources about the pandemic, etc.) [ 1 , 54 ]. Furthermore, there may be additional positive effects of these recommendations when published in different languages or languages that are spoken by a wide range of nationalities. Besides, cognitive behavioral therapy techniques, some of which are available online during the current pandemic crisis, can further reduce anxiety. Also, if someone does not feel safe or fear prevails, there are helplines to get in touch with professionals, and this applies to the University of Debrecen in Hungary, and to a certain extent internationally.

Naturally, our study had certain limitations that should be acknowledged and considered. The temporality of events could not be assessed as we employed a cross-sectional study design, that is, we did not have information on the previous conditions of the participants which means that it is possible that some of these conditions existed in the past, while others de facto occurred with COVID-19 crisis. The survey questionnaires were completed by those who felt interested and involved, i.e., a convenience sampling technique was used, this impairs the representativeness of the sample (in terms of sociodemographic variables) and the generalizability of our results. Also, the type of recruitment (including social media) as well as the online nature of the study, probably appealed more to people with an affinity with this kind of instrument. Besides, each questionnaire represented self-reported states; thus, over-reporting or under-reporting could be present. It is also important to note that international students were answering the survey questionnaire in a language that might not have been their mother language. Nevertheless, English fluency is a prerequisite to enroll in a study program at the University of Debrecen for international students. As the options for gender were only male/female in our survey questionnaire, we might have missed the views of students who do not identify themselves according to these gender categories. Also, no data on medical history/current medical status were collected. Lastly, we had to make minor changes to the used scales in the different languages for comparability.

The COVID-19 pandemic crisis has imposed a significant burden on the physical and psychological wellbeing of humans. Crises like the current pandemic can trigger unprecedented emotional and behavioral responses among individuals to adapt or cope with the situation. The elevated perceived stress levels during major life events can be further deepened by disengagement from home and by using inadequate coping strategies. By following and adhering to the international recommendations, adopting proper coping strategies, and equipping oneself with the required coping and stress management skills, the associated high levels of perceived stress and anxiety might be mitigated.

Availability of data and materials

The datasets generated and/or analyzed during the current study are not publicly available due to compliance with institutional guidelines but they are available from the corresponding author (LRK) on a reasonable request.

Abbreviations

Centers for Disease Control and Prevention

Coronavirus Disease 2019

Perceived Stress Scale

Short Health Anxiety Inventory

Middle East Respiratory Syndrome

Severe Acute Respiratory Syndrome

Ways of Coping Questionnaire

World Health Organization

World Health Organization. Advice for the public on COVID-19. [Online]. 2020. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public . Accessed 9 Sep 2020.

National Center for Public Health. Opportunities to reduce contact numbers—Community events in relation to COVID-19 virus infection. [Online]. 2020. https://www.nnk.gov.hu/index.php/koronavirus-tajekoztato/549-opportunities-to-reduce-contact-numbers-community-events-in-relation-to-covid-19-virus-infection . Accessed 20 Sep 2020.

GardaWorld. Crisis24 News Alert. [Online]. 2020. https://www.garda.com/crisis24/news-alerts?search_api_fulltext=&na_countries%5B%5D=1431&field_news_alert_categories=All&field_news_alert_crit=All&items_per_page=20 . Accessed 20 Sep 2020.

Tanne JH, Hayasaki E, Zastrow M, Pulla P, Smith P, Rada AG. Covid-19: how doctors and healthcare systems are tackling coronavirus worldwide. Br Med J. 2020;368:m1090.

Article   Google Scholar  

Sohrabi C, Alsafi Z, O’Neill N, Khan M, Kerwan A, Al-Jabir A, et al. World Health Organization declares global emergency: A review of the 2019 novel coronavirus (COVID-19). Int J Surg. 2020;76:71–6.

Rosenbaum L. Facing Covid-19 in Italy—ethics, logistics, and therapeutics on the epidemic’s front line. N Engl J Med. 2020;382:1873–5. https://doi.org/10.1056/NEJMp2005492 .

Article   PubMed   Google Scholar  

Wang C, Pan R, Wan X, Tan Y, Xu L, Ho CS, et al. Immediate Psychological Responses and Associated Factors during the Initial Stage of the 2019 Coronavirus Disease (COVID-19) epidemic among the general population in China. Int J Environ Res Public Health. 2020;17:1729.

Akour A, Al-Tammemi AB, Barakat M, Kanj R, Fakhouri HN, Malkawi A, et al. The impact of the COVID-19 pandemic and emergency distance teaching on the psychological status of university teachers: a cross-sectional study in Jordan. Am J Trop Med Hyg. 2020;103:2391–9.

Al-Tammemi AB, Akour A, Alfalah L. Is it just about physical health? An online cross-sectional study exploring the psychological distress among university students in Jordan in the Midst of COVID-19 Pandemic. Front Psychol. 2020;11:562213.

Roddenberry A, Renk K. Locus of control and self-efficacy: potential mediators of stress, illness, and utilization of health services in college students. Child Psychiatry Hum Dev. 2010;41:353–70.

Racic M, Todorovic R, Ivkovic N, Masic S, Joksimovic B, Kulic M. Self- perceived stress in relation to anxiety, depression and health-related quality of life among health professions students: a cross-sectional study from Bosnia and Herzegovina. Slov J Public Heal. 2017;56:251–9.

Görgen SM, Hiller W, Witthöft M. Health anxiety, cognitive coping, and emotion regulation: a latent variable approach. Int J Behav Med. 2014;21:364–74.

Abramowitz JS, Braddock A. Psychological treatment of health anxiety and hypochondriasis: a biopsychosocial approach. Boston: Hogrefe Publishing; 2008.

Google Scholar  

Taylor S, Asmundson GJG. Treating health anxiety: a cognitive-behavioral approach. 1st ed. New York: Guilford Press; 2004.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders,(DSM-5). 5th edition. Washington, DC: American Psychiatric Association; 2013.

Wheaton MG, Abramowitz JS, Berman NC, Fabricant LE, Olatunji BO. Psychological predictors of anxiety in response to the H1N1 (swine flu) pandemic. Cognit Ther Res. 2012;36:210–8.

Asmundson GJG, Taylor S. How health anxiety influences responses to viral outbreaks like COVID-19: What all decision-makers, health authorities, and health care professionals need to know. J Anxiety Disord. 2020;71:102211.

Taha SA, Matheson K, Anisman H. The 2009 H1N1 influenza pandemic: the role of threat, coping, and media trust on vaccination intentions in Canada. J Health Commun. 2013;18:278–90.

Phua DH, Tang HK, Tham KY. Coping responses of emergency physicians and nurses to the 2003 severe acute respiratory syndrome outbreak. Acad Emerg. 2005;12:322–8.

Teasdale E, Yardley L, Schlotz W, Michie S. The importance of coping appraisal in behavioural responses to pandemic flu. Br J Health Psychol. 2012;17:44–59.

Sim K, Huak Chan Y, Chong PN, Chua HC, Wen SS. Psychosocial and coping responses within the community health care setting towards a national outbreak of an infectious disease. J Psychosom Res. 2010;68:195–202.

Lazarus RS. Stress and emotion: a new synthesis. London: Free Association Books; 1999.

Lazarus RS, Folkman S. Stress, appraisal, and coping. New York: Springer; 1984.

Lazarus RS, Folkman S. Transactional theory and research on emotions and coping. Eur J Pers. 1987;1:141–69.

Cai H, Tu B, Ma J, Chen L, Fu L, Jiang Y, et al. psychological impact and coping strategies of frontline medical staff in hunan between January and March 2020 during the Outbreak of Coronavirus Disease 2019 (COVID-19) in Hubei. China Med Sci Monit. 2020;26:e924171.

PubMed   Google Scholar  

Chua SE, Cheung V, Cheung C, McAlonan GM, Wong JWS, Cheung EPT, et al. Psychological effects of the SARS outbreak in Hong Kong on high-risk health care workers. Can J Psychiatry. 2004;49:391–3. https://doi.org/10.1177/070674370404900609 .

Flesia L, Monaro M, Mazza C, Fietta V, Colicino E, Segatto B, et al. Predicting perceived stress related to the Covid-19 outbreak through stable psychological traits and machine learning models. J Clin Med. 2020;9:3350.

Khalid I, Khalid TJ, Qabajah MR, Barnard AG, Qushmaq IA. Healthcare workers emotions, perceived stressors and coping strategies during a MERS-CoV outbreak. Clin Med Res. 2016;14:7–14.

Gee S, Skovdal M. The role of risk perception in willingness to respond to the 2014–2016 West African Ebola outbreak: a qualitative study of international health care workers. Glob Heal Res Policy. 2017;2:21.

Gerhold L. COVID-19: Risk perception and Coping strategies. Results from a survey in Germany. PsyArXiv Prepr. 2020.

Auerbach RP, Alonso J, Axinn WG, Cuijpers P, Ebert DD, Green JG, et al. Mental disorders among college students in the World Health Organization World Mental Health Surveys. Psychol Med. 2016;46:2955–70.

Bíró É. Studies on the mental health of students in higher education. University of Debrecen; 2014. https://dea.lib.unideb.hu/dea/handle/2437/195979 . Accessed 15 Feb 2021.

The University of Debrecen. Facts and Figures. [Online]. 2020. https://www.edu.unideb.hu/page.php?id=28 . Accessed 25 Dec 2020.

Cohen S. Perceived stress in a probability sample of the United States. In: The social psychology of health. Thousand Oaks: Sage Publications, Inc; 1988. p. 31–67.

Strauder A, Thege BK. Az Észlelt Stressz Kérdőív (PSS) Magyar Verziójának Jellemzői. Mentálhigiéné És Pszichoszomatika. 2006;7:203–16.

Sørlie T, Sexton HC. The factor structure of “The Ways of Coping Questionnaire” and the process of coping in surgical patients. Pers Individ Dif. 2001;30:961–75.

Rózsa S, Purebl G, Susánszky É, Kő N, Szádóczky E, Réthelyi J, et al. Dimensions of coping: Hungarian adaptation of the Ways of Coping Questionnaire. Mentálhigiéné És Pszichoszomatika. 2008; 217–241.

Salkovskis PM, Rimes KA, Warwick HMC. The Health Anxiety Inventory: development and validation of scales for the measurement of health anxiety and hypochondriasis. Psychol Med. 2002;32:843.

Alberts NM, Hadjistavropoulos HD, Jones SL, Sharpe D. The Short Health Anxiety Inventory: a systematic review and meta-analysis. J Anxiety Disord. 2013;27:68–78.

Köteles F, Simor P, Bárdos G. Validation and psychometric evaluation of the Hungarian version of the Short Health Anxiety Inventory (SHAI). Mentálhigiéné és Pszichoszomatika. 2011;12:191–213.

Artino AR. Regulation of Emotion. In: Encyclopedia of Child Behavior and Development. Boston, MA: Springer US; 2011. p. 1236–8.

Gross JJ. The emerging field of emotion regulation: an integrative review. Rev Gen Psychol. 1998;2:271–99.

Kaczkurkin AN, Foa EB. Cognitive-behavioral therapy for anxiety disorders: an update on the empirical evidence. Dialogues Clin Neurosci. 2015;17:337–46.

Jungmann SM, Witthöft M. Health anxiety, cyberchondria, and coping in the current COVID-19 pandemic: Which factors are related to coronavirus anxiety? J Anxiety Disord. 2020;73:102239.

Taha S, Matheson K, Cronin T, Anisman H. Intolerance of uncertainty, appraisals, coping, and anxiety: the case of the 2009 H1N1 pandemic. Br J Health Psychol. 2014;19:592–605.

Umucu E, Lee B. Examining the impact of COVID-19 on stress and coping strategies in individuals with disabilities and chronic conditions. Rehabil Psychol. 2020;65:193–8.

Main A, Zhou Q, Ma Y, Luecken LJ, Liu X. Relations of SARS-related stressors and coping to Chinese college students’ psychological adjustment during the 2003 Beijing SARS epidemic. J Couns Psychol. 2011;58:410–23.

Özdin S, Bayrak ÖŞ. Levels and predictors of anxiety, depression and health anxiety during COVID-19 pandemic in Turkish society: the importance of gender. Int J Soc Psychiatry. 2020;66:504–11.

Taylor S, Landry CA, Paluszek MM, Fergus TA, McKay D, Asmundson GJG. COVID stress syndrome: concept, structure, and correlates. Depress Anxiety. 2020;37:706–14.

Gamonal Limcaoco RS, Mateos EM, Fernández JM, Roncero C. Anxiety, worry and perceived stress in the world due to the COVID-19 pandemic, March 2020. Preliminary results. MedRxiv Prepr. 2020.

Abouammoh N, Irfan F, AlFaris E. Stress coping strategies among medical students and trainees in Saudi Arabia: a qualitative study. BMC Med Educ. 2020;20:124.

Gade S, Chari S, Gupta M. Perceived stress among medical students: to identify its sources and coping strategies. Arch Med Health Sci. 2014;2:80–6.

Leszko M, Iwański R, Jarzębińska A. The relationship between personality traits and coping styles among first-time and recurrent prisoners in Poland. Front Psychol. 2020;10:2969.

Centers for Disease Control and Prevention. Mental Health and Coping During COVID-19 Pandemic. [Online]. 2020. https://www.cdc.gov/coronavirus/2019-ncov/daily-life-coping/managing-stress-anxiety.html . Accessed 9 Sep 2020.

Download references

Acknowledgments

We would like to provide our extreme thanks and appreciation to all students who participated in our study. ABA is currently supported by the Tempus Public Foundation’s scholarship at the University of Debrecen.

This research project did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author information

Authors and affiliations.

Doctoral School of Health Sciences, University of Debrecen, Debrecen, Hungary

Szabolcs Garbóczy, Szilvia Harsányi, Ala’a B. Al-Tammemi & László Róbert Kolozsvári

Department of Psychiatry, Faculty of Medicine, University of Debrecen, Debrecen, Hungary

Szabolcs Garbóczy

Department of Personality and Clinical Psychology, Institute of Psychology, University of Debrecen, Debrecen, Hungary

Anita Szemán-Nagy

Faculty of Medicine, University of Debrecen, Debrecen, Hungary

Mohamed S. Ahmad & Viktor Rekenyi

Department of Social and Work Psychology, Institute of Psychology, University of Debrecen, Debrecen, Hungary

Dorottya Ocsenás

Doctoral School of Human Sciences, University of Debrecen, Debrecen, Hungary

Department of Family and Occupational Medicine, Faculty of Medicine, University of Debrecen, Móricz Zs. krt. 22, Debrecen, 4032, Hungary

Ala’a B. Al-Tammemi & László Róbert Kolozsvári

You can also search for this author in PubMed   Google Scholar

Contributions

All authors SG, ASN, MSA, SH, DO, VR, ABA, and LRK have worked on the study design, text writing, revising, and editing of the manuscript. DO, SG, and VR have done data management and extraction, data analysis. Drafting and interpretation of the manuscript were made in close collaboration by all authors SG, ASN, MSA, SH, DO, VR, ABA, and LRK. All authors read and approved the final manuscript.

Corresponding author

Correspondence to László Róbert Kolozsvári .

Ethics declarations

Ethics approval and consent to participate.

Ethical permission was obtained from the Hungarian Ethical Review Committee for Research in Psychology (Reference number: 2020-45). All methods were carried out following the institutional guidelines and conforming to the ethical standards of the declaration of Helsinki. All participants were informed about the study and written informed consent was obtained before completing the survey.

Consent for publication

Not Applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Garbóczy, S., Szemán-Nagy, A., Ahmad, M.S. et al. Health anxiety, perceived stress, and coping styles in the shadow of the COVID-19. BMC Psychol 9 , 53 (2021). https://doi.org/10.1186/s40359-021-00560-3

Download citation

Received : 07 January 2021

Accepted : 26 March 2021

Published : 06 April 2021

DOI : https://doi.org/10.1186/s40359-021-00560-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Health anxiety
• Perceived stress
• Coping styles
• University students

BMC Psychology

ISSN: 2050-7283

• General enquiries: [email protected]

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Individual stress response patterns: Preliminary findings and possible implications

Roles Conceptualization, Project administration, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

¶ ‡ These authors are joint senior authors on this work.

Affiliation Stress, Hope and Cope Lab., School of Behavioral Sciences, Tel-Aviv Yaffo Academic College, Tel-Aviv Yaffo, Israel

Roles Investigation, Writing – original draft

Roles Conceptualization, Investigation, Resources, Writing – original draft

Roles Conceptualization, Resources

Roles Conceptualization, Formal analysis, Methodology, Project administration, Writing – review & editing

• Rebecca Jacoby, 
• Keren Greenfeld Barsky, 
• Tal Porat, 
• Stav Harel, 
• Tsipi Hanalis Miller, 
• Gil Goldzweig

• Published: August 13, 2021
• https://doi.org/10.1371/journal.pone.0255889
• Reader Comments

Research on stress occupied a central position during the 20 th century. As it became evident that stress responses affect a wide range of negative outcomes, various stress management techniques were developed in attempt to reduce the damages. However, the existing interventions are applied for a range of different stress responses, sometimes unsuccessfully.

The aim of this study was to examine whether there are specific clusters of stress responses representing interpersonal variation. In other words, do people have dominant clusters reflecting the different aspects of the known stress responses (physiological, emotional, behavioral, and cognitive)?

The researchers derived a measure of stress responses based on previous scales and used it in two studies in order to examine the hypothesis that stress responses can be grouped into dominant patterns according to the type of response.

The results of Study 1 revealed four distinctive response categories: psychological (emotional and cognitive), physiological gastro, physiological muscular, and behavioral. The results of Study 2 revealed five distinctive response categories: emotional, cognitive, physiological gastro, physiological muscular, and behavioral.

By taking into consideration each person’s stress response profile while planning stress management interventions and then offering them a tailored intervention that reduces the intensity of these responses, it might be possible to prevent further complications resulting in a disease (physical or mental).

Citation: Jacoby R, Greenfeld Barsky K, Porat T, Harel S, Hanalis Miller T, Goldzweig G (2021) Individual stress response patterns: Preliminary findings and possible implications. PLoS ONE 16(8): e0255889. https://doi.org/10.1371/journal.pone.0255889

Editor: Georgia Panayiotou, University of Cyprus, CYPRUS

Received: February 13, 2021; Accepted: July 26, 2021; Published: August 13, 2021

Copyright: © 2021 Jacoby et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data is registered and available through the Open Science Framework (OSF): Goldzweig, G., Jacoby, R., Barsky, K. G., Porat, T., Harel, S., & Miller, T. H. (2021, July 10). Individual stress response patterns: Preliminary findings and possible implications. https://doi.org/10.17605/OSF.IO/XQ2TA .

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The concept of stress in its various forms (such as anxiety and fear) has been known since the 18 th century but became central during the 20 th century. Stress models were, consequently, developed in order to explain the processes undertaken in response to stressors. Early stress research underlined the adaptive nature of organisms’ physiological response to acute stress and the potentially deleterious effects of prolonged stressors [ 1 , 2 ]. Later on, the research expanded to humans emphasizing the cognitive system as mediating between stressors and stress responses [ 3 , 4 ]. As it became evident that continuing or repeated stress responses affect a wide range of negative outcomes, both physical and psychological [ 5 , 6 ], various stress management techniques were developed in an attempt to reduce the harms [ 7 – 10 ]. These include psychophysiological techniques (such as relaxation, biofeedback and more) aiming to reduce stress responses and regaining control, as well as cognitive behavioral techniques aiming to challenge misleading perceptions. A substantial body of research has explored the efficacy of these techniques (e.g. [ 10 ]). However, in most cases, interventions were not tailored to specific patterns of responses, and thus, the same interventions were used, sometimes unsuccessfully, for various stress responses. Since stress responses influence organisms’ coping and health outcomes, their study is important. We believe that by better understanding the complexity of stress responses, appropriate interventions can be developed and implemented.

