key: cord-0692338-d5ee12na
authors: Hopf, D.; Schneider, E.; Aguilar-Raab, C.; Scheele, D.; Ditzen, B.; Eckstein, M.
title: Loneliness and diurnal cortisol levels during COVID-19 lockdown: the roles of living situation, relationship status and relationship quality
date: 2022-02-26
journal: nan
DOI: 10.1101/2022.02.25.22271461
sha: 80d20290ae7def8223c3c0ab3f2b2b3a70d41c12
doc_id: 692338
cord_uid: d5ee12na

Loneliness and social isolation have become increasing concerns during COVID-19 lockdown through neuroendocrine stress-reactions, physical and mental health problems. We investigated living situation, relationship status and quality as potential moderators for trait and state loneliness and salivary cortisol levels (hormonal stress-responses) in healthy adults during the first lockdown in Germany. N=1242 participants (mean age = 36.32, 78% female) filled out an online questionnaire on demographics, trait loneliness and relationship quality. Next, N=247 (mean age = 32.6, 70% female) completed ecological momentary assessment (EMA), collecting twelve saliva samples on two days and simultaneously reporting their momentary loneliness levels. Divorced/widowed showed highest trait loneliness, followed by singles and partnerships. The latter displayed lower momentary loneliness and cortisol levels compared to singles. Relationship satisfaction significantly reduced loneliness levels in participants with a partner and those who were living apart from their partner reported loneliness levels similar to singles living alone. Living alone was associated with lower loneliness levels. Hierarchical linear models revealed a significant cross-level interaction between relationship status and momentary loneliness in predicting cortisol. The results imply that widowhood, being single, living alone and low relationship quality represent risk factors for loneliness and having a partner buffers neuroendocrine stress responses during lockdown.

The recent Corona virus pandemic has been occupying mental and physical health facilities for two years now. Hard lockdown regulations in almost all countries early during the pandemic (April until June 2020) to prevent further spreading of the virus entail increased social isolation. The steady and massive health threat from the virus in combination with the missing social buffering effect of everyday social encounters lead to or amplified psychosocial problems that could have long-term consequences for mental and physical health [1] [2] [3] [4] . E.g., loneliness, as the subjective and emotional component of social exclusion, is a highly topical and public health issue in modern societies, where social isolation and anonymity become increasingly prevalent 5, 6 . It entails the perceived lack of intimacy or social companionship and the feeling that social relationships are deficient in either quality or quantity 7 . By contrast, social isolation is defined as the objective state of being alone 7, 8 .

According to the belongingness-hypothesis, loneliness is rooted in the human need to socially belong, or the pervasive drive to form and maintain lasting positive and significant social relationships 9 . It has been shown that the sense of belonging in early adolescents is mainly achieved through the acceptance by peers, whereas in late adolescence and adulthood, it is achieved especially by romantic relationships, marital status and close friends 10 . On the other hand, lacking feelings of belonging are assumed to be associated with loneliness and negative physical and mental health outcomes in a long-term 9 . Both loneliness and social isolation are significantly associated with indices of physical and mental health, such as psychosocial stress 11 , depression 12 , generalized anxiety 6 , cardiovascular diseases 13 , chronic obstructive pulmonary disease 14 , and mortality 8, [15] [16] [17] [18] [19] . Chronic loneliness may hamper the formation of new social relationships by inducing negative cognitive biases such as interpersonal distrust 20 .

Furthermore, loneliness is associated with neuroendocrine parameters, like elevated cortisol levels [21] [22] [23] and altered cortisol awakening responses 23, 24 . As one of the main effector hormones quality might become an important moderator, especially if couples do not live together and thus are unable to see their partner and potentially have to rely on non-physical relationship qualities. Living alone has become increasingly prevalent, with one-person households accounting for more than 40% of all households in Scandinavian nations, more than 33% of all households in France, Germany, and England; and more than 25% of all households in the United States, Russia, Canada, Spain, and Japan 54 . In Germany, in the young adult age of 18 to 30 years, more than 30% live without a partner 55 . An important distinction in this context is between partnerships with and without a common household (the latter being called "living apart together"). In general, living alone has been seen as a risk factor for poor physical and mental health 54, 56 . For instance, the living situation predicts mortality risk 57, 58 and people who are living alone show higher loneliness levels 59 .Cross-sectional studies suggest that during the pandemic, being married served as a protective factor against loneliness 60 , whereas being divorced or widowed increased the risk of loneliness 61 . Furthermore, living with others has been found to protect against loneliness 62 , even when controlling for relationship status 63 and loneliness during lockdown predicted psychological distress 64 . It has not been investigated yet, however, whether relationship status and living situation during lockdown affected biological, specifically neuroendocrine, health parameters, such as cortisol levels. In previous studies, living alone had been positively associated with cortisol levels 65 . Furthermore, the buffering effect of living situation and relationship status with regard to psychobiological outcomes during stress-exposure (i.e. the world-wide considerable psychological stress through COVID- 19) has not been investigated yet. Previous research suggests that the separation from a partner is associated with elevated feelings of loneliness and cortisol levels in general [66] [67] [68] . In adolescents, significant correlations between self-reported loneliness and cortisol awakening responses during COVID-19 lockdown were found 69 