Theories on stress responses

In the past, stress responses were studied and described mainly by physiologists,. Darwin (1809–1882), who studied the manifested responses of animals and humans to emotional situations, was mainly interested in the observable responses and not in the biochemical changes that occur in response to stress [ 11 ]. Immense strides toward understanding stress responses and their physiological basis were made by Cannon (1871–1945) and Selye (1907–1982). Cannon [ 2 ] studied the homeostatic mechanisms underlying the “fight or flight” response to stressful situations, while Selye [ 1 ] later developed the general adaptation syndrome (GAS) theory, which describes the process of responding to an ongoing stress.

According to these theories, exposure to a stressful event activates a series of autonomic system reactions that cause changes within organs [ 12 , 13 ]. The reactions found were the activation of the hypothalamic-pituitary-adrenal axis (HPA) and the sympathetic nervous system (SNS). Activation of the HPA axis and the SNS causes hormonal secretions of adrenaline and cortisol and the behavioral “fight or flight” reaction [ 2 ]. According to Berger et al. [ 14 ], the “fight or flight” response is triggered by osteocalcin, a protein released by the skeleton as a hormone, which, they claimed, is a messenger, sent by bone to regulate crucial processes all over the body, including how we respond to danger.

Porges [ 15 ] asserted that although these arousal theories have empirical support when measuring the effects of acute stress, they neglect other aspects of the physiological stress response such as parasympathetic nervous system (PNS) influences and interaction between sympathetic and parasympathetic processes. This neglect limits the theories’ ability to explain a wide variety of stress responses such as freezing, tonic immobility, fainting, and syncope. Porges’ polyvagal theory looks to explain the mechanism underlying the interpersonal differences of physiological and psychological stress responses. Other theories on the role of oxytocin for moderating the autonomic nervous system (e.g., [ 16 ]) and on gender differences, such as “tend and befriend” [ 17 ], have focused on responses directed toward safety behaviors.

Although these researchers have tried to include a psychological dimension in their models, this was mainly cast in terms of stimulus-response relationships, consistent with the dominant physiological and behavioral approaches of the period, and therefore could not explain why different people who are exposed to the same stimulus respond differently. These models, which were derived from animal behavior, were criticized for their universal approach of focusing solely on biological mechanisms and disregarding humans’ subjective perception of the stress experience [ 18 ].

From the middle of the 20 th century, the concept of stress came to occupy a central position in the psychological literature, and new stress models were developed emphasizing the interactions between individuals and their environment. The leading model in psychological stress research is the Transactional Model developed by Lazarus and Folkman [ 3 ]. This model focuses on the cognitive processes preceding the stress response and promotes the understanding of interpersonal variance in the stress responses to the same events. It emphasizes the importance of the individuals’ appraisal of the meaning of the stressful event and their own resources for coping with this event to help mediate between the stressor and stress responses of the organism. It also established the understanding that different individuals will react to the same event with varying intensity or duration: one will find an event threatening, while the other will find it neutral or realize that they have the required coping resources. The Transactional Model does not, however, explain the interpersonal variance of the stress response patterns and of stress influences on health.

Other theories have proposed a more integrative outlook on the stress-related cascade of events, starting even before the encounter with a potential stressor and resulting in various health outcomes. They have suggested a process that is mediated by cognitive appraisal, behavioral outcomes, and physiological mechanisms [ 19 , 20 ]. For example, Brosschot, Gerin, & Thayer [ 21 ] argue that perseverative cognition as manifested in worry, rumination and anticipatory stress should be considered as they are associated with enhanced cardiovascular, endocrinological, immunological, and neurovisceral activity. Others [ 22 , 23 ] have suggested that personality traits are also likely to influence how people respond to stress. These approaches consider all of the main aspects depicted by prior models and provide a wider perspective for both researchers and clinicians.

The aforementioned theories notwithstanding, individual differences of stress responses as represented by different clusters in a non-pathological population have not, to the best of our knowledge, been studied. The purpose of the current study was, therefore, to address this gap and examine whether reported stress responses do, in fact, reflect clusters of the common stress responses: physiological, emotional, behavioral, and cognitive. We also strived to assess interpersonal variation in stress responses; in other words, do people have dominant clusters of stress responses?

Measuring stress responses

Different scales were developed to measure stress responses. For example, Terluin [ 24 ] developed the Four-Dimensional Symptom Questionnaire (4DSQ) in order to differentiate between general distress and what he considered as psychiatric symptoms, namely, depression, anxiety, and somatization. Schlebusch [ 25 ] developed the Stress Symptom Checklist (SSCL) which consists of three categories: physical, psychological, and behavioral. The checklist was intended to be a diagnostic tool that measures specific stress-related psychopathological conditions or disorders, particularly the intensity (or severity) of stress as reflected by an individual’s physical, psychological, and behavioral reactions.

These scales were primarily intended to measure the total intensity of the stress response in order to identify either pathological or intense stress responses, assuming the existence of a unified stress response for all. They ignored the different patterns people exhibit when confronted by a stressor, thus limiting their ability to characterize an individual’s dominant stress response pattern. Based on these works and others, stress responses were generally classified into four categories: physiological, emotional, behavioral, and cognitive [ 26 , 27 ].

Two studies were conducted in order to examine the hypothesis that stress responses can be grouped into dominant patterns according to the type of response (physiological, emotional, behavioral, and cognitive). Although the existing scales include various items representing the above mentioned categories they are too long and didn’t meet our research purposes. Therefore we have decided to derive a short scale of stress responses, representing the four categories, based on the above mentioned scales (see details under " items selection" in the Study 1 description). Participants in the first study were students while participants in the second study were a sample of people suffering from the stress-related medical syndromes of fibromyalgia (FM), irritable bowel syndrome (IBS), or both. Participants in both studies were asked to rate the extent to which different stress responses characterize their typical responses to stress. The results are presented separately for each study.

The same statistical analysis was used for both studies. Descriptive statistics was calculated for each item (stress response). We conducted an exploratory factor analysis (principle component analyses with Varimax rotation) of all items. For the second study we also calculated sub-scale scores (base and factor analysis) and compared these scores between the study groups.

The two studies were approved by the ethics committee of the Academic College of Tel Aviv-Yaffo, and all participants signed an electronic consent form prior to the study’s initiation.

Step 1: Items selection

In order to create a comprehensive list of stress response measures, the authors screened the two validated stress response questionnaires: the Four-Dimensional Symptom Questionnaire [ 24 ] and the Stress Symptom Checklist [ 25 ]. The responses were pre-classified separately by each of the authors into the four categories: physiological, emotional, behavioral, and cognitive. Discrepancies between the authors were discussed and resolved when at least four of the six authors agreed on the classification. Other items were newly added in order to encompass stress responses in all four categories. Four external experts examined and discussed the content validity of the new items as expressing stress responses and matching the relevant stress response categories. The final list included 66 items.

Step 2: Identifying the partition of stress responses

Participants..

The participants in Study 1 were first-year psychology undergraduate students at the Academic College of Tel Aviv-Yaffo. They participated in the study as part of their undergraduate program requirements and were recruited via the college’s credit database. A total of 100 participants enrolled in the study, with 91 fully completing the questionnaire. All 91 were first-year undergraduate students (84.6% female, 15.4% male) and the mean age was 23.56 years (SD = 1.37, range = 21–29).

The 66 selected items scale.

A short sociodemographic questionnaire including data on: age and gender.

The items were presented to the participants through the Qualtrics XM online platform. Participants were asked to recall a stressful event and to rate each response item on a scale ranging from “not at all” (1) to “always” (5) reflecting the extent to which each item characterize their response to stressful situations. All participants signed an electronic consent form before answering the questionnaires. The data was gathered and stored anonymously.

Data analysis.

In the first stage of analysis we conducted an exploratory factor analysis (principle component analyses with Varimax rotation) of all 66 items, which revealed four factors (eigenvalue>1.0), accounting for 48% of overall variance. We then screened the items and excluded 36 items based on both content analysis and factor loadings (items with loading< 0.5 were omitted) following two-step analyses. We ended with a final set of 30 items which we found as satisfying for our research purposes (hereinafter the 30 items scale).

In the second stage of analysis, we determined the number of factors according to the Kaiser criterion of eigenvalue> = 1 [ 28 ] and identified 7 factors accounting for 70.38% of the total variance. However, according to the scree plot, 3–4 factors could have been retained. We chose a conservative approach and determined 4 factors. The 4 factors accounted for 58.48% of the total variance.

Table 1 presents the factor loadings and descriptive statistics for each item. It is evident that items loaded on Factor 1 include mostly psychological (emotional and cognitive) responses (introversion, loneliness, confusion, etc.). Factor 2 items include mostly physiological-gastro responses (digestive upset, stomach pains, etc.). Factor 3 items include mostly physiological-muscular responses (neck and shoulder pain, backaches, etc.). Factor 4 items include mostly unregulated behavioral responses (temper flare-ups, nervousness, etc.). The item “physical unrest” was loaded on both Factor 1 and Factor 2.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0255889.t001

We calculated a mean score for each factor. Factor 1, psychological, had the highest score (mean = 3.16; SD = 0.93; reliability Cronbach’s Alpha = 0.94, McDonalds Omega = 0.94), followed by Factor 4, behavioral, (mean = 3.13; SD = 0.89;reliability Cronbach’s Alpha = 0.84, McDonalds Omega = 0.84), and Factor 2, physiological-gastro, (mean = 2.56; SD = 1.00; reliability Cronbach’s Alpha = 0.65, McDonalds Omega = 0.67), and finally, Factor 3, physiological-muscular, (mean = 2.27; SD = 0.90; reliability: Cronbach’s Alpha = 0.77, McDonalds Omega = 0.79). Differences between all pairs of factors were significant except for Factor 3 vs. Factor 4.

In order to get further insight into the structure of the stress response items we conducted a smallest space analysis (SSA). SSA is a method of non-metric multidimensional scaling (NMDS) in which a set of variables and their inter-correlations are geometrically portrayed in a multidimensional space [ 29 ]. SSA treats each variable (i.e., each questionnaire item) as a point in a Euclidean space—the higher the correlation between two variables, the closer the points in the space. It attempts to find the space with the minimum number of dimensions in which the rank order of relations is preserved. The regional partition of the SSA space can be studied in conjunction with the corresponding content of the mapped variables. All points within a region should be associated with a specific set of variables of the same content [ 30 – 33 ].

As can be seen, the SSA space in Fig 1 is partitioned into four polar (or angular) regions. Each polar region corresponds to one of the four categories—psychological, physiological-gastro, physiological-muscular, and behavioral—with their respective items. Polar regions divide the space into pie-shaped sections, all emanating from a common point. The elements of a polar facet are considered to be unordered but related [ 37 ]; they differ in kind but not necessarily in complexity. It should be noted that each two adjacent categories are close to each other in some respect.

https://doi.org/10.1371/journal.pone.0255889.g001

Participants

Participants were people over 18 years old diagnosed with fibromyalgia (FM), irritable bowel syndrome (IBS) or both. Participants were recruited from four different online forums based in Israel—two of which were dedicated to IBS the rest to FM. All participants volunteered for the study and none were offered any compensation. A total of 217 participants enrolled in the study but only 143 completed the questionnaire. Amongst these, 62 participants reported having FM (43.35%), 45 reported having IBS (31.47%) and 36 reported having both IBS and FM (25.17%). 95.1% of all participants reported being officially diagnosed by a doctor. The reported mean time from diagnosis was 8.43 years (SD = 6.82). 129 were female (90.2%) and 14 were male (9.8%). This gender difference might be partially explained by the fact that both IBS and FM are more common in women worldwide. Mean age was M = 37.67 years SD = 13.2.

The 30 items scale (see Study 1 ).

A short sociodemographic questionnaire including data on: age, gender, diagnosis, time since diagnosis and who gave the diagnosis.

The items were presented to the participants through the Qualtrics XM online platform. Participants were asked to recall a stressful event and to rate each response item on a scale ranging from “not at all” (1) to “always” (5) reflecting the extent to which each item characterize their response to stressful situations. All participants signed an electronic consent form before answering the questionnaires. All data was gathered and stored anonymously.

Data analysis

We determined the number of factors according to the Kaiser criterion of eigenvalue> = 1 [ 36 ] and identified 7 factors accounting for 65.77% of the total variance. We identified 4–5 factors according to the scree plot. The fit to comparison data method (CD) revealed that the 4 factors solutions added significantly to the eigenvalue of 3 factors solution. Nevertheless, we decided on a conservative approach and we set the number of factors at 5. The 5 factors accounted for 58.01% of the total variance.

Table 2 presents the factor loadings and descriptive statistics for each item. Factor 1 included emotional responses identical to those included in factor 1 (psychological) in study 1. Three items that were included in this factor in study 1 (confusion, difficulty concentrating and attention dispersion) were now included in the additional factor 5 that consists of cognitive responses. Factor 2 included physiological- muscular items (identical to factor 3 in study 1) and the items insomnia and fatigue that were included in Factor 1 in study 1. Factor 3 included behavioural items and was identical to factor 4 in study 1. Factor 4 included physiological-gastro items, identical to the items included in Factor 2 in study 1 (except for the physical unrest item that in study 2 was included in Factor 1).

https://doi.org/10.1371/journal.pone.0255889.t002

We calculated a mean score for each factor (see Table 2 ) there were no significant differences between the factors (largest difference was 0.0962). Factors reliability measures: Factor 1—Cronbach’s Alpha = 0.88, McDonalds Omega = 0.88; Factor 2—Cronbach’s Alpha = 0.82, McDonalds Omega = 0.83; Factor 3—Cronbach’s Alpha = 0.85, McDonalds Omega = 0.86; Factor 4—Cronbach’s Alpha = 0.74, McDonalds Omega = 0.75; Factor 5—Cronbach’s Alpha = 0.85, McDonalds Omega = 0.85.

In order to get further insight into the structure of the stress response items we conducted a smallest space analysis (SSA) similar to the SSA conducted in study 1.

As can be seen, the SSA space in Fig 2 is partitioned into five polar (or angular) regions. Each polar region corresponds to one of the five categories—emotional, physiological-gastro, physiological-muscular, behavioral and cognitive—with their respective items. Polar regions divide the space into pie-shaped sections, all emanating from a common point. The elements of a polar facet are considered to be unordered but related; they differ in kind but not necessarily in complexity. It should be noted that each two adjacent categories are close to each other in some respect.

https://doi.org/10.1371/journal.pone.0255889.g002

Table 3 which presents comparisons of the 5 factors between the study groups of study 2 indicate that the IBS group reported on significantly lower levels of physiological-muscular distress in comparison to the other two groups. The IBS group reported significantly higher levels of physiological-gastro distress in comparison to the FM group. The IBS group was also found to be significantly lower in comparison to the other two groups on the cognitive factor.

https://doi.org/10.1371/journal.pone.0255889.t003

Both the emotional and behavioral factors did not differ significantly between the groups.

In this study, we aimed to examine whether there are specific clusters of stress responses, representing interpersonal variation. More specifically, we hypothesized that different stress response clusters will reflect the different aspects of the known physiological, emotional, behavioral, and cognitive stress responses, thus allowing the identification of individual patterns.

The results obtained from Study 1 revealed four distinctive response categories: psychological (emotional and cognitive), physiological-gastro, physiological-muscular and behavioral. The psychological category entails mostly clinical symptoms of depression and anxiety as well as cognitive responses. Two of the categories are physiological responses: one mostly gastro-related symptoms and the other mostly muscle tension symptoms. The fourth category entails unregulated behaviors. The results obtained from Study 2 revealed five distinctive response categories: emotional, cognitive, physiological-gastro, physiological-muscular, and behavioral. As can be seen, the psychological category is divided into emotional and cognitive. These results thus portray an interesting classification that, if understood, may help to shed new light on stress response patterns and to highlight potential psychological and physiological susceptibilities.

It is already well established that psychological stress plays a role in negative physical and mental health conditions [ 20 , 34 – 36 ]. Thus, each of these response patterns may reflect a specific time point or dimension in the cascade of events vulnerability that may have a pathological outcome. For example, it was found that the link between stress and depression and anxiety is underlined by biological mechanisms such as HPA axis activation and inflammatory processes [ 37 ]. It can therefore be postulated that a person characterized by a high score in the emotional category may, in fact, be at risk of not only depression and anxiety disorder but also high cortisol-related illnesses. Such an assumption may be even more pronounced when an individual presents a high score in one or both of the physiological response categories. For instance, irritable bowel syndrome (IBS) was found to be adversely affected by psychological stress via several possible biological pathways including gastrointestinal function [ 38 ]. The digestion-related symptoms characterizing the physiological-gastro category may therefore, indicate susceptibility to such illnesses.

As a result of these findings, we decided to proceed with a study testing whether stress responses in a sample of people suffering from the stress-related medical syndromes of fibromyalgia (FM), irritable bowel syndrome (IBS) or both (FM+IBS) will reflect our assumptions.

Our results (see Table 2 ) indicate that both the IBS and the FM+IBS groups reported experiencing physiological-gastro stress responses during stressful events more often than the FM group. We also found that both the FM and the FM+IBS groups reported experiencing physiological-muscular stress responses during stressful events more often than the IBS group.

These results are compatible with previous research findings regarding pain sensitivity patterns in these groups. Two-thirds of IBS patients have been found to have lower visceral pain thresholds [ 39 ], while their musculoskeletal pain thresholds are normal [ 40 ] or higher than in normal controls [ 41 ]. In contrast, FM patients have decreased musculoskeletal pain thresholds but normal visceral pain thresholds [ 42 , 43 ]. It appears that only the subgroup of patients who have both IBS and FM suffer both from visceral and somatic hypersensitivity [ 42 ]. However, we find that the direction of the correlation needs further study. For example, it is possible that people who have physiological-gastro responses to stress are more likely to develop IBS later, but it is also possible that people who already have IBS are more likely to respond to stress with gastrointestinal symptoms. We suggest that future longitudinal studies inspect the nature of this correlation.

We also found that both the FM and the FM+IBS groups reported experiencing emotional stress during stressful events more often than the IBS group. An earlier finding by Janssens, Zijlema, Joustra, and Rosmalen [ 44 ] that major depressive disorder is more common in FM than in IBS may explain our findings.

Conclusions

Our individual responses to stressful events embody much about who we are and what we have gone through. Our genetics, past experiences, gender, beliefs, and even smoking habits play a key role in how we react to stressors [ 19 , 45 – 47 ]. In mapping these reactions and patterns, we can obtain a clearer image of each person’s stress responses profile. A possible clinical implication of the findings of this study is the understanding that if we take into consideration the individual’s stress responses profile while planning stress management interventions and offer them a tailored intervention that reduces the intensity of these responses, we might prevent further complications resulting in physical or mental disease. Therefore, stress management interventions should be considered seriously and evidence based. An improved validated scale of stress responses may serve in the future as an important tool that will allow for the implementation of such tailored psychological interventions in various settings with minimal resources.

Limitations

Despite these important implications, there are some methodological limitations in the current study. First, the scale we have used for our research purposes is composed of items selected from previous scales and has not been validated. Second, the participants in study 2 (people who suffer from FM or IBS or both) differ from those of study 1 (students). Third, the majority of the participants were female, which might have led to a bias due to gender differences.

In addition, there are some theoretical concerns. Since our results are based on retrospective reports, participants may not accurately remember how they usually act and feel and may appraise how they have always responded to stress according to the salience of events rather than actual frequency. Previous research has suggested that people who suffer from chronic pain have an attention bias that makes pain more salient to them than it would be in normal controls (e.g., [ 48 ]). We therefore propose that future studies use a daily log of stressful events and subsequent stress reactions in order to circumvent possible memory biases.

Another major challenge is differentiating between stress responses and stress coping strategies. Some theorists have even preferred to limit the concept of coping to voluntary responses [ 49 ], while others have included automatic and involuntary responses as well [ 50 , 51 ]. However, it is difficult to distinguish between voluntary and involuntary responses—if “volition” even exists at all. Libet [ 52 ] posited that if volition does indeed exist, it is only expressed when we use a conscious effort to think or behave differently than we are used to. Furthermore, thoughts and behaviors that are intentional and effortful when first used may, he claimed, become automatic and involuntary with repetition.

These limitations notwithstanding, our study calls for a detailed observation of the components of the existing stress models, focusing specifically on the function of stress responses and their impact on health while implementing tailored interventions that take into consideration individuals’ specific response clusters.

• 1. Selye H. The Stress of Life. Hans Selye. MD New York, McGraw-Hill Book Company, Inc. 1956:479.
• 2. Cannon WB. The wisdom of the body. 1939.
• 3. Lazarus RS, Folkman S. Stress, appraisal, and coping: Springer publishing company; 1984.
• View Article
• PubMed/NCBI
• Google Scholar
• 5. Sapolsky RM. Why zebras don’t get ulcers: The acclaimed guide to stress, stress-related diseases, and coping: Holt paperbacks; 2004.
• 7. Jaremko M, Meichenbaum D. Stress reduction and prevention: Springer Science & Business Media; 2013.
• 11. Darwin C. The Expression of the Emotions in Man and Animals. Chicago (University of Chicago Press) 1965. 1965.
• 26. Taylor SE. Health psychology: Tata McGraw-Hill Education; 2006.
• 30. Brown J. An introduction to the uses of facet theory. Facet theory: Springer; 1985. p. 17–57.
• 32. Canter D. An introduction to the uses of facet theory. Facet theory approaches to social research: Springer; 1985. p. 18–57.
• 33. Borg IS, Samuel . Facet Theory: Form and Content: 5 (Advanced Quantitative Techniques in the Social Sciences),. CA: Sage: Newbury Park; 1995.
• 35. Hjemdahl P, Rosengren A, Steptoe A. Stress and cardiovascular disease: Springer Science & Business Media; 2011.
• 50. Eisenberg N, Fabes RA, Guthrie IK. Coping with stress. Handbook of children’s coping: Springer; 1997. p. 41–70.

• Open access
• Published: 24 February 2024

Physical activity improves stress load, recovery, and academic performance-related parameters among university students: a longitudinal study on daily level

• Monika Teuber 1 ,
• Daniel Leyhr 1 , 2 &
• Gorden Sudeck 1 , 3  

BMC Public Health volume  24 , Article number:  598 ( 2024 ) Cite this article

3857 Accesses

30 Altmetric

Metrics details

Physical activity has been proven to be beneficial for physical and psychological health as well as for academic achievement. However, especially university students are insufficiently physically active because of difficulties in time management regarding study, work, and social demands. As they are at a crucial life stage, it is of interest how physical activity affects university students' stress load and recovery as well as their academic performance.

Student´s behavior during home studying in times of COVID-19 was examined longitudinally on a daily basis during a ten-day study period ( N  = 57, aged M  = 23.5 years, SD  = 2.8, studying between the 1st to 13th semester ( M  = 5.8, SD  = 4.1)). Two-level regression models were conducted to predict daily variations in stress load, recovery and perceived academic performance depending on leisure-time physical activity and short physical activity breaks during studying periods. Parameters of the individual home studying behavior were also taken into account as covariates.

While physical activity breaks only positively affect stress load (functional stress b = 0.032, p  < 0.01) and perceived academic performance (b = 0.121, p  < 0.001), leisure-time physical activity affects parameters of stress load (functional stress: b = 0.003, p  < 0.001, dysfunctional stress: b = -0.002, p  < 0.01), recovery experience (b = -0.003, p  < 0.001) and perceived academic performance (b = 0.012, p  < 0.001). Home study behavior regarding the number of breaks and longest stretch of time also shows associations with recovery experience and perceived academic performance.

Conclusions

Study results confirm the importance of different physical activities for university students` stress load, recovery experience and perceived academic performance in home studying periods. Universities should promote physical activity to keep their students healthy and capable of performing well in academic study: On the one hand, they can offer opportunities to be physically active in leisure time. On the other hand, they can support physical activity breaks during the learning process and in the immediate location of study.

Peer Review reports

Introduction

Physical activity (PA) takes a particularly key position in health promotion and prevention. It reduces risks for several diseases, overweight, and all-cause mortality [ 1 ] and is beneficial for physical, psychological and social health [ 2 , 3 , 4 , 5 ] as well as for academic achievement [ 6 , 7 ]. However, PA levels decrease from childhood through adolescence and into adulthood [ 8 , 9 , 10 ]. Especially university students are insufficiently physically active according to health-oriented PA guidelines [ 11 ] because of academic workloads as well as difficulties in time management regarding study, work, and social demands [ 12 ]. Due to their independence and increasing self-responsibility, university students are at a crucial life stage. In this essential and still educational stage of the students´ development, it is important to study their PA behavior. Furthermore, PA as health behavior represents one influencing factor which is considered in the analytical framework of the impact of health and health behaviors on educational outcomes which was developed by the authors Suhrcke and de Paz Nieves [ 13 , 14 ]. In light of this, the present study examines how PA affects university students' academic situations.

Along with the promotion of PA, the reduction of sedentary behavior has also become a crucial part of modern health promotion and prevention strategies. Spending too much time sitting increases many health risks, including the risk of obesity [ 15 ], diabetes [ 16 ] and other chronic diseases [ 15 ], damage to muscular balances, bone metabolism and musculoskeletal system [ 17 ] and even early death [ 15 ]. University students are a population that has shown the greatest increase in sedentary behavior over the last two decades [ 18 ]. In Germany, they show the highest percentage of sitting time among all working professional groups [ 19 ]. Long times sitting in classes, self-study learning, and through smartphone use, all of which are connected to the university setting and its associated behaviors, might be the cause of this [ 20 , 21 ]. This goes along with technological advances which allow students to study in the comfort of their own homes without changing locations [ 22 ].

To counter a sedentary lifestyle, PA is crucial. In addition to its physical health advantages, PA is essential for coping with the intellectual and stress-related demands of academic life. PA shows positive associations with stress load and academic performance. It is positively associated with learning and educational success [ 6 ] and even shows stress-regulatory potential [ 23 ]. In contrast, sedentary behavior is associated with lower cognitive performance [ 24 ]. Moreover, theoretical derivations show that too much sitting could have a negative impact on brain health and diminish the positive effects of PA [ 16 ]. Given the theoretical background of the stressor detachment model [ 25 ] and the cybernetic approach to stress management in the workplace [ 26 ], PA can promote recovery experience, it can enhance academic performance, and it is a way to reduce the impact of study-related stressors on strain. Load-related stress response can be bilateral: On the one hand, it can be functional if it is beneficial to help cope with the study demands. On the other hand, it can be dysfunctional if it puts a strain on personal resources and can lead to load-related states of strain [ 27 ]. Thus, both, the promotion of PA and reduction of sedentary behavior are important for stress load, recovery, and performance in student life, which can be of particular importance for students in an academic context.

A simple but (presumably) effective way to integrate PA and reduce sedentary behavior in student life are short PA breaks. Due to the exercises' simplicity and short duration, students can perform them wherever they are — together in a lecture or alone at home. Short PA breaks could prevent an accumulation of negative stressors during the day and can help with prolonged sitting as well as inactivity. Especially in the university setting, evidence of the positive effects of PA breaks exists for self-perceived physical and psychological well-being of the university students [ 28 ]. PA breaks buffer university students’ perceived stress [ 29 ] and show positive impacts on recovery need [ 30 ] and better mood ratings [ 31 , 32 ]. In addition, there is evidence for reduction in tension [ 30 ], overall muscular discomfort [ 33 ], daytime sleepiness or fatigue [ 33 , 34 ] and increase in vigor [ 34 ] and experienced energy [ 30 ]. This is in line with cognitive, affective, behavioral, and biological effects of PA, all categorized as palliative-regenerative coping strategies, which addresses the consequences of stress-generating appraisal processes aiming to alleviate these consequences (palliative) or restore the baseline of the relevant reaction parameter (regenerative) [ 35 , 36 ]. This is achieved by, for example, reducing stress-induced cortisol release or tension through physical activity (reaction reduction) [ 35 ]. Such mechanisms are also in accordance with the previously mentioned stressor detachment model [ 25 ]. Lastly, there is a health-strengthening effect that impacts the entire stress-coping-health process, relying on the compensatory effects of PA which is in accordance to the stress-buffering effect of exercise [ 37 ]. Health, in turn, effects educational outcomes [ 13 , 14 ]. Therefore, stress regulating effects are also accompanied with the before mentioned analytical framework of the impact of health and health behaviors on educational outcomes [ 13 , 14 ].

Focusing on the effects of PA, this study is guided by an inquiry into how PA affects university students' stress load and recovery as well as their perceived academic performance. For that reason, the student´s behavior during home studying in times of COVID-19 is examined, a time in which reinforced prolonged sitting, inactivity, and a negative stress load response was at a high [ 38 , 39 , 40 , 41 , 42 ]. Looking separately on the relation of PA with different parameters based on the mentioned evidence, we assume that PA has a positive impact on stress load, recovery, and perceived academic performance-related parameters. Furthermore, a side effect of the home study behavior on the mentioned parameters is assumed regarding the accumulation of negative stressors during home studying. These associations are presented in Fig.  1 and summarized in the following hypotheses:

Overview of the assumed effects and investigated hypotheses of physical activity (PA) behavior on variables of stress load and recovery and perceived academic performance-related parameters

Hypothesis 1 (path 1): Given that stress load always occurs as a duality—beneficial if it is functional for coping, or exhausting if it puts a strain on personal resources [ 27 ] – we consider two variables for stress load: functional stress and dysfunctional stress. In order to reduce the length of the daily surveys, we focused the measure of recovery only on the most obvious and accessible component of recovery experience, namely psychological detachment. PA (whether performed in leisure-time or during PA breaks) encourages functional stress and reduce dysfunctional stress (1.A) and has a positive effect on recovery experience through psychological detachment (1.B).

Hypothesis 2 (path 2): The academic performance-related parameters attention difficulties and study ability are positively influenced by PA (whether done in leisure-time or during PA breaks). We have chosen to assess attention difficulties for a cognitive parameter because poor control over the stream of occurring stimuli have been associated with impairment in executive functions or academic failure [ 43 , 44 , 45 , 46 ]. Furthermore, we have assessed the study ability to refer to the self-perceived feeling of functionality regarding the demands of students. PA reduces self-reported attention difficulties (2.A) and improves perceived study ability, indicating that a student feels capable of performing well in academic study (2.B).

Hypothesis 3: We assume that a longer time spent on studying at home (so called home studying) could result in higher accumulation of stressors throughout the day which could elicit immediate stress responses, while breaks in general could reduce the influence of work-related stressors on strain and well-being [ 47 , 48 ]. Therefore, the following covariates are considered for secondary effects:

the daily longest stretch of time without a break spent on home studying

the daily number of breaks during home studying

Study setting

The study was carried out during the COVID-19 pandemic containment phase. It took place in the middle of the lecture period between 25th of November and 4th of December 2020. Student life was characterized by home studying and digital learning. A so called “digital semester” was in effect at the University of Tübingen when the study took place. Hence, courses were mainly taught online (e.g., live or via a recorded lecture). Other events and actions at the university were not permitted. As such, the university sports department closed in-person sports activities. For leisure time in general, there were contact restrictions (social distancing), the performance of sports activities in groups was not permitted, and sports facilities were closed.

Thus, the university sports department of the University of Tübingen launched various online sports courses and the student health management introduced an opportunity for a new digital form of PA breaks. This opportunity provided PA breaks via videos with guided physical exercises and health-promoting explanations for a PA break for everyday home studying: the so called “Bewegungssnack digital” [in English “exercise snack digital” (ESD)] [ 49 ]. The ESD videos took 5–7 min and were categorized into three thematic foci: activation, relaxation, and coordination. Exercises were demonstrated by one or two student exercise leaders, accompanied by textual descriptions of the relevant execution features of each exercise.

Participants

Participants were recruited within the framework of an intervention study, which was conducted to investigate whether a digital nudging intervention has a beneficial effect on taking PA breaks during home study periods [ 49 ]. Students at the University of Tübingen which counts 27,532 enrolled students were approached for participation through a variety of digital means: via an email sent to those who registered for ESD course on the homepage of the university sports department and to all students via the university email distribution list; via advertisement on social media of the university sports department (Facebook, Instagram, YouTube, homepage). Five tablets, two smart watches, and one iPad were raffled off to participants who engaged actively during the full study period in an effort to motivate them to stick with it to the end. In any case, participants knew that the study was voluntary and that they would not suffer any personal disadvantages should they opt out. There was a written informed consent prompt together with a prompt for the approval of the data protection regulations immediately within the first questionnaire (T0) presented in a mandatory selection field. Positive ethical approval for the study was given by the first author´s institution´s ethics committee of the faculty of the University of Tübingen.

Participants ( N  = 57) who completed the daily surveys on at least half of the days of the study period, were included in the sample (male = 6, female = 47, diverse = 1, not stated = 3). As not all subjects provided data on all ten study days, the total number of observations was between 468 and 540, depending on the variable under study (see Table  1 ). The average number of observations per subject was around eight. Their age was between 18 and 32 years ( M  = 23.52, SD  = 2.81) and they were studying between the 1st to 13th semester ( M  = 5.76, SD  = 4.11) within the following major courses of study: mathematical-scientific majors (34.0%), social science majors (22.6%), philosophical majors (18.9%), medicine (13.2%), theology (5.7%), economics (3.8%), or law (1.9%). 20.4% of the students had on-site classroom teaching on university campus for at least one day a week despite the mandated digital semester, as there were exceptions for special forms of teaching.

Design and procedures

To examine these hypothesized associations, a longitudinal study design with daily surveys was chosen following the suggestion of the day-level study of Feuerhahn et al. (2014) and also of Sonnentag (2001) measuring recovery potential of (exercise) activities during leisure time [ 50 , 51 ]. Considering that there are also differences between people at the beginning of the study period, initial base-line value variables respective to the outcomes measured before the study period were considered as independent covariates. Therefore, the well-being at baseline serves as a control for stress load (2.A), the psychological detachment at baseline serves as a control for daily psychological detachment (2.B), the perception of study demands serves as a control for self-reported attention difficulties (1.A), and the perceived study ability at baseline serves as a control for daily study ability (2.B).

Subjects were asked to continue with their normal home study routine and additionally perform ESD at any time in their daily routine. Data were collected one to two days before (T0) as well as daily during the ten-day study period (Wednesday to Friday). The daily surveys (t 1 -t 10 ) were sent by email at 7 p.m. every evening. Each day, subjects were asked to answer questions about their home studying behavior, study related requirements, recovery experience from study tasks, attention, and PA, including ESD participation. The surveys were conducted online using the UNIPARK software and were recorded and analyzed anonymously.

Measures and covariates

In total, five outcome variables, two independent variables, and seven covariates were included in different analyses: three variables were used for stress load and recovery parameters, two variables for academic performance-related parameters, two variables for PA behavior, two variables for study behavior, four variables for outcome specific baseline values and one variable for age.

Outcome variables

Stress load & recovery parameters (hypothesis 1).

Stress load was included in the analysis with two variables: functional stress and dysfunctional stress. Followingly, a questionnaire containing a word list of adjectives for the recording of emotions and stress during work (called “Erfassung von Emotionen und Beanspruchung “ in German, also known as EEB [ 52 ]) was used. It is an instrument which were developed and validated in the context of occupational health promotion. The items are based on mental-workload research and the assessment of the stress potential of work organization [ 52 ]. Within the questionnaire, four mental and motivational stress items were combined to form a functional stress scale (energetic, willing to perform, attentive, focused) (α = 0.89) and four negative emotional and physical stress items were combined to form dysfunctional stress scale (nervous, physically tensioned, excited, physically unwell) (α = 0.71). Participants rated the items according to how they felt about home studying in general on the following scale (adjustment from “work” to “home studying”): hardly, somewhat, to some extent, fairly, strongly, very strongly, exceptionally.

Recovery experience was measured via psychological detachment. Therefore, the dimension “detachment” of the Recovery Experience Questionnaire (RECQ [ 53 ]) was adjusted to home studying. The introductory question was "How did you experience your free time (including short breaks between learning) during home studying today?". Students responded to four statements based on the extent to which they agreed or disagreed (not at all true, somewhat true, moderately true, mostly true, completely true). The statements covered subjects such as forgetting about studying, not thinking about studying, detachment from studying, and keeping a distance from student tasks. The four items were combined into a score for psychological detachment (α = 0.94).

Academic performance-related parameters (hypothesis 2)

Attention was assessed via the subscale “difficulty maintaining focused attention performance” of the “Attention and Performance Self-Assessment” (ASPA, AP-F2 [ 54 ]). It contains nine items with statements about disturbing situations regarding concentration (e.g. “Even a small noise from the environment could disturb me while reading.”). Participants had to answer how often such situations happened to them on a given day on the following scale: never, rarely, sometimes, often, always. The nine items were combined into the AP-F2 score (α = 0.87).

The perceived study ability was assessed using the study ability index (SAI [ 55 ]). The study ability index captures the current state of perceived functioning in studying. It is based on the Work Ability Index by Hasselhorn and Freude ([ 56 ]) and consists of an adjusted short scale of three adapted items in the context of studying. Firstly, (a) the perceived academic performance was asked after in comparison to the best study-related academic performance ever achieved (from 0 = completely unable to function to 10 = currently best functioning). Secondly, the other two items were aimed at assessing current study-related performance in relation to (b) study tasks that have to be mastered cognitively and (c) the psychological demands of studying. Both items were answered on a five-point Likert scale (1 = very poor, 2 = rather poor, 3 = moderate, 4 = rather good, 5 = very good). A sum index, the SAI, was formed which can indicate values between 2 and 20, with higher values corresponding to higher assessed functioning in studies (α = 0.86). In a previous study it already showed satisfying reliability (α = 0.72) [ 55 ].

Independent variables

Pa behavior.

Two indicators for PA behavior were included via self-reports: the time spent on ESD and the time spent on leisure-time PA (LTPA). Participants were asked the following overarching question daily: “How much time did you spend on physical activity today and in what context”. For the independent variable time spent on PA breaks, participants could answer the option “I participated in the Bewegungssnack digital” with the amount of time they spent on it (in minutes). To assess the time spent on LTPA besides PA breaks, participants could report their time for four different contexts of PA which comprised two forms: Firstly, structured supervised exercise was reported via time spent on (a) university sports courses and (b) other organized sports activities. Secondly, self-organized PA was indicated via (c) independent PA at home, such as a workout or other physically demanding activity such as cleaning or tidying up, as well as via (d) independent PA outside, like walking, cycling, jogging, a workout or something similar. Referring to the different domains of health enhancing PA [ 57 ], the reported minutes of these four types of PA were summed up to a total LTPA value. The total LTPA value was included in the analysis as a metric variable in minutes.

Covariates (hypothesis 3)

Regarding hypothesis 3 and home study behavior, the longest daily stretch of time without a break spent on home studying (in hours) and the daily number of breaks during home studying was assessed. Therein, participants had to answer the overarching question “How much time did you spend on your home studying today?” and give responses to the items: (1) longest stretch of time for home studying (without a break), and (2) number of short and long breaks you took during home studying.

In principle, efforts were made to control for potential confounders at the individual level (level 2) either by including the baseline measure (T0) of the respective variable or by including variables assessing related trait-like characteristics for respective outcomes. The reason why related trait-like characteristics were used for the outcomes was because brief assessments were used for daily surveys that were not concurrently employed in the baseline assessment. To enable the continued use of controlling for person-specific baseline characteristics in the analysis of daily associations, trait-like characteristics available from the baseline assessment were utilized as the best possible approximation.To sum up, four outcome specific baseline value variables were measured before the study period (at T0). The psychological detachment with the RECQ (α = 0.87) [ 53 ] was assessed at the beginning to monitor daily psychological detachment. Further, the SAI [ 55 ] was assessed at the beginning of the study period to monitor daily study ability. To monitor daily stress load, which in part measures mental stress aspects and negative emotional stress aspects, the well-being was assessed at the beginning using the WHO-Five Well-being Index (WHO-5 [ 58 ]). It is a one-dimensional self-report measure with five items. The index value is the sum of all items, with higher values indicating better well-being. As the well-being and stress load tolerance may linked with each other, this variable was assumed to be a good fit with the daily stress load indicating mental and emotional stress aspects. With respect to student life, daily academic performance-related attention was monitored with an instrument for the perception of study demands and resources (termed “Berliner Anforderungen Ressourcen-Inventar – Studierende” in German, the so-called BARI-S [ 59 ]). It contains eight items which capture overwork in studies, time pressure during studies, and the incompatibility of studies and private life. All together they form the BARI-S demand scale (α = 0.85) which was included in the analysis. As overwork and time pressure may result in attention difficulties (e.g. Elfering et al., 2013), this variable was assumed to have a good fit with academic performance-related attention [ 60 ]. Additionally, age in years at T0 was considered as a sociodemographic factor.

Statistical analysis

Since the study design provided ten measurement points for various people, the hierarchical structure of the nested data called for two-level analyses. Pre-analyses of Random-Intercept-Only models for each of the outcome variables (hypothesis 1 to 3) revealed an Intra-Class-Correlation ( ICC ) of at least 0.10 (range 0.26 – 0.64) and confirmed the necessity to perform multilevel analyses [ 61 ]. Specifically, the day-level variables belong to Level 1 (ESD time, LTPA time, longest stretch of time without a break spent on home studying, daily number of breaks during home studying). To analyze day-specific effects within the person, these variables were centered on the person mean (cw = centered within) [ 50 , 62 , 63 , 64 ]. This means that the analyses’ findings are based on a person’s deviations from their average values. The variables assessed at T0 belong to Level 2, which describe the person level (psychological detachment baseline, SAI baseline, well-being, study demands scale, age). These covariates on person level were centered around the grand mean [ 50 ] indicating that the analyses’ findings are based how far an individual deviates from the sample's mean values. As a result, the models’ intercept reflects the outcome value of an average student in the sample at his/her daily average behavior in PA and home study when all parameters are zero. For descriptive statistics SPSS 28.0.1.1 (IBM) and for inferential statistics R (version 4.1.2) were used. The hierarchical models were calculated using the package lme4 with the lmer-function in R in the following steps [ 65 ]. The Null Model was analyzed for all models first, with the corresponding intercept as the only predictor. Afterwards, all variables were entered. The regression coefficient estimates (”b”) were considered for statistical significance for the models and the respective BIC was provided.

In total, five regression models with ‘PA break time’ and ‘LTPA time’ as independent variables were computed due to the five measured outcomes of the present study. Three models belonged to hypothesis 1 and two models to hypothesis 2.

Hypothesis 1: To test hypothesis 1.A two outcome variables were chosen for two separate models: ‘functional stress’ and ‘dysfunctional stress’. Besides the PA behavior variables, the ‘number of breaks’, the ‘longest stretch of time without a break spent on home studying’, ‘age’, and the ‘well-being’ at the beginning of the study as corresponding baseline variable to the output variable were also included as independent variables in both models. The outcome variable ‘psychological detachment’ was utilized in conjunction with the aforementioned independent variables to test hypotheses 1.B, with one exception: psychological detachment at the start of the study was chosen as the corresponding baseline variable.

Hypothesis 2: To investigate hypothesis 2.A the outcome variable ‘attention difficulties’ was selected. Hypothesis 2.B was tested with the outcome variables ‘study ability’. Both models included both PA behavior variables as well as the ‘number of breaks’, the ‘longest stretch of time without a break spent on home studying’, ‘age’ and one corresponding baseline variable each: the ‘study demand scale’ at the start of the study for ‘attention difficulties’ and the ‘SAI’ at the beginning of the study for the daily ‘study ability’.

Hypothesis 3: In addition to both PA behavior variables, age and one baseline variable that matched the outcome variable, the covariates ‘daily longest stretch of time spent on home studying’ and ‘daily number of breaks during home studying’ were included in the models for all five outcome variables.

Handling missing data

The dataset had up to 18% missing values (most exhibit the variables ‘daily longest stretch of time without a break spent on home studying’ with 17.89% followed by ‘daily number of breaks during homes studying’ with 16.67%, and ‘functional / dysfunctional stress’ with 12.45%). Therefore, a sensitivity analysis was performed using the multiple imputation mice-package in the statistical program R [ 66 ], the package howManyImputation based on Von Hippel (2020, [ 67 ]), and the additional broom package [ 68 ]. The results of the models remained the same, with one exception for the Attention Difficulties Model: The daily longest stretch of time without a break spent on home studying showed a significant association (Table  1 in supplement). Due to this almost perfect consistency of results between analyses based on the dataset with missing data and those with imputed data alongside the lack of information provided by the packages for imputed datasets, we decided to stick with the main analysis including the missing data. Thus, in the following the results of the main analysis without imputations are presented.

Table 1 shows the descriptive statistics of the variables used in the analysis. An overview of the analysed models is presented in Table  2 .

Effects on stress load and recovery (hypothesis 1)

Hypothesis 1.A: The Model Functional Stress explained 13% of the variance by fixed factors (marginal R 2  = 0.13), and 52% by both fixed and random factors (conditional R 2  = 0.52). The time spent on ESD as well as the time spent on PA in leisure showed a positive significant influence on functional stress (b = 0.032, p  < 0.01). The same applied to LTPA (b = 0.003, p  < 0.001). The Model Dysfunctional Stress (marginal R 2  = 0.027, conditional R 2  = 0.647) showed only one significant result. The dysfunctional stress was only significantly negatively influenced by the time spent on LTPA (b = 0.002, p  < 0.01).

Hypothesis 1.B: With the Model Detachment, fixed factors contributed 18% of the explained variance and fixed and random factors 46% of the explained variance for psychological detachment. Only the amount of time spent on LTPA revealed a positive impact on psychological detachment (b = 0.003, p  < 0.001).

Effects on academic performance-related parameters (hypothesis 2)

Hypothesis 2.A: The Model Attention Difficulties showed 13% of the variance explained by fixed factors, and 51% explained by both fixed and random factors. It showed a significant negative association only for the time spent on LTPA (b = 0.003, p  < 0.001).

Hypothesis 2.B: The Model SAI showed 18% of the variance explained by fixed factors, and 39% explained by both fixed and random factors. There were significant positive associations for time spent on ESD (b = 0.121, p  < 0.001) and time spent on LTPA (b = 0.012, p  < 0.001). The same applied to LTPA (b = 0.012, p  < 0.001).

Effects of home study behavior (hypothesis 3)

Regarding the independent covariates for the outcome variables functional and dysfunctional stress, there were no significant results for the number of breaks during homes studying or the longest stretch of time without a break spent on home studying. Considering the outcome variable ‘psychological detachment’, there were significant results with negative impact for both study behavior variables: breaks during home studying (b = 0.058, p  < 0.01) and daily longest stretch of time without a break (b = 0.120, p  < 0.01). Evaluating the outcome variables ‘attention difficulties’, there were no significant results for the number of breaks during home studying or the longest stretch of time without a break spent on home studying. Testing the independent study behavior variables for the SAI, it increased with increasing number in daily breaks during homes studying relative to the person´s mean (b = 0.183, p  < 0.05). No significant effect was found for the longest stretch of time without a break spent on home studying ( p  = 0.07).

The baseline covariates of the models showed expected associations and thus confirmed their inclusion. The baseline variables well-being showed a significant impact on functional stress (b = 0.089, p  < 0.001), psychological detachment showed a positive effect on the daily output variables psychological detachment (b = 0.471, p  < 0.001), study demand scale showed a positive association on difficulties in attention (b = 0.240, p  < 0.01), and baseline SAI had a positive effect on the daily SAI (b = 0.335, p  < 0.001).

The present study theorized that PA breaks and LTPA positively influence the academic situation of university students. Therefore, impact on stress load (‘functional stress’ and ‘dysfunctional stress’) and ‘psychological detachment’ as well as academic performance-related parameters ‘self-reported attention difficulties’ and ‘perceived study ability’ was taken into account. The first and second hypotheses assumed that both PA breaks and LTPA are positively associated with the aforementioned parameters and were confirmed for LTPA for all parameters and for PA breaks for functional stress and perceived study ability. The third hypothesis assumed that home study behavior regarding the daily number of breaks during home studying and longest stretch of time without a break spent on home studying has side effects. Detected negative effects for both covariates on psychological detachment and positive effects for the daily number of breaks on perceived study ability were partly unexpected in their direction. These results emphasize the key position of PA in the context of modern health promotion especially for students in an academic context.

Regarding hypothesis 1 and the detected positive associations for stress load and recovery parameters with PA, the results are in accordance with the stress-regulatory potential of PA from the state of research [ 23 ]. For hypothesis 1.A, there is a positive influence of PA breaks and LTPA on functional stress and a negative influence of LTPA on dysfunctional stress. Given the bilateral role of stress load, the results indicate that PA breaks and LTPA are beneficial for coping with study demands, and may help to promote feelings of joy, pride, and learning progress [ 27 ]. This is in line with previous evidence that PA breaks in lectures can buffer university students’ perceived stress [ 29 ], lead to better mood ratings [ 29 , 31 ], and increase in motivation [ 28 , 69 ], vigor [ 34 ], energy [ 30 ], and self-perceived physical and psychological well-being [ 28 ]. Looking at dysfunctional stress, the result point that LTPA counteract load-related states of strain such as inner tension, irritability and nervous restlessness or feelings of boredom [ 27 ]. In contrast, short PA breaks during the day could not have enough impact in countering dysfunctional stress at the end of the day regarding the accumulation of negative stressors during home studying which might have occurred after the participant took PA breaks. Other studies have been able to show a reduction in tension [ 30 ] and general muscular discomfort [ 33 ] after PA breaks. However, this was measured as an immediate effect of PA breaks and not with general evening surveys. Blasche and colleagues [ 34 ] measured effects immediately and 20 min after different kind of breaks and found that PA breaks led to an additional short‐ and medium‐term increase in vigor while the relaxation break lead to an additional medium‐term decrease in fatigue compared to an unstructured open break. This is consistent with the results of the present study that an effect of PA breaks is only observed for functional stress and not for dysfunctional stress. Furthermore, there is evidence that long sitting during lectures leads to increased fatigue and lower concentration [ 31 , 70 ], which could be counteracted by PA breaks. For both types of stress loads, functional and dysfunctional stress, there is an influence of students´ well-being in this study. This shows that the stress load is affected by the way students have mentally felt over the last two weeks. The relevance of monitoring this seems important especially in the time of COVID-19 as, for example, 65.3% of the students of a cross-sectional online survey at an Australian university reported low to very low well-being during that time [ 71 ]. However, since PA and well-being can support functional stress load, they should be of the highest priority—not only as regards the pandemic, but also in general.

Looking at hypothesis 1.B; while there is a positive influence of LTPA on experienced psychological detachment, no significant influence for PA breaks was detected. The fact that only LTPA has a positive effect can be explained by the voluntary character of the activity [ 50 ]. The voluntary character ensures that stressors no longer affect the student and, thus, recovery as detachment can take place. Home studying is not present in leisure times, and thus detachment from study is easier. The PA break videos, on the other hand, were shot in a university setting, which would have made it more difficult to detach from study. In order to further understand how PA breaks affect recovery and whether there is a distinction between PA breaks and LTPA, future research should also consider other types of recovery (e.g. relaxation, mastery, and control). Additionally, different types of PA breaks, such as group PA breaks taken on-site versus video-based PA breaks, should be taken into account.

Considering the confirmed positive associations for academic performance-related parameters of hypothesis 2, the results are in accordance with the evidence of positive associations between PA and learning and educational success [ 6 ], as well as between PA breaks and better cognitive functioning [ 28 ]. Looking at the self-reported attention difficulties of hypothesis 2.A, only LTPA can counteract it. PA breaks showed no effects, contrary to the results of a study of Löffler and collegues (2011, [ 31 ]), in which acute effects of PA breaks could be found for higher attention and cognitive performance. Furthermore, the perception of study demands before the study periods has a positive impact on difficulties in attention. That means that overload in studies, time pressure during studies, and incompatibility of studies and private life leads to higher difficulties with attention in home studying. In these conditions, PA breaks might have been seen as interfering, resulting in the expected beneficial effects of exercise on attention and task-related participation behavior [ 72 , 73 ] therefore remaining undetected. With respect to the COVID-19 pandemic, accompanying education changes, and an increase in student´s worries [ 74 , 75 ], the perception of study demands could be affected. This suggests that especially in times of constraint and changes, it is important to promote PA in order to counteract attention difficulties. This also applies to post-pandemic phase.

Regarding the perceived academic performance of hypothesis 2.B, both PA breaks and LTPA have a positive effect on perceived study ability. This result confirms the positive short-term effects on cognition tasks [ 76 ]. It is also in line with the positive function of PA breaks in interrupting sedentary behavior and therefore counteracting the negative association between sitting behavior and lower cognitive performance [ 24 ]. Additionally, this result also fits with the previously mentioned positive relationship between LTPA and functional stress and between PA breaks and functional stress.

According to hypothesis 3, in relation to the mentioned stress load and recovery parameters, there are negative effects of the daily number of breaks during home studying and the longest stretch of time without a break spent on home studying on psychological detachment. As stressors result in negative activation, which impede psychological detachment from study during non-studying time [ 25 ], it was expected and confirmed that the longest stretch of time without a break spent on home studying has a negative effect on detachment. Initially unexpected, the number of breaks has a negative influence on psychological detachment, as breaks could prevent the accumulation of strain reactions. However, if the breaks had no recovery effect through successful detachment, the number might not have any influence on recovery via detachment. This is indicated by the PA breaks, which had no impact on psychological detachment. Since there are other ways to recover from stress besides psychological detachment, such as relaxation, mastery, and control [ 53 ], PA breaks must have had an additional impact in relation to the positive results for functional stress.

In relation to the mentioned academic performance-related parameters, only the number of breaks has a positive influence on the perceived study ability. This indicates that not only PA breaks but also breaks in general lead to better perceived functionality in studying. Paulus and colleagues (2021) found out that an increase in cognitive skills is not only attributed to PA breaks and standing breaks, but also to open breaks with no special instructions [ 28 ]. Either way, they found better improvement in self-perceived physical and psychological well-being of the university students with PA breaks than with open breaks. This is also reflected in the present study with the aforementioned positive effects of PA breaks on functional stress, which does not apply to the number of breaks.

Overall, it must be considered that the there is a more complex network of associations between the examined parameters. The hypothesized separate relation of PA with different parameters do not consider associations between parameters of stress load / recovery and academic performance although there might be a interdependency. Furthermore, moderation aspects were not examined. For example, PA could be a moderator which buffer negative effects of stress on the study ability [ 55 ]. Moreover, perceived study ability might moderate stress levels and academic performance. Further studies should try to approach and understand the different relationships between the parameters in its complexity.

Limitations

Certain limitations must be taken into account. Regarding the imbalanced design toward more female students in the sample (47 female versus 6 male), possible sampling bias cannot be excluded. Gender research on students' emotional states during COVID-19, when this study took place, or students´ acceptance of PA breaks is diverse and only partially supplied with inconsistent findings. For example, during the COVID-19 pandemic, some studies reported that female students were associated with lower well-being [ 71 ] or worse mental health trajectories [ 75 , 77 ]. Another study with a large sample of students from 62 countries reported that male students were more strongly affected by the pandemic because they were significantly less satisfied with their academic life [ 74 ]. However, Keating and colleges (2020) discovered that, despite the COVID-19 pandemic, females rated some aspects of PA breaks during lectures more positively than male students did. However, this was also based on a female slanted sample [ 78 ]. Further studies are needed to get more insights into gender bias.

Furthermore, the small sample size combined with up to 16% missing values comprises a significant short-coming. There were a lot of possibilities which could cause such missing data, like refused, forgotten or missed participation, technical problems, or deviation of the personal code for the questionnaire between survey times. Although the effects could be excluded by sensitive analysis due to missing data, the sample is still small. To generalize the findings, future replication studies are needed.

Additionally, PA breaks were only captured through participation in the ESD, the specially instructed PA break via video. Effects of other short PA breaks were not include in the study. However, participants were called to participate in ESD whenever possible, so the likelihood that they did take part in PA breaks in addition to the ESD could be ignored.

With respect to the baseline variables, it must be considered that two variables (stress load, attention difficulties) were adjusted not with their identical variable in T0, but with other conceptually associated variables (well-being index, BARI-S). Indeed, contrary to the assumption the well-being index does only show an association with functional stress, indicating that it does not control dysfunctional stress. Although the other three assumed associations were confirmed there might be a discrepancy between the daily measured variables and the variables measured in T0. Further studies should either proof the association between these used variables or measure the same variables in T0 for control the daily value of these variables.

Moreover, the measuring instruments comprised the self-assessed perception of the students and thus do not provide an objective information. This must be considered, especially for measuring cognitive and academic-performance-related measures. Here, existing objective tests, such as multiple choice exams after a video-taped lecture [ 72 ] might have also been used. Nevertheless, such methods were mostly used in a lab setting and do not reflect reality. Due to economic reasons and the natural learning environment, such procedures were not applied in this study. However, the circumstances of COVID-19 pandemic allowed a kind of lab setting in real life, as there were a lot of restrictions in daily life which limited the influence of other covariates. The study design provides a real natural home studying environment, producing results that are applicable to the healthy way that students learn in the real world. As this study took place under the conditions of COVID-19, new transformations in studying were also taken into account, as home studying and digital learning are increasingly part of everyday study.

However, the restrictions during the COVID-19 pandemic could result in a greater extent of leisure time per se. As the available leisure time in general was not measured on daily level, it is not possible to distinguish if the examined effects on the outcomes are purely attributable to PA. It is possible that being more physical active is the result of having a greater extent of leisure time and not that PA but the leisure time itself effected the examined outcomes. To address this issue in future studies, it is necessary to measure the proportion of PA in relation to the leisure time available.

Furthermore, due to the retrospective nature of the daily assessments of the variables, there may be overstated associations which must be taken into account. Anyway, the daily level of the study design provides advantages regarding the ability to observe changes in an individual's characteristics over the period of the study. This design made it possible to find out the necessity to analyze the hierarchical structure of the intraindividual data nested within the interindividual data. The performed multilevel analyses made it possible to reflect the outcome of an average student in the sample at his/her daily average behavior in PA and home study.

Conclusion and practical implications

The current findings confirm the importance of PA for university students` stress load, recovery experience, and academic performance-related parameters in home studying. Briefly summarized, it can be concluded that PA breaks positively affect stress load and perceived study ability. LTPA has a positive impact on stress load, recovery experience, and academic performance-related parameters regarding attention difficulties and perceived study ability. Following these results, universities should promote PA in both fashions in order to keep their students healthy and functioning: On the one hand, they should offer opportunities to be physically active in leisure time. This includes time, environment, and structural aspects. The university sport department, which offers sport courses and provides sport facilities on university campuses for students´ leisure time, is one good example. On the other hand, they should support PA breaks during the learning process and in the immediate location of study. This includes, for example, providing instructor videos for PA breaks to use while home studying, and furthermore having instructors to lead in-person PA breaks in on-site learning settings like universities´ libraries or even lectures and seminars. This not only promotes PA, but also reduces sedentary behavior and thereby reduces many other health risks. Further research should focus not only on the effect of PA behavior but also of sedentary behavior as well as the amount of leisure time per se. They should also try to implement objective measures for example on academic performance parameters and investigate different effect directions and possible moderation effects to get a deeper understanding of the complex network of associations in which PA plays a crucial role.

Availability of data and materials

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

Attention and Performance Self-Assessment

"Berliner Anforderungen Ressourcen-Inventar – Studierende" (instrument for the perception of study demands and resources)

Centered within

Grand centered

“Erfassung von Emotionen und Beanspruchung “ (questionnaire containing a word list of adjectives for the recording of emotions and stress during work)

Exercise snack digital (special physical activity break offer)

Intra-Class-Correlation

Leisure time physical activity

• Physical activity

Recovery Experience Questionnaire

Study ability index

World Health Organization-Five Well-being index

Knight JA. Physical inactivity: associated diseases and disorders. Ann Clin Lab Sci. 2012;42(3):320–37.

PubMed   Google Scholar  

Kemel PN, Porter JE, Coombs N. Improving youth physical, mental and social health through physical activity: a systematic literature review. Health Promot J Austr. 2002;33(3):590–601.

Article   Google Scholar  

Gothe NP, Ehlers DK, Salerno EA, Fanning J, Kramer AF, McAuley E. Physical activity, sleep and quality of life in older adults: influence of physical, mental and social well-being. Behav Sleep Med. 2020;18(6):797–808.

Article   PubMed   Google Scholar  

Eime RM, Young JA, Harvey JT, Charity MJ, Payne WR. A systematic review of the psychological and social benefits of participation in sport for children and adolescents: informing development of a conceptual model of health through sport. Int J Behav Nutr Phys Act. 2013;10(1):1–21.

Google Scholar  

Iannotti RJ, Janssen I, Haug E, Kololo H, Annaheim B, Borraccino A. Interrelationships of adolescent physical activity, screen-based sedentary behaviour, and social and psychological health. Int J Public Health. 2009;54:191–8.

Dadaczynski K, Schiemann S. Welchen Einfluss haben körperliche Aktivität und Fitness im Kindes-und Jugendalter auf Bildungsoutcomes? German J Exerc Sport Res. 2015;4(45):190–9.

Kari JT, Pehkonen J, Hutri-Kähönen N, Raitakari OT, Tammelin TH. Longitudinal associations between physical activity and educational outcomes. Med Sci Sports  Exerc. 2017;49(11).

Grim M, Hortz B, Petosa R. Impact evaluation of a pilot web-based intervention to increase physical activity. Am J Health Promot. 2011;25(4):227–30.

Irwin JD. Prevalence of university students’ sufficient physical activity: A systematic review. Percept Mot Skills. 2004;98:927–43.

Kwan MY, Cairney J, Faulkner GE, Pullenayegum EE. Physical activity and other health-risk behaviors during the transition into early adulthood: a longitudinal cohort study. Am J Prev Med. 2012;42(1):14–20.

John JM, Gropper H, Thiel A. The role of critical life events in the talent development pathways of athletes and musicians: A systematic review. Psychol Sport Exerc. 2019;45.

Bopp M, Bopp C, Schuchert M. Active transportation to and on campus is associated with objectively measured fitness outcomes among college students. J Phys Act Health. 2015;12(3):418–23.

Dadaczynski K. Stand der Forschung zum Zusammenhang von Gesundheit und Bildung. Überblick und Implikationen für die schulische Gesundheitsförderung. Zeitschrift für Gesundheitspsychologie. 2012;20(3):141–53

Suhrcke M, de Paz NC. The impact of health and health behaviours on educational outcomes in high-income countries: a review of the evidence. Copenhagen: WHO Regional Offi ce for Europe; 2011.

Lynch BM, Owen N. Too much sitting and chronic disease risk: steps to move the science forward. Ann Intern Med. 2015;16(2):146–7.

Voss MW, Carr LJ, Clark R, Weng T. Revenge of the “sit” II: Does lifestyle impact neuronal and cognitive health through distinct mechanisms associated with sedentary behavior and physical activity? Ment Health Phys Act. 2014;7(1):9–24.

Huber G. Ist Sitzen eine tödliche Aktivität? B&G Bewegungstherapie und Gesundheitssport. 2014;30(01):13–6.

Peterson NE, Sirard JR, Kulbok PA, DeBoer MD, Erickson JM. Sedentary behavior and physical activity of young adult university students. Res Nurs Health. 2018;4(1):30–8.

Rupp R, Dold C, Bucksch J. Sitzzeitreduktion und Bewegungsaktivierung in der Hochschullehre – Entwicklung und Implementierung der Mehrebenen-Intervention Kopf-Stehen. Die Hochschullehre. 2019;5:525–42.

Ickes MJ, McMullen J, Pflug C, Westgate PM. Impact of a University-based Program on Obese College Students’ Physical Activity Behaviors, Attitudes, and Self-efficacy. Am J Health Educ. 2016;47(1):47–55.

Lepp A, Barkley JE, Karpinski AC. The relationship between cell phone use and academic performance in a sample of US college students. Sage Open. 2015;5(1).

Stapp AC, Prior LF. The Impact of Physically Active Brain Breaks on College Students’ Activity Levels and Perceptions. J Physic Activ Res. 2018;3(1):60–7.

Fuchs R, Klaperski S. Sportliche Aktivität und Stressregulation. In: Fuchs R, Schlicht W, editors. Sportaktivität und seelische Gesundheit Göttingen: Hogrefe; 2012. p. 100–21.

Falck RS, Davis JC, Liu-Ambrose T. What is the association between sedentary behaviour and cognitive function? A systematic review. Br J Sports Med. 2017;51(10):800–11.

Sonnentag S, Fritz C. Recovery from job stress: The stressor-detachment model as an integrative framework. J Organ Behav. 2015;36:72–103.

Edwards JR. A cybernetic theory of stress, coping, and well-being in organizations. Acad Manag Rev. 1992;17(2):238–74.

Wieland R. Status-Bericht: Psychische Gesundheit in der betrieblichen Gesundheitsförderung – eine arbeitspsychologische Perspektive. In: Nold H, Wenninger G, editors. Rückengesundheit und psychische Gesundheit. Rückengesundheit und psychische Gesundheit.: Asanger Verlag; 2013.

Paulus M, Kunkel J, Schmidt SCE, Bachert P, Wäsche H, Neumann R, et al. Standing breaks in lectures improve university students’ self-perceived physical, mental, and cognitive condition. Int J Environ Res Public Health. 2021;18.

Marschin V, Herbert C. A Short, Multimodal Activity Break Incorporated Into the Learning Context During the Covid-19 Pandemic: Effects of Physical Activity and Positive Expressive Writing on University Students’ Mental Health — Results and Recommendations From a Pilot Study. Front Psychol. 2021;12.

Gollner E, Savil M, Schnabel F, Braun C, Blasche G. Unterschiede in der Wirksamkeit von Kurzpausenaktivitäten im Vergleich von Bewegungspausen zu psychoregulativen Pausen bei kognitiver Belastung. Bewegungstherapie Gesundheitssport. 2019;35:134–43.

Löffler SN, Dominok E, von Haaren B, Schellhorn R, Gidion G. Aktivierung, Konzentration, Entspannung: Interventionsmöglichkeiten zur Förderung fitnessrelevanter Kompetenzen im Studium: KIT Scientific Publishing; 2011.

Marschin V, Herbert C. A short, multimodal activity break incorporated into the learning context during the Covid-19 pandemic: effects of physical activity and positive expressive writing on university students' mental health—results and recommendations from a pilot study. Front Psychol. 2021.

Kowalsky RJ, Farney TM, Hearon CM. Resistance Exercise Breaks Improve Ratings of Discomfort and Sleepiness in College Students. Res Q Exerc Sport. 2022;94(1):210–5.

Blasche G, Szabo B, Wagner-Menghin M, Ekmekcioglu C, Gollner E. Comparison of rest-break interventions during a mentally demanding task. Stress Health. 2018;34(5):629–38.

Article   PubMed   PubMed Central   Google Scholar  

Fuchs R, Klaperski S. Stressregulation durch Sport und Bewegung. In: Fuchs R, Gerber M, editors. Handbuch Stressregulation und Sport. Berlin: Springer; 2018. p. 205–26.

Chapter   Google Scholar  

Kaluza G, Renneberg B. Stressbewältigung. In: Bengel J, Jerusalem M, editors. HandbuchGesundheitspsychologie und medizinische Psychologie. Göttingen: Hogrefe; 2009. p. 265–72.

Klaperski S. Exercise, Stress and Health: The Stress-Buffering Effect of Exercise. In: Fuchs R, Gerber M, editors. Handbuch Stressregulation und Sport. Berlin: Springer; 2018. p. 227–50.

Lesser IA, Nienhuis CP. The impact of COVID-19 on physical activity behavior and well-being of Canadians. Int J Environ Res Public Health. 2020;17(11):3899.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Moore SA, Faulkner G, Rhodes RE, Brussoni M, Chulak-Bozzer T, Ferguson LJ, et al. Impact of the COVID-19 virus outbreak on movement and play behaviours of Canadian children and youth: a national survey. Int J Behav Nutr Phys Act. 2020;17(1):1–11.

Rodríguez-Larrad A, Mañas A, Labayen I, González-Gross M, Espin A, Aznar S, et al. Impact of COVID-19 confinement on physical activity and sedentary behaviour in Spanish university students: Role of gender. Int J Environ Res Public Health. 2021;18(2):369.

Stanton R, To QG, Khalesi S, Williams SL, Alley SJ, Thwaite TL, et al. Depression, anxiety and stress during COVID-19: associations with changes in physical activity, sleep, tobacco and alcohol use in Australian adults. Int J Environ Res Public Health. 2020;17(11):4065.

Zheng C, Huang WY, Sheridan S, Sit CHP, Chen XK, Wong SHS. COVID-19 pandemic brings a sedentary lifestyle in young adults: a cross-sectional and longitudinal study. Int J Environ Res Public Health. 2020;17(17):6035.

Commodari E. Attention Skills and Risk of Developing Learning Difficulties. Curr Psychol. 2012;31:17–34.

Commodari E, Guarnera M. Attention and reading skills. Percept Mot Skills. 2005;100:3753–86.

Raaijmakers MAJ, Smidts DP, Sergeant JA, Maassen GH, Posthumus JA, van Engeland H, et al. Executive functions in preschool children with aggressive behavior: impairments in inhibitory control. J Abnorm Child Psychol. 2008;36:1097–107.

Vellutino FR, Scanlon DM, Sipay ER, Small SG, Pratt A, Chen RS, et al. Cognitive profiles of difficulty to remediate and readily remediate poor readers: early intervention as a vehicle for distinguishing between cognitive and experiential deficits as basic of specific reading disability. J Educ Psychol. 1996;88:601–38.

Ilies R, Dimotakis N, De Pater IE. Psychological and physiological reactions to high workloads: Implications for well-being. Pers Psychol. 2010;63(2):407–36.

Rodell JB, Judge TA. Can “good” stressors spark “bad” behaviors? The mediating role of emotions in links of challenge and hindrance stressors with citizenship and counterproductive behaviors. J Appl Psychol. 2009;94(6).

Teuber M, Leyhr D, Moll J, Sudeck G. Nudging digital physical activity breaks for home studying of university students—A randomized controlled trial during the COVID-19 pandemic with daily activity measures. Front Sports Active Living. 2022;4.

Feuerhahn N, Sonnentag S, Woll A. Exercise after work, psychological mediators, and affect: A day-level study. Eur J Work Organ Psy. 2014;23(1):62–79.

Sonnentag S. Work, Recovery Activities, ans Individual Well-Being: A Diary Study. J Occup Health Psychol. 2001;6(3):196–210.

Article   CAS   PubMed   Google Scholar  

Wieland R. Gestaltung gesundheitsförderlicher Arbeitsbedingungen. In: Kleinbeck U, Schmidt K-H, editors. Arbeitspsychologie (Enzyklopädie der Psychologie, Serie Wirtschafts-, Organisations- und Arbeitspsychologie). 1. Göttingen: Hogrefe; 2010. p. 869–919.

Sonnentag S, Fritz C. The Recovery Experience Questionnaire: development and validation of a measure for assessing recuperation and unwinding from work. J Occup Health Psychol. 2007;12(3):204–21.

Bankstahl US, Görtelmeyer R. Measuring subjective complaints of attention and performance failures development and psychometric validation in tinnitus of the self-assessment scale APSA. Health and Quality of Life Outcomes. 2013;11(86).

Teuber M, Arzberger I, Sudeck G. Körperliche Aktivität, Gesundheit und Funktionsfähigkeit im Studium: Sportliche Freizeitaktivitäten und aktive Fortbewegung als Ressource im Studium? In: Göring A, Mayer J, Jetzke M, editors. Sport und Studienerfolg - Analysen zur Bedeutung sportlicher Aktivität im Setting Hochschule. Hochschulsport: Bildung und Wissenschaft, 4. Göttingen: Universitätsverlag Göttingen; 2020. p. 27–49.

Hasselhorn H-M, Freude G. Der Work-Ability-Index: ein Leitfaden In: Arbeitsmedizin BfAu, editor. Dortmund/Berlin/Dresden: Wirtschaftsverl. NW, Verlag für Neue Wissenschaft GmbH; 2007.

Rütten A, Pfeifer K. Nationale Empfehlungen für Bewegung und Bewegungsförderung. Köln: Bundeszentrale für Gesundheitliche Aufklärung (BZgA); 2017.

Brähler E, Mühlan H, Albani C, Schmidt S. Teststatistische Prüfung und Normierung der deutschen Versionen des EUROHIS-QOL Lebensqualität-Index und des WHO-5 Wohlbefindens-Index. Diagnostica. 2007;53(2):83–96.

Gusy B, Wörfel F, Lohmann K. Erschöpfung und Engagement im Studium. Zeitschrift für Gesundheitspsychologie. 2016;24(1):41–53.

Elfering A, Grebner S, de Tribolet-Hardy F. The long arm of time pressure at work: Cognitive failure and commuting near-accidents. Eur J Work Organ Psy. 2013;22(6):737–49.

Kreft IG, de Leeuw J. Introducing multilevel modeling. London: Sage; 1998.

Book   Google Scholar  

Bates D, Mächler M, Bolker BM, Walker SC. Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw. 2015;67(1):1–48.

Hofmann DA, Gavin MB. Centering decisions in hierarchical linear models: Implications for research in organizations. J Manag. 1998;24(5):623–41.

Nezlek J. Diary Studies in Social and Personality Psychology: An Introduction With Some Recommendations and Suggestions. Social Psychological Bulletin. 2020;15(2).

Knapp G. Gemischte Modelle in R. Begleitskriptum zur Weiterbildung. In: Dortmund TU, editor. Braunschweig2019.

Van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate imputation by chained equations in R. J Stat Softw. 2011;45:1–67.

von Hippel PT. How Many Imputations Do You Need? A Twostage Calculation Using a Quadratic Rule. Sociological Methods & Research. 2020;49(3):699–718.

Article   MathSciNet   Google Scholar  

Robinson D. broom: An R package for converting statistical analysis objects into tidy data frames. arXiv preprint arXiv:14123565. 2014.

Young-Jones A, McCain J, Hart B. Let’s Take a Break: The Impact of Physical Activity on Academic Motivation. Int J Teach Learn High Educ. 2022;33(3):110–8.

Barr-Anderson DJ, AuYoung M, Whitt-Glover MC, Glenn BA, Yancey AK. Integration of short bouts of physical activity into organizational routine: A systematic review of the literature. Am J Prev Med. 2011;40(1):76–93.

Dodd RH, Dadaczynski K, Okan O, McCaffery KJ, Pickles K. Psychological Wellbeing and Academic Experience of University Students in Australia during COVID-19. Int J Environ Res Public Health 2021;18.

Fenesi B, Lucibello K, Kim JA, Heisz JJ. Sweat so you don’t forget: exercise breaks during a university lecture increase on-task attention and learning. J Appl Res Mem Cogn. 2018;7(2):261–9.

Ruhland S, Lange KW. Effect of classroom-based physical activity interventions on attention and on-task behavior in schoolchildren: A systematic review. Sports Med Health Sci. 2021;3:125–33.

Aristovnik A, Keržič D, Ravšelj D, Tomaževič N, Umek L. Impacts of the COVID-19 Pandemic on Life of Higher Education Students: A Global Perspective. Sustainability. 2020;12(20).

Browning MHEM, Larson LR, Sharaievska I, Rigolon A, McAnirlin O, Mullenbach L, et al. Psychological impacts from COVID-19 among university students: Risk factors across seven states in the United States. PLoS ONE 2021;16(1).

Chang Y-K, Labban JD, Gapin JI, Etnier JL. The effects of acute exercise on cognitive performance: a meta-analysis. Brain Res. 2012;1453:87–101.

Elmer T, Mepham K, Stadtfeld C. Students under lockdown: Comparisons of students’ social networks and mental health before and during the COVID-19 crisis in Switzerland. PLoS ONE. 2020;15(7):e0236337.

Keating R, Ahern S, Bisgood L, Mernagh K, Nicolson GH, Barrett EM. Stand up, stand out. Feasibility of an active break targeting prolonged sitting in university students. J Am Coll Health. 2020;70(7).

Download references

Acknowledgements

We would like to thank Juliane Moll, research associate of the Student Health Management of University of Tübingen, for the support in the coordination and realization study. We would like to express our thanks also to Ingrid Arzberger, Head of University Sports at the University of Tübingen, for providing the resources and co-applying for the funding. We acknowledge support by Open Access Publishing Fund of University of Tübingen.

Open Access funding enabled and organized by Projekt DEAL. This research regarding the conduction of the study was funded by the Techniker Krankenkasse, health insurance fund.

Author information

Authors and affiliations.

Institute of Sports Science, Faculty of Economics and Social Sciences, University of Tübingen, Tübingen, Germany

Monika Teuber, Daniel Leyhr & Gorden Sudeck

Methods Center, Faculty of Economics and Social Sciences, University of Tübingen, Tübingen, Germany

Daniel Leyhr

Interfaculty Research Institute for Sports and Physical Activity, University of Tübingen, Tübingen, Germany

Gorden Sudeck

You can also search for this author in PubMed   Google Scholar

Contributions

M.T. and G.S. designed the study. M.T. coordinated and carried out participant recruitment and data collection. M.T. analyzed the data and M.T. and D.L. interpreted the data. M.T. drafted the initial version of the manuscript and prepared the figure and all tables. All authors contributed to reviewing and editing the manuscript and have read and agreed to the final version of the manuscript.

Corresponding author

Correspondence to Monika Teuber .

Ethics declarations

Ethics approval and consent to participate.

The study involves human participants and was reviewed and approved by the Ethics Committee of the Faculty of Social Sciences and Economics, University of Tübingen (ref. A2.54-127_kr). The participants provided their written informed consent to participate in this study. All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary file 1., supplementary file 2., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Teuber, M., Leyhr, D. & Sudeck, G. Physical activity improves stress load, recovery, and academic performance-related parameters among university students: a longitudinal study on daily level. BMC Public Health 24 , 598 (2024). https://doi.org/10.1186/s12889-024-18082-z

Download citation

Received : 30 June 2023

Accepted : 12 February 2024

Published : 24 February 2024

DOI : https://doi.org/10.1186/s12889-024-18082-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Physical activity breaks
• Stress load
• Psychological detachment
• Academic performance
• Study ability
• University students

BMC Public Health

ISSN: 1471-2458

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Transforming stress through awareness, education and collaboration.

Stress Research

“The difficulty in science is often not so much how to make the discovery but rather to know that one has made it.” – J.D. Bernal

2024 Stress Statistics

The 2024 results of the American Psychiatric Association’s annual mental health poll show that U.S. adults are feeling increasingly anxious. In 2024, 43% of adults say they feel more anxious than they did the previous year, up from 37% in 2023 and 32% in 2022. Adults are particularly anxious about current events (70%) — especially the economy (77%), the 2024 U.S. election (73%), and gun violence (69%).

When asked about a list of lifestyle factors potentially impacting mental health, adults most commonly say stress (53%) and sleep (40%) have the biggest impact on their mental health. Younger adults (18-34 years old) are more likely than older adults (50+) to say social connection has the biggest impact on their mental health. Despite the increasing anxiety, most adults have not sought professional mental health support. In 2024, just one in four (24%) adults say they talked with a mental health care professional in the past year. Notably, younger adults (18-34) are more than twice as likely as older adults (50+) to have done so.

“Living in a world of constant news of global and local turmoil, some anxiety is natural and expected,” said APA President Petros Levounis, M.D., M.A. “But what stands out here is that Americans are reporting more anxious feelings than in past years. This increase may be due to the unprecedented exposure that we have to everything that happens in the world around us, or to an increased awareness and reporting of anxiety. Either way, if people have these feelings, they are not alone, and they can seek help from us.”

Among adults who have used mental health care this year, more than half prefer to meet with a mental health professional in person (55%) rather than via telehealth; 30% prefer telehealth; and 15% have no preference. Also among adults who have used mental healthcare this year, more than half (59%) are worried about losing access to mental healthcare, and 39% of insured adults are worried about losing their health insurance, as a result of the election this year.

Americans perceive broad impacts of untreated mental illness: 83% of adults say it negatively impacts families and 65% say it negatively impacts the U.S. economy. Also, 71% of adults feel that children and teens have more mental health problems than they did 10 years ago. That said, more than half of adults (55%) think there is less mental health stigma than 10 years ago.

“Over the past ten years, we’ve grown more comfortable talking about mental health, and that’s absolutely key to helping us through the current crisis,” said APA CEO and Medical Director Saul Levin, M.D., M.P.A. “The continued work of APA is to ensure that people can access care when they need it, especially in areas that need it badly, like child and adolescent psychiatry.”

Other issues people said they were anxious about include:

• Keeping themselves or their families safe, 68%.
• Keeping their identity safe, 63%.
• Their health, 63%.
• Paying bills or expenses, 63%.
• The opioid epidemic, 50%.
• The impact of emerging technology on day-to-day life, 46%.

In addition, 57% of adults are concerned about climate change.

This annual poll was conducted April 9 to 11, 2024, among a sample of more than 2,200 adults. This annual survey is complemented by APA’s Healthy Minds Monthly series, conducted by Morning Consult on behalf of APA. See  past Healthy Minds Monthly polls . For a copy of the results, contact us at  [email protected] .

American Psychiatric Association

The American Psychiatric Association, founded in 1844, is the oldest medical association in the country. The APA is also the largest psychiatric association in the world with more than 38,900 physician members specializing in the diagnosis, treatment, prevention, and research of mental illnesses. APA’s vision is to ensure access to quality psychiatric diagnosis and treatment.

Causes and Sources of Stress

Living conditions, the political climate, financial insecurity, and work issues are some stressors US adults cite as the cause of their stress. Ineffective communications increase work stress to the point of frustration that workers want to quit.  These stressors, unfortunately, are not something people can just ignore. Quitting a job would result in debt and financial instability which, in turn, would be added stressors.

• 35% of workers say their boss is a cause of their workplace stress.
• 80% of US workers experience work stress because of ineffective company communications.
• 39% of North American employees report their workload the main source of the work stress.
• 49% of 18 – 24 year olds who report high levels of stress felt comparing themselves to others is a stressor.
• 71% of US adults with private health insurance say the cost of healthcare causes them stress while 53% with public insurance say the same.
• 54% of Americans want to stay informed about the news but following the news causes them stress.
• 42% of US adults cite personal debt as a source of significant stress.
• 1 in 4 American adults say discrimination is a significant source of stress.
• Mass shootings are a significant source of stress across all races; 84% of Hispanic report this, the highest among the races.

Stress and Relationships

People under stress admit to taking out their frustration on other people. Targets for venting out include strangers and those they have personal relationships with. Men and women report different levels of how work stress affects their relationships with their spouses.

• 76% of US workers say their workplace stress has had a negative impact on their personal relationships.
• Seven in 10 adults report work stress affects their personal relationships.
• 79% of men report work stress affects their personal relationship with their spouse compared to 61% for women.
• 36% of adults reported experiencing stress caused by a friend or loved one’s long-term health condition.

Stress Management Statistics

A look at the stress management techniques employed by US adults to deal with their stress, an overwhelming majority are self-care practices. Though very helpful, it does not address the stressor at the root of the problem. Stress management programs would be beneficial not only for employees but for the company in the long run.

• 30% of Us adults eat comfort food “more than the usual” when faced with a challenging or stressful event.
• 51% of US adults engage in prayer—a routine activity—when faced with a challenge or stressful situation.
• Coping mechanisms of Gen Z and Millenials experiencing stress in the US 44% of Gen Z and 40% of Millenials sleep in while exercising counts for 14% and 20% respectively.
• 49% of US adults report enduring stressful situations as a coping behavior to handle stress.
• Less than 25% of those with depression worldwide have access to mental health treatments.

CompareCamp

American Psychological Association

Cardiac Coherence and Post-traumatic Stress Disorder in Combat Veterans

Jay P. Ginsberg, Ph.D.; Melanie E. Berry, M.S.; Donald A Powell, Ph.D.

Alternative Therapies in Health and Medicine, A Peer-Reviewed Journal, 2010;16 (4):52-60. PDF version of the complete paper: Cardiac Coherence and PTSD in Combat Veterans

Abstract-PTSD

Background: The need for treatment of posttraumatic stress disorder (PTSD) among combat veterans returning from Afghanistan and Iraq is a growing concern. PTSD has been associated with reduced cardiac coherence (an indicator of heart rate variability [HRV]) and deficits in early-stage information processing (attention and immediate memory) in different studies. However, the co-occurrence of reduced coherence and cognition in combat veterans with PTSD has not been studied before.

Primary Study Objective: A pilot study was undertaken to assess the covariance of coherence and information processing in combat veterans. An additional study goal was an assessment of the effects of HRV biofeedback (HRVB) on coherence and information processing in these veterans.

Methods/Design: A two-group (combat veterans with and without PTSD), a pre-post study of coherence and information processing was employed with baseline psychometric covariates.

Setting: The study was conducted at a VA Medical Center outpatient mental health clinic.

Participants: Five combat veterans from Iraq or Afghanistan with PTSD and five active-duty soldiers with comparable combat exposure who were without PTSD.

Intervention: Participants met with an HRVB professional once weekly for 4 weeks and received visual feedback in HRV patterns while receiving training in resonance frequency breathing and positive emotion induction.

Primary Outcome Measures: Cardiac coherence, word list learning, commissions (false alarms) in go—no go reaction time, digits backward.

Results: Cardiac coherence was achieved in all participants, and the increase in coherence ratio was significant post-HRVB training. Significant improvements in the information processing indicators were achieved. Degree of increase in coherence was the likely mediator of cognitive improvement.

Conclusion: Cardiac coherence is an index of the strength of control of parasympathetic cardiac deceleration in an individual that has cardinal importance for the individual’s attention and affect regulation.

The Effect of a Biofeedback-based Stress Management Tool on Physician Stress: A Randomized Controlled Clinical Trial

Jane B. Lemaire, Jean E. Wallace, Adriane M. Lewin, Jill de Grood, Jeffrey P. Schaefer

Open Medicine 2011; 5(4)E154. PDF version of the complete paper: physician-stress-randomized-controlled-clinical-trial

Abstract- Biofeedback-based Stress Management

Background: Physicians often experience work-related stress that may lead to personal harm and impaired professional performance. Biofeedback has been used to manage stress in various populations.

Objective: To determine whether a biofeedback-based stress management tool, consisting of rhythmic breathing, actively self-generated positive emotions and a portable biofeedback device, reduces physician stress.

Design: Randomized controlled trial measuring the efficacy of a stress-reduction intervention over 28 days, with a 28-day open-label trial extension to assess effectiveness.

Setting: Urban tertiary care hospital.

Participants: Forty staff physicians (23 men and 17 women) from various medical practices (1 from primary care, 30 from a medical specialty and 9 from a surgical specialty) were recruited by means of electronic mail, regular mail and posters placed in the physicians’ lounge and throughout the hospital.

Intervention: Physicians in the intervention group were instructed to use a biofeedback-based stress management tool three times daily. Participants in both the control and intervention groups received twice-weekly support visits from the research team over 28 days, with the intervention group also receiving re-inforcement in the use of the stress management tool during these support visits. During the 28-day extension period, both the control and the intervention groups received the intervention, but without intensive support from the research team.

Main outcome measure: Stress was measured with a scale developed to capture short-term changes in global perceptions of stress for physicians (maximum score 200).

Results: During the randomized controlled trial (days 0 to 28), the mean stress score declined significantly for the intervention group (change -14.7, standard deviation [SD] 23.8; p = 0.013) but not for the control group (change -2.2, SD 8.4; p = 0.30). The difference in mean score change between the groups was 12.5 (p = 0.048). The lower mean stress scores in the intervention group were maintained during the trial extension to day 56. The mean stress score for the control group changed significantly during the 28-day extension period (change -8.5, SD 7.6; p < 0.001).

Conclusion: A biofeedback-based stress management tool may be a simple and effective stress-reduction strategy for physicians.

Coherence Training In Children With Attention-Deficit Hyperactivity Disorder: Cognitive Functions and Behavioral Changes

Anthony Lloyd, Ph.D.; Davide Brett, B.Sc.; Ketith Wesnes, Ph.D.

Alternative Therapies in Health and Medicine, A Peer-Reviewed Journal, 2010; 16 (4):34-42

PDF version of the complete paper: coherence-training-in-children-with-adhd

Abstract-ADHD

Attention-deficit hyperactivity disorder (ADHD) is the most prevalent behavioral diagnosis in children, with an estimated 500 000 children affected in the United Kingdom alone. The need for an appropriate and effective intervention for children with ADHD is a growing concern for educators and childcare agencies. This randomized controlled clinical trial evaluated the impact of the HeartMath self-regulation skills and coherence training program (Institute of HeartMath, Boulder Creek, California) on a population of 38 children with ADHD in academic year groups 6, 7, and 8. Learning of the skills was supported with heart rhythm coherence monitoring and feedback technology designed to facilitate self-induced shifts in cardiac coherence. The cognitive drug research system was used to assess cognitive functioning as the primary outcome measure. Secondary outcome measures assessed teacher and student reposted changes in behavior. Participants demonstrated significant improvements in various aspects of cognitive functioning such as delayed word recall, immediate word recall, word recognition, and episodic secondary memory. Significant improvements in behavior were also found. The results suggest that the intervention offers a physiologically based program to improve cognitive functioning in children with ADHD and improve behaviors that is appropriate to implement in a school environment.

Coherence and Health Care Cost – RCA Actuarial Study: A Cost-Effectiveness Cohort Study

Woody Bedell; Mariette Kaszkin-Bettag, Ph.D.

Alternative Therapies in Health and Medicine, A Peer-Reviewed Journal, 2010;16 (4):26-31. PDF version of the complete paper: rca-actuarial-study-coherence-and-health-care

Abstract-Health and Medicine

Chronic stress is among the most costly health problems in terms of direct health costs, absenteeism, disability, and performance standards. The Reformed Church in America (RCA) identified stress among its clergy as a major cause of higher-than-average health claims and implemented HeartMath (HM) to help its participants manage stress and increase physiological resilience. The 6-week HM program Revitalize You! was selected for the intervention including the emWave Personal Stress Reliever technology.

From 2006 to 2007, completion of a health risk assessment (HRA) provided eligible clergy with the opportunity to participate in the HM program or a lifestyle management program (LSM). Outcomes for that year were assessed with the Stress and Well-being Survey. Of 313 participants who completed the survey, 149 completed the Revitalize You! The program and 164 completed the LSM. Well-being, stress management, resilience, and emotional vitality were significantly improved in the HM group as compared to the LSM group.

In an analysis of the claims costs data for 2007 and 2008, 144 pastors who had participated in the HM program were compared to 343 non-participants (control group). Adjusted medical costs were reduced by 3.8% for HM participants in comparison with an increase of 9.0% for the control group. For the adjusted pharmacy costs, an increase of 7.9% was found compared with an increase of 13.3% for the control group. Total 2008 savings as a result of the HM program are estimated at $585 per participant, yielding a return on investment of 1.95:1. These findings show that HM stress-reduction and coherence-building techniques can reduce health care costs.

View my collection, “Stress and Cardiovascular Disease” from NCBI

View my collection, “Stress and Cancer” from NCBI

View my collection, “Stress and Diabetes” from NCBI

View my collection, “Post Traumatic Stress Disorder” from NCBI

View my collection, “Stress and Aging” from NCBI

View my collection, “Stress in Adolenscents” from NCBI

View my collection, “Stress and Meditation” from NCBI

View my collection, “Stress and Yoga” from NCBI

View my collection, “Workplace Stress” from NCBI

Want to receive the latest news and info from AIS? Sign up to receive our enewsletter.

What about our FREE magazines?

Contentment Magazine Combat Stress Magazine

Premium Content

Can scientists ‘solve’ stress? They’re trying.

From cardiovascular disease and obesity to a weakened immune system, the side effects of stress can be life-altering. But there may be a way to prevent those outcomes.

As modern-day stress ratchets up to what feels like unbearable levels, researchers are striving to learn more about the precise mechanisms through which it affects our body and mind. The hope is that by unlocking more about how stress works physiologically, we can find ways to prevent it from permanently harming people.

Over the last five decades, scientists have established beyond doubt that persistent stress really can poison our overall health. In addition to increasing the risk of cardiovascular disease , stress plays a role in obesity and diabetes and can weaken the immune system , leaving us more vulnerable to infectious diseases. You can recover swiftly from an episode of acute stress—for example, the alarm one might feel when caught unprepared for a presentation. Chronic stress, on the other hand, is more toxic as it is an unrelenting circumstance that offers little chance for a return to normalcy. Financial strain, having a bully for a boss, and social isolation are all examples.

Today chronic stress seems to be increasing worldwide, as people grapple with rapid socioeconomic and environmental change.   A 2023 national survey by the American Psychological Association found that stress has taken a serious toll since the start of the pandemic , with the incidence of chronic illnesses and mental health problems going up significantly, especially among those ages 35 to 44.

( Do you have chronic stress? Look for these signs. )

So far, one of the major realizations among scientists is that stress harms all of us in different and powerful ways. But is there any way to avoid it—or at least recover more quickly? Some promising avenues of research offer hope for the future.

Preventing chronic stress from harming you in the first place

Groundbreaking studies in orphans showed how stress in early life can leave an indelible mark on the brain.

For Hungry Minds

“Chronic stress in early life has more serious and lasting effects, because that’s when a lot of connections are being laid down in the brain,” says Aniko Korosi, a researcher at the University of Amsterdam who has been conducting experiments on mice to elucidate that link between early-life stress and brain development.

Korosi may have found a surprising link between stress and the resulting nutrient composition in the brain . She and her colleagues noticed that mouse pups that had been exposed to stress in the first week of their lives—having been moved from their mother’s care to a cage—had lower levels of certain fatty acids and amino acids in their brains compared with pups being raised in a stress-free environment.

She wondered if it was possible to normalize a stressed pup’s development by feeding it a diet rich in the specific nutrients its brain would be lacking. To find out, the researchers first fed a supplemented diet to the mothers so it would pass through their milk, then continued to provide it in the pups’ feed for two weeks after they were weaned. A few months later, the researchers tested the now adult mice in learning and memory. Unlike stressed mice that had never received an enriched diet, these mice did not display cognitive impairments.

( How wild animals cope with stress—from overeating to sleepless nights. )

“I was surprised that changing the nutrition could have such a powerful effect, because it’s such an easy intervention,” Korosi says.

If further studies provide more evidence of the nutritional pathway, she says, there would be a strong basis for supplementing the diets of infants born to mothers living in stressful conditions.

Developing an early warning system for stress

Katie McLaughlin, a psychologist at the University of Oregon, is investigating how mental health problems arise in adolescents as they’re going through a particularly vulnerable time in their lives, transitioning to adulthood.

She and her colleagues are still collecting data , but a smaller, precursor study tracking 30 teenagers offers clues about what the researchers might learn—and how it might help them identify stress before it goes too far.  

In that study, McLaughlin found that the extent of stress experienced by a subject in the month before their lab visit changed how their brain responded to emotionally impactful information such as when they were shown a picture of a threatening face. The brain’s prefrontal cortex, which helps regulate emotions, showed less activation when the subject had experienced higher levels of stress.

McLaughlin is optimistic that data from the ongoing study will help pinpoint changes in behavior as well as brain activity that predict the emergence of mental health problems like anxiety and depression. This could enable the development of targeted interventions delivered to teenagers at just the right time, she says. If the identified marker of stress were a sudden decrease in sleep duration or a sharp decline in social interactions, for example, it would be possible to push the intervention out to the individual on their smartphone.

“Like, here’s a reminder about good sleep hygiene, or this might be a good time to check in with your counselor at school about what’s been going on in your life,” McLaughlin explains.

( ‘Hysterical strength’? Fight or flight? This is how your body reacts to extreme stress. )

Learn more about stress and how to manage it

Preventing inflammation caused by chronic stress.

Gaining a deeper understanding of how stress affects the immune system may also help find a way to reverse those effects.

You May Also Like

Does meditation actually work? Here’s what the science says.

How to cope with stress at work—and avoid burning out

20 stress-relief gifts for the frazzled friend in your life

In the 1980s, psychologist Janice Kiecolt-Glaser and her virologist husband, Ronald Glaser, began exploring the physiological impact of stress on two notably stressed segments of society: medical students and older caregivers. The researchers found the students’ immune systems were less robust when they were taking exams than during non-exam times—and that stress altered the body’s response to vaccines.

Researchers then administered the flu and pneumonia vaccines to individuals responsible for a spouse with dementia. Unlike medical students taking exams, who were likely stressed only in the short term, these people were experiencing unrelenting stress. When tested at set periods after inoculation, they had fewer antibodies compared with a control group —they couldn’t maintain their protective response. “That gave us good evidence that the changes brought on by stress were biologically meaningful,” says Kiecolt-Glaser, now an emeritus professor at the Ohio State University.

Around the same time, researchers led by Sheldon Cohen, now emeritus professor of psychology at Carnegie Mellon University, delivered cold-causing viruses into the nostrils of about 400 adult volunteers in the U.K. “The more stress they reported prior to our exposing them to a virus, the higher the risk was for them to develop a cold,” says Cohen. The duration and type of stress mattered: Chronic economic or interpersonal stress were what really put people at high risk—and the longer it went on, the greater the susceptibility to falling sick.

Cohen and his colleagues also learned that when exposed to viruses, chronically stressed people tended to produce an excess of cytokines—proteins that serve as messengers of the immune system, traveling to sites of infection and injury and activating inflammation and other cellular processes to protect the body. Too many cytokines cause an excess of inflammation.

Researchers still don’t know enough about how stress alters the immune system’s ability to regulate cytokines to devise an intervention to reduce the inflammation, but in one way, these findings signal some hope: There are clear targets for more work to be done.  

Understanding stress on a cellular level

The future of understanding and combating stress may lie in our DNA.

In 2023, Ursula Beattie, then a doctoral student at Tufts University, and her colleagues found possible evidence that stress can overwhelm DNA’s repair mechanisms . In their study, researchers repeatedly tapped on sparrow cages with pens, played the radio loudly, and other actions designed to cause distress but no physical harm. Blood and tissue samples from the sparrows after three weeks of this unpleasant treatment revealed damage to the DNA. “It’s like if you had two pieces of string coiled up, just like DNA, and you took a pair of scissors and cut them,” Beattie says.

While these kinds of double-strand breaks in DNA occur all the time in sparrows and other species, including humans, the damage is typically reversed through self-repair mechanisms. In a chronic-stress setting, “those repair mechanisms get overwhelmed, which is how we see a buildup of DNA damage,” Beattie explains. The damage in the birds appears to be the most severe in cells of the liver, she adds, suggesting that for humans, too, the extent and type of damage inflicted by stress might be different for different tissues of the body.

Separately, Kiecolt-Glaser and psychologist Lisa Christian at OSU are conducting a longitudinal study to determine whether chronic stress ages you more quickly. If results support a smaller, earlier study, it appears that chronically stressed caregivers not only are more likely to get sick and heal more slowly but they also show signs of accelerated aging.

We’re still learning how deep stress goes into our bodies. But these exploratory findings mean we’re getting closer to solving the puzzle that is stress, which promises a future where we can better meet the ongoing demand for change.

( 20 stress-relief gifts for the frazzled friend in your life. )

Related Topics

• MENTAL HEALTH
• PUBLIC HEALTH

‘Hysterical strength’? Fight or flight? This is how your body reacts to extreme stress

Extreme heat can be deadly – here’s how to know if you’re at risk

The benefits—and downsides—of taking ashwagandha

What your biological age can reveal about your health

Many people wean off antidepressants too quickly. That can be dangerous.

• Environment
• Perpetual Planet

History & Culture

• History & Culture
• History Magazine
• Mind, Body, Wonder
• Paid Content
• Terms of Use
• Privacy Policy
• Your US State Privacy Rights
• Children's Online Privacy Policy
• Interest-Based Ads
• About Nielsen Measurement
• Do Not Sell or Share My Personal Information
• Nat Geo Home
• Attend a Live Event
• Book a Trip
• Inspire Your Kids
• Shop Nat Geo
• Visit the D.C. Museum
• Learn About Our Impact
• Support Our Mission
• Advertise With Us
• Customer Service
• Renew Subscription
• Manage Your Subscription
• Work at Nat Geo
• Sign Up for Our Newsletters
• Contribute to Protect the Planet

Copyright © 1996-2015 National Geographic Society Copyright © 2015-2024 National Geographic Partners, LLC. All rights reserved

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• AIMS Public Health
• v.8(1); 2021

Risk factors associated with stress, anxiety, and depression among university undergraduate students

Mohammad mofatteh.

1 Lincoln College, University of Oxford, Turl Street, Oxford OX1 3DR, United Kingdom

2 Merton College, University of Oxford, Merton Street, Oxford OX1 4DJ, United Kingdom

3 Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, United Kingdom

It is well-known that prevalence of stress, anxiety, and depression is high among university undergraduate students in developed and developing countries. Students entering university are from different socioeconomic background, which can bring a variety of mental health risk factors. The aim of this review was to investigate present literatures to identify risk factors associated with stress, anxiety, and depression among university undergraduate students in developed and developing countries. I identified and critically evaluated forty-one articles about risk factors associated with mental health of undergraduate university students in developed and developing countries from 2000 to 2020 according to the inclusion criteria. Selected papers were analyzed for risk factor themes. Six different themes of risk factors were identified: psychological, academic, biological, lifestyle, social and financial. Different risk factor groups can have different degree of impact on students' stress, anxiety, and depression. Each theme of risk factor was further divided into multiple subthemes. Risk factors associated with stress, depression and anxiety among university students should be identified early in university to provide them with additional mental health support and prevent exacerbation of risk factors.

1. Introduction

Mental health is one of the most significant determinants of life quality and satisfaction. Poor mental health is a complex and common psychological problem among university undergraduate students in developed and developing countries [1] . Different psychological and psychiatric studies conducted in multiple developed and developing countries across the past decades have shown that prevalence of stress, anxiety, and depression (SAD) is higher among university students compared with the general population [2] – [4] . It is well established that as a multi-factorial problem, SAD cause personal, health, societal, and occupational issues [5] which can directly influence and be influenced by the quality of life. The level of stress cited in self-reported examinations and surveys is inversely correlated with life quality and well-being [6] .

Untreated poor mental health can cause distress among students and, hence, negatively influence their quality of lives and academic performance, including, but not limited to, lower academic integrity, alcohol and substance abuse as well as a reduced empathetic behaviour, relationship instability, lack of self-confidence, and suicidal thoughts [7] – [9] .

A 21-item self-evaluating questionnaire, Beck Depression Inventory (BDI), is the most common tool used for diagnoses of depression [10] . A BDI-based survey in five developed countries in Europe (European Outcome of Depression International Network-ODIN in the United Kingdom, Netherlands, Greece, Norway, and Spain) concluded that overall 8.6% (95% CI, 7.95–10.37) of the resident population are dealing with depression [11] . Similar studies confirmed that about 8% of the population in developed and developing countries suffer from depression [12] . Data from systematic review studies revealed that this depression rate is much higher among university students and around one third of all students in the majority of the developed countries have some degree of SAD disorders; and depression prevalence has been increasing in academic environments over the past few decades [3] .

Despite all the efforts to increase awareness and tackle mental health problems among university students, there is still an increasing number of depression and suicide among students [13] , indicating a lack of effectiveness of the measures adopted. In addition to an increase in the prevalence of mental health issues, comparing students and non-college-attending peers demonstrated that the severity of psychological disorders that students receive treatment for has also increased [14] . For example, the rate of suicide among adolescents has increased significantly over the past few decades [15] . In fact, suicide as a result of untreated mental health is the second cause of death among American college students [16] , emphasizing the importance of identifying and treating risk factors associated with SAD.

SAD can be manifested in different forms; however, some common overt symptoms include loss of appetite, sleep disturbance, lack of concentration, apathy (lack of enthusiasm and concern), and poor hygiene. Studying SAD is particularly important among university students who are future representatives and leaders of a country. Furthermore, most undergraduate students enter university at an early age; and dealing with SAD early in life can have long-term negative consequences on the mental and social life of students [3] . For example, a longitudinal study in New Zealand over 25 years demonstrated that depression among people aged 16–21 could increase their unemployment and welfare-dependence in long-term [17] .

A better understanding of SAD among students in developed and developing countries not only helps governments, universities, families, and healthcare agencies to identify risk factors associated with mental health problems in order to minimise such risk factors, but also provides them with an opportunity to study how these factors have been changing in the academia.

This review aims to provide an updated understanding of risk factors associated with SAD among post-secondary undergraduate and college students in developed and developing countries by using existing literature resources available to answer the following question:

“Aetiology of depression and anxiety: What are risk factors associated with stress, anxiety and depression among university and college undergraduate students studying in developed and developing countries?”

It is worth mentioning that this review focuses on SAD risk factors of university students in developed and developed countries, and does not cover underdeveloped countries which can have their own niche problems (such as poverty). However, this review takes into account international students who migrate from underdeveloped countries to developed and developing countries to pursue their education.

2.1. Aims and objectives

The aims of this review were to identifying principal themes associated with depression and anxiety risk factors among university undergraduate students. The objectives of this review are to design a rigorous searching methodology approach by using appropriate inclusion and exclusion criteria, to conduct literature searches of publicly available databases using the designed methodology approach, to investigated collected literature resources to identify risk factors associated with the depression and anxiety which have not changed, and to identify principal themes associated with SAD risk factors among university undergraduate students.

2.2. Designed approach for literature review

A narrative review based on a comprehensive and replicable search strategy is used in this review. This approach is justified and preferred, over other approaches such as primary data gathering, because of the timescale of the research (2000 to 2020-temporal reasons), and extent of the research (developed and developing countries-spatial reasons).

2.3. Criteria for inclusion and exclusion of articles

Inclusion and exclusion criteria for articles and academic writings used in this review are as follows:

2.3.1. Date

2000 to 2020 are included Academic writings which are published between in this review. Initially, during a pilot search, search strategies covered 1990 to 2020. However, the majority of the search results (more than 80% of the search results and more than 88% of applicable search results) were from 2000 to 2020, which indicates the importance of mental health issue and increased awareness over the past two decades. Therefore, for the final search, papers from 2000 to 2020 were included.

2.3.2. Study design

Literatures included in this narrative review were primary research articles, review articles, systematic reviews, mini-reviews, opinion pieces, correspondence, clinical trials, and cases reports published in peer reviewed journals.

2.3.3. Country

The narrative review was limited to developed and developing countries definition by the United Nations Department of Economic and Social Affairs [18] . Abstract and method sections of search results were screened to check the country of research.

2.3.4. Language

Peer-reviewed articles published in English were only included in this narrative review.

2.3.5. The explanation for papers exclusion

The main reason for papers excluded from consideration after search results was that they focused on intervention and therapies associated with SAD. Other reasons for exclusion was that studies were conducted on a mixture of undergraduate and graduate students or focused solely only graduate students. Studies which focused on other types of mental disorders such as eating disorders but did not focus on SAD were excluded too. The conducted search did not exclude any gender or specific age category.

2.4. Strategies used for search and limitations

In this review, a robust and replicable search strategy was designed to identify appropriate articles by searching PubMed, MEDLINE via Ovid, and JSTOR electronic databases. These databases were selected because they encompass biopsychosocial papers published on SAD. The date chosen for this search was for articles published between 2000 to 2020 which covers the past two decades. Once key articles were identified, a search for citation of those papers was conducted, and the bibliography of those papers were further screened to identify potential articles which can be relevant.

2.5. Search terminologies used

To conduct searches in databases mentioned above, the following search terms were used: students stress, anxiety, depression risk factors, university stress, anxiety, depression risk factors, student mental health developed and developing countries, students stress, anxiety and depression developed and developing countries. The operation AND was used to connect stress, anxiety, depression, mental health, developed, developing, countries, students. The search for each term was conducted in all fields (title, abstract, full text, etc.).

2.6. Screening, selecting search results, and data extraction

The search results were exported into separate Excel and EndNote X8 files. Titles and abstracts from all articles were screened to determine their relevance to the topic of this review. Potentially relevant articles were fully read to establish their relevance. Each paper which was included according to the inclusion criteria described above was read fully. A word file was created to identify themes associated with SAD risk factors which is included in the Results. An initial search resulted in 1305 articles. The title and abstract of individual papers were read for relevance, resulting in 60 papers which were relevant for the research question asked in this review. All 60 papers were read completely, and from those, 19 were excluded based on the criteria mentioned before. Therefore, the total number of papers for consideration was 41. A flowchart explaining the procedure for identification, screening, eligibility, and inclusion of papers is shown in Figure 1 .

Figure 2 provides a quantitative summary of the papers included in this narrative review. In terms of the distribution of the countries where the research was conducted, included papers were mainly articles which carried out studies in the USA (n = 17), followed by China and Canada (each n = 5), UK (n = 4), Japan (n = 3), Germany and Australia (each n = 2), South Korea, Hungary, Switzerland (each n =1) ( Figure 2A ). As for article types included in this review, original research articles, including quantitative and qualitative studies, which relied on obtaining data including cross-sectional studies, interviews, case-control studies, surveys, and questionnaire, were the highest (n = 37) followed by meta-analysis, literature and systematic reviews ( Figure 2B ). Another interesting observation was that although the search was carried out from 2000–2020, most papers were concentrated in the period from 2016 to 2020 ( Figure 2C ). This can be due to the reason that mental health is becoming more important over the past few years. Alternatively, a higher number of papers included from 2016 onward can be due to unintended selection bias. The smallest study covered in this narrative review was conducted on 19 students and the largest one on 153,635 students, adding up to 236,104 students, who were included in articles covered in this narrative review in total. Most studies on mental health, anxiety, and depression use standardised approaches such as patient-filled general health questionnaires, Pearling coping questionnaire, internally regulated surveys, BDI, DSM-IV symptomology, and general anxiety and burnout scales such as Maslach Burnout Inventory.

3.1. Literature search results

Following the search protocol shown in Figure 3 , a list of included papers identified which can be found in the Table 1 .

3.2. Prevalence of mental health disorders in students

Literature showed that mental health problems are common phenomenon among students with a higher prevalence compared to the general public. For example, surveying more than 2800 students in five American large public universities demonstrated that more than half of them experienced anxiety and depression in their last year of studies [19] . Similarly, a survey of Coventry University undergraduate students in the UK showed that more than one-third of them had experienced mental health issues such as anxiety and depression over the past one year since they were surveyed [20] . In agreements with these results, Maser et al. [21] found that prevalence of mental health disorders including anxiety and depression was higher among medical students compared to the general non-student population of the same age. These studies demonstrated that the prevalence of SAD among students has remained higher than the average population over the past two decades.

SAD are not only prevalent among students, but also persistent. By conducting a follow-up survey study of students over two years, Zivin et al. [19] demonstrated that more than half of students retain their higher levels of anxiety and depression over time. This can be due to a lack of SAD treatment or persistence of existing risk factors over time.

3.3. Risk factors associated with stress, anxiety, and depression

SAD are multifactorial, complex psychological issues which can have underlying biopsychosocial reasons. Multiple risk factors which affect the formation of SAD among undergraduate university students in developed and developing countries were identified in this review. These factors can be categorized into multiple themes including psychological, academic, biological, lifestyle, social and financial. A summary of risk factors and their associated publications are shown in Table 2 .

3.3.1. Psychological factors

Self-esteem, self-confidence, personality types, and loneliness can be associated with SAD among university students. Students who have a lower level of self-esteem are more susceptible to develop anxiety and depression [22] . Also, students with high neuroticism and low extraversion in five-factor personality inventory [23] are more likely to develop SAD during university years [24] . Other psychological factors such as feeling of loneliness plays important roles in increasing SAD risk factors [24] . Moving away from family and beginning an independent life can pose challenges for fresher students such as loneliness until they adjust to university life and expand their social network. Indeed, Kawase et al. [24] showed that students who live in other cities than their hometown for studying purposes are more likely to develop anxiety and depression.

Some students enter the university with underlying mental conditions, which can become exacerbated as they transition into the independent life at university. While depression is higher among university and college students compared to the general public, students with a history of mental health problems, such as post-traumatic stress disorder (PTSD), are more prone to development of anxiety and depression during their university lives compared to students who did not have such experience before starting their degrees [25] . Furthermore, exposure to violence in childhood either at the household or the community correlates with SAD formation later in life and at University [26] . Therefore, low self-esteem and self-confidence, having an underlying mental health condition before beginning the university, personality type (high neuroticism and low extravasation), and loneliness can increase the probability of SAD formation in students.

3.3.2. Academic factors

Multiple university-related academic stressors can lead to SAD among students. One of these factors which was strongly present in many studies evaluated in this review was the subject of the degree. Medical, nursing, and health-related students have a higher prevalence of depression and anxiety compared to their non-medical peers [24] , [27] – [28] . Medical and nursing students who have both theoretical duties and patient-related work usually have the highest level of workload among university students, consequently deal more with anxiety and depression [27] , [29] . In addition, students who major in psychology and philosophy, similar to nursing and medical students, are more likely to develop depression during their studies compared to others [24] . These studies did not identify whether students who have underlying mental health conditions are more likely to choose certain subjects such as philosophy, psychology, or subjects which lead to caring roles such as nursing and medicine. Because of the nature of their work, medical and nursing students who deal with people's health can experience depression and anxiety as a result of fears of making mistakes which can result in harming patients [27] . Students with practical components in their degree are required to travel to unfamiliar places for fieldwork and work experience which can add to their stress and anxiety [27] .

Also, some prospective students, especially those who study nursing and medicine, usually do not have a clear understanding of the curriculum and workload associated with the subject before entering the university, therefore, they can face a state of disillusionment once they begin their studies at university [27] – [29] . It is worth mentioning that not all studies found a significant correlation between the subject of study and SAD development [30] . This can be explained by differences in sample type and size which results in variations existing in the amount of workload and curriculum in similar subjects taught in various universities in different countries.

Studying a higher degree can be a challenging task which requires mental effort. Mastery of the subject can negatively correlate with self-esteem, anxiety, and depression among university students with students who have a mastery of subject demonstrating a lower level of stress and anxiety [31] . Also, students who study in a non-native language report the highest level of anxiety and depression during their freshman years, and their stress levels decrease during the subsequent study years [32] . This can be explained by the fact that students who are studying in a foreign language usually are those who have migrated abroad, therefore, require some time to adjust to their new lives. Different studies showed that the level of anxiety and depression among both international and home students could correlate with the year of study with fresher students who enter the university and students at the final year of their studies experience the greatest amount of anxiety and depression with different risk factors [22] , [32] . While fresher students experience SAD because of challenges in adjustments to university life, past negative family experience, social isolation and not having many friends, final year students report uncertainty about their future, prospective employment, university debt repayment and adjusting to the life after university as major risk factors for their SAD [22] , [32] . Therefore, a shift in SAD risk factors themes are observed as students make a progress in their degrees.

Students spend a significant portion of their time at university being engaged with their academic activities, and unpleasant academic outcomes can influence their mental health. Receiving lower grades during the time of studies can negatively influence students' mental health, causing them to develop SAD [33] , [34] . Academic performance during undergraduate studies can determine the degree classification, which can, subsequently, influence students' opportunities such as employment success rate or access to postgraduate courses [27] . Conversely, both the number of students with mental health problem symptoms and the severity of students' SAD increase during exam time [35] , reflecting a direct relationship between academic pressure and students' mental health states. However, the causal relationship is not well-established; it is possible that depression and associated problems such as temporary memory loss and lack of concentration [36] are reasons for poor academic grades or inversely, students feel stressed leading to depression because of their poor performance in their exams. A mutual relationship can exist between grades and mental health, as having a poor mental health can reciprocally cause students to get lower grades [34] , leading to a vicious cycle of mental health and academic performance. Interestingly, students' sense of social belonging and coherence to the university community was reduced during exam periods [35] . This can be explained by the reduced participation rate of students in university social activities and clubs as well as an increased sense of competition with their peers. Furthermore, students interact directly and indirectly with teachers, lecturers, tutors, and other staff; therefore, the relationship between students and academic staff can influence students' mental health. A negative and abusive relationship with teachers and mentors can be another factor causing SAD among undergraduate students [27] .

On the other hand, being a part-time student is a protective factor for anxiety and depression, and part-time students have better mental health compared to students with full-time status [34] . This can be explained by financial securities which have a source of income can bring or because part-time students are usually older than full-time students [34] , and therefore, more emotionally stable. In conclusion, risk factors increasing SAD among university students include high workload pressure, fear of poor performance in exams and assessments, wrong expectations from the course and university, insufficient mastery in the subject, year of study, and a negative relationship with academic staff.

3.3.3. Biological factors

Mental health can be influenced by ones' physical health. Presence of an underlying health condition or a chronic disease before entering the university can be a predictor of having SAD during university years [31] , [33] . Students with physical and mental disabilities can be in a more disadvantaged position and do not fully participate in university life leading to SAD formation [33] .

An association between gender and depressive disorders have been observed in several studies [21] , [27] , [34] , [37] . Female students had a higher prevalence of SAD compared to male students. Interestingly, while female students demonstrated a higher level of SAD, the dropout rate of female students with a mental health problem from university was lower compared to their male counterparts [33] . On the other hand, while females are at a higher risk of developing depressive disorders, males with depressive disorders are less willing to seek professional help and ask for support due to the stigma attached to mental health [38] , causing exacerbation of their problem over time [20] .

Age can be another factor related to SAD. Younger students report a higher level of SAD compared to older students [34] , [37] . However, other meta-analysis studies did not find a significant correlation between students' age and their mental health which can be due to sampling differences [39] . Some studies showed that while older undergraduate students have a higher determination to do well in the university [40] , those who have family commitments are more prone to develop SAD during their degrees [27] . These discrepancies in findings can be explained by different sample sizes and types of studies which can be influenced by various confounding factors such as nationality, country of study, degree of studies, gender, and socioeconomic status. Similarly, a lack of correlation between depression prevalence and year of study is observed as some studies have reported a higher prevalence among earlier years of studies, while others have shown a higher prevalence among students as they move closer to the end of their studies [41] . These differences can be explained by different causes of depression in a different age; for example, while depression in younger adults can be due to changes in their environment and difficulties in adapting to a new life, older adults can have depression symptoms because of a lack of certainty for their future and employment. Nevertheless, differences exist between SAD risk factors associated with young and older students. Overall, biological risk factors affecting SAD include age of students, gender, and underlying physical conditions before entering the university.

3.3.4. Lifestyle factors

Moving away from families and beginning a new life requires flexibility and adaptation to adjust to a new lifestyle. As most undergraduate students leave their family environment and enter a new life with their peers, friends, and classmates, their behaviour and lifestyle change too. Multiple lifestyle factors such as alcohol consumption, tobacco smoking, dietary habits, exercise, and drug abuse can affect SAD. Alcohol consumption is high among students with SAD [28] ; a causal relationship was not been established in this study though.

Tobacco smoking is another risk factor associated with SAD which is common among students, especially students who study in Eastern developed and developing countries such as China, Japan and South Korea [24] , [42] . Most students, especially male students, smoke because of social bonding and the rate of social smoking is directly correlated with SAD [24] , [42] . Social smokers are less willing to quit smoking, and more likely to persist in their habit, resulting in long term negative physical and psychological health consequences [42] . Illegal substance abuse can be another factor important in SAD among young people [43] . Academic-related stress and social environment in university dormitories and student accommodations can encourage students to use illegal drugs, smoke tobacco and consume alcohol excessively as a coping mechanism, resulting in SAD [44] . Interestingly, students who perceived they had support from the university were feeling less stressed and were less at the risk of substance abuse [45] , indicating the important role of social support in preventing and alleviating depression symptoms. This is of particular importance as a new social habit and behaviour adapted early during life can last for a long time. Furthermore, students who do not have a healthy lifestyle can feel guilt, which can worsen their SAD condition [46] . Interestingly, Rosenthal et al. [47] showed that negative behaviours resulting from alcohol consumption such as missing the next day class, careless behaviour and self-harm, verbal argument or physical fight, being involved in unwanted sexual behaviour, and personal regret and shame could be the main reasons for depression associated with drinking alcohol, rather than the amount of alcohol consumed.

In contrast, a moderate to vigorous level of physical activity can be a protective factor against developing SAD during university life [37] , [48] . Students who have a perception of having inadequate time during their studies do not spend enough time for exercise and can develop SAD symptoms [27] .

Another lifestyle-related risk fact associated with SAD is sleep. Many young people do not receive sufficient sleep, and sleep deprivation is a serious risk factor for low mood and depression [28] , [47] . Self-reported high level of stress and sleep deprivation is common among American students [31] , [49] . Insufficient sleep can act as a vicious cycle- academic stress can cause sleep deprivation, and insufficient sleep can cause stress due to poor academic performance since both sleep quality and quantity is associated with academic performance [28] . Overall, poor sleeping habit is associated with a decreased learning ability, increase in anxiety and stress, leading to depression.

Different negative lifestyle behaviours such as tobacco smoking, excessive alcohol consumption, unhealthy diet, lack of adequate physical activity, and insufficient sleep can increase the risk of SAD formation among university students.

3.3.5. Social factors

Having a supportive social network can influence students' social and emotional wellbeing, and subsequently lower their probability of having anxiety and depression in university [27] , [37] , [50] . The quality of relationship with family and friends is important in developing SAD. Having a well-established and supportive relationship with family members can be a protective factor against SAD development, which, in turn, can affect the sense of students' fulfilment from their university life [27] . The frequency of family visits during university years negatively correlates with SAD development [33] . Family visits can be more challenging for international students who live far away from their families, therefore adding to existing problems of international students who live and study abroad.

In contrast, having a negative relationship with family members, especially parents, can cause SAD formation among students in university [51] . Similarly, having a strict family who posed restrictions on behaviours and activities during childhood can be a predictor of developing SAD during university years [51] .

Also, it is shown that being in a committed relationship has a beneficial protective factor against developing depressive symptoms in female, but not male, students [52] . Interestingly, both male and female students who were in committed relationships reported a lower alcohol consumption compared to their peers who were not in committed relationships [52] .

Involvement in social events such as participating in sporting events and engaging in club activities can be a protective factor for mental health [32] , [37] . Assessing preclinical medical students' social, mental, and psychological wellbeing showed that while first year students demonstrate a decrease in their mental wellbeing during the academic year, they have an increase in their social wellbeing and social integration [53] . This can be explained by the time period required for fresher students who enter the university to adjust to the social environment, make new friends, and integrate into the social life of the university.

Access to social support from university is another factor which is negatively correlated with developing anxiety and depression [31] . It is worth mentioning that different universities provide different degrees of social support for students which can reflect on different anxiety and depression observed among students of different universities.

Importantly, sexual victimization during university life can be a predictor of depression. By surveying female Canadian undergraduate students, McDougall et al. [54] found that students who were sexually victimized and had non-consensual sex were at a higher chance of developing depression following their experience, emphasizing the importance of safeguarding mechanism for students at university campuses.

While the internet and social media can be great tools for maintaining a social relationship with classmates, pre-university friends and family members, it can have negative mental health effects. Excessive usage of social media and the internet during freshman year can be a predictor of developing SAD during the following years [55] . Students who have a higher dependence on the social media report a higher feeling of loneliness, which can result in SAD [56] . Students with internet addiction and excessive usage of social media are usually in first year of their degrees [55] , [56] which can reflect a lack of adjustment to university life and forming a social network. Also, students who use social media more often have a lower level of self-esteem and prefer to recreate their sense of self [56] , indicating an intertwined relationship between biopsychosocial factors in developing SAD among students.

Demographic status, ethnic and sexual minority groups including international Asian students, black and bisexual students were at an elevated risk of depression and suicidal behaviour [16] , [50] . The frequency of mental health is usually more common among ethnic minorities. For example, Turner et al. [20] showed that ethnic minority students report a higher level of anxiety and depression compared to their white peers; however, they do not ask for help as much. Other studies supported these findings by showing that students from ethnic and religious minorities, regardless of their country of origin and country in which they study, have a higher prevalence of anxiety and depression compared to their peers [50] . Also, students' expectations from university can be different among ethnic minorities students, and most of them do not have a sufficient understanding of the services that university can provide for them [40] .

Therefore, lack of support from family and university, adverse relationships with family, lack of engagement in social activities, sexual victimization, excessive social media usage, belonging to ethnic and religious minority groups, and stigma associated with the mental health are among risk factors for SAD in university students.

3.3.6. Economic factors

Students' family economic status can influence their mental health. A low family income and experiencing poverty can be predictors of SAD development during university years [22] , [50] , [57] , [58] . A higher family income can even ameliorate negative psychological experiences during childhood, which can have long-term negative consequences on the mental health of students once they enter university [57] . Also, experiencing poverty during childhood can have negative long-term consequences on adults, leading to SAD development during university life [58] .

Some students take up part-time job to partially fund their studies. Vaughn et al. [59] showed that relationship of employed students with their colleagues in the workplace could affect students' mental health; and those students who had a poor relationship with their colleagues had worse mental health. However, it is worth mentioning that a causal relationship was not established. It can be possible that students who have poor mental health cannot get along with their co-workers, resulting in an adverse working relationship.

Because of paying higher tuition fees and less access to scholarships and bursaries available, international students can have more financial problems, causing a higher degree of anxiety and depression compared to home students [60] .

Lack of adequate financial support, low family income and poverty during childhood are risk factors of SAD in students of undergraduate courses in developed and developing countries.

3.4. Stigma associated with mental health

While efforts have been put to reduce the stigma associated with receiving help for mental health problems, this still remains a challenge. For example, more than half of students who had SAD did not receive any help or treatment for their condition because of the stigma associated with mental health [19] , [61] . This is not related to the awareness of the availability of mental health resources which was ruled out by authors, as most of the students who did not receive any help for their mental health problem were aware of available help and support to them [19] .

Furthermore, the social stigma associated with receiving help for mental health problems was significantly associated with suicidal behaviour, acting as a preventive barrier to seek help (planning and attempt) [16] . Among students, those with a history of mental health problem such as veterans with PTSD are less likely to seek for help compared to non-veteran students [25] , making them more susceptible to struggling with untreated mental health.

4. Discussion

This review tried to identify and summarise risk factors associated with SAD in undergraduate students studying in developed and developing countries. The prevalence of SAD is high among undergraduate university students who study in developed and developing countries. Untreated SAD can lead to eating disorders, self-harm, suicide, social problems [28] . Similar to a complex society, differences exist among students leading to complicated risk factors causing SAD. Because different themes influencing SAD has been investigated as a distinct body of research by different literature, a concept map is created to demonstrate the relationship between various risk factors contributing to the development of SAD in undergraduate students in developed and developing countries. Figure 3 bridges risk factors concepts between different literature. For most students, entering university is a new step in their lives which is associated with certain challenges such as moving into independent accommodation, social identity, financial management, making decisions, and forming a social network. Different students have different needs depending on the stage of their degree, which needs to be fulfilled. For example, coping with a new university life style can be a challenging task for students who enter the university. This becomes more significant for students moving abroad for their studying who need to adapt to a new lifestyle, speak in a different language, and live away from their families. In agreement with this, different levels of anxiety and depression with different risk factors are observed among students as they progress in their degrees. On the other hand, students who are finishing their degrees can have SAD because of uncertainties about their future.

Students learn different modules in different degrees and have different abilities. Mastery of the subject can be a factor affecting students' sense of self-esteem, influencing their anxiety level and developing depressive symptoms. This partially can explain changes in risk factors observed as students' progress in their degrees. Final year students who adjust to the university environment and develop mastery in their subject can deal with academic pressure better compared to freshers who transform from secondary school life to university lifestyle.

Students can come with a varied and challenging background such as those who experienced household and domestic violence, sexual abuse, and child poverty which can make them susceptible to developing anxiety and depression once academic pressure is mounted. As universities are diverse environment which enrol students from different socioeconomic background and different cultures, universities need to identify risk factors for different students and have robust plans to tackle them to provide a fostering environment for future leaders of the society. Therefore, early mental health screening can help to identify those students who are at risk to provide them with special and additional mental health support. Students not only should be screened for their mental health state as they enter university, but also regular follow up check-ups should be conducted to monitor their conditions as they progress in their degrees to detect early signs of SAD.

University and academic staff can play a significant role in either exaggerating or ameliorating risk factors associated with anxiety and depression. While teachers and mentors can support students to cope with SAD, they can be a source of problem too by discriminating, bullying, and hampering students' progress.

Managing finance and expenses can be a challenging task for students who are stepping into an independent lifestyle and need to pay for their tuitions in addition to their maintenance fees. While some students have access to private funding, bursaries, and scholarships, other students receive loans which they need to pay back or have part-time jobs to meet their expenses. Students who work need to have a work-life balance and the time spent in their jobs can affect the quality of their education.

Fresher students try to establish their social network and might feel isolated, which can push them to excessive usage of social media to fill their social gap. While internet addiction and excessive usage of social media can have a negative impact on students' mental health, technology, such as mobile phone applications can be used in universities campuses to promote a healthier lifestyle and reducing risk factors among students. For example, many students refuse to receive face-to-face mental health counselling support during their anxiety and depression due to stigma associated with disclosure of mental health issues. Providing students with anonymized counselling services through mobile phone applications can be one way of delivering help to students at universities.

With the advent of social media platforms such as Facebook, Twitter, Instagram, TikTok, etc., more and more students rely on such networks for socialisation. While the internet and online platforms can have beneficial consequences for students, such as rapid access to a variety of online learning resources and keeping in contact with friends and families, excessive usage of social media and internet can have negative consequences on students' academic performance. A poor mental health state at the beginning of university life is a predictor of internet addiction later during the degree. Heavy reliance on the internet can be a coping mechanism for students with anxiety and depression to overcome their mental health problems.

As governments and educational bodies in developed and developing countries are emphasising recruitment of ethnic minority students to university to increase the range of equality and diversity among students, it is important to consider the mental health of those students in the university as well. Students in minority groups such as black, international Chinese and bisexual student report a higher level of anxiety and depression compared to other non-minority group students. This can be due to either pre-existing conditions which student experience before entering the university, and can be exacerbated during the university, or can be because of problems which can develop during university life.

Also, more mental health support is available in universities as the number of university students is increasing, and there is a better understanding of the importance of mental health in academia; however, the stigma associated with mental health has not changed proportionately.

While research and understanding of mental health have changed significantly over the past two decades and many more articles are present, risk factors associated with SAD remain unchanged.

One caveat with studies of mental health among student is that most studies have been conducted among medical and nursing students and neglected non-medical students. One potential explanation for the tendency to conduct depression surveys among medical students is the higher response rate as medical students are more willing to fill out the questionnaires and surveys. It is understandable that students studying medical subjects, who directly interact with the public and treating them once they enter the healthcare profession should have a reasonably sound mental health to be able to conduct their duties, but it does not justify neglecting the mental health of other students. Therefore, more research on mental health and risk factors associated with SAD of non-medical students is required in the future.

Another caveat with most mental health studies is that they are based on self-reports and surveys. Different people can have different perception and understanding of mental health and anxiety, and many confounding factors can influence the response of participants in the time of participation. Furthermore, students with severe mental health conditions are less likely to participate in any activity including surveys and questionnaires, leading to a non-response bias.

Another area which requires improvement in future studies of mental health is the categorisation of different types of depression and their severity. Depression and anxiety are a spectrum which can comprise of minor and major symptoms; however, most studies did not specify the scale of depression in their findings. Furthermore, while various risk factors were identified, a causal relationship between mental health and behaviours were not established.

While counselling services provided by universities in Western countries such as the UK and USA have increased over the past few years [62] , it is still not clear how effective such services are; therefore, more research is required to assess the effectiveness of counselling services at universities.

Therefore, a better understanding of the aetiology, associated factors is required for an effective intervention to reduce the disease incidence and prevalence among students in the population and providing them with a fostering environment to achieve their potential.

University undergraduate students are at a higher risk of developing SAD in developed and developing countries. Promoting the mental health of students is an important issue which should be addressed in the education and healthcare systems of developed and developing countries. Since students entering university are from different socioeconomic background, screening should be carried out early as students.

A personalized approach is required to assess mental health of different students. In addition, a majority of mental health risk factors can be related to the academic environment. A personalised, student-centred approach to include needs and requirements of different students from different background can help students to foster their talent to reach their full potential. Furthermore, more training should be provided for teaching and university staff to help students identify risk factors, and provide appropriate treatment.

5. Conclusion

Despite all the efforts over the past two decades to destigmatise mental health, the stigma associated with mental health is still a significant barrier for students, especially male students and students from ethnic and religious minorities to seek help for SAD treatment. Universities need to continue to destigmatise mental health in university campuses to enable students to receive more in campus support by providing designated time for positive metal health activities such as group exercise, physical activities, and counselling services. There is no shortage of athletic and group activities in form of clubs and social classes in most universities in developed and developing countries; however, more incentives such as athletic bursaries and prizes should be provided to students to encourage their participation in such activities which can act as protective factors against SAD development. Therefore, universities need to allocate more resources for sporting and social activities which can impact the mental health of students. Furthermore, an increase in mental health problems in universities has created a huge burden on university counselling services to meet the demands of students. More novel approaches, such as online counselling services can help universities to meet those increased demands.

Students in different years of studies deal with different risk factors from the time that they enter the university until they graduate, therefore, different coping strategies are required for students at different levels. Universities should be aware of these risk factors and implement measures to minimise those factors while providing mental health treatments to students.

Future studies are required to investigate long-term effects of experiencing SAD on students. A longitudinal study with a large randomly recruited sample size (different age, sex, degree of study, – socioeconomic status, etc.) is required to address how students' mental health change from entering the university until they graduate. Also, more extended follow up studies can be included to address the effect of depression and poor mental health on people's lives after they graduate from the university.

Abbreviations

Conflict of interest: All authors declare no conflicts of interest in this paper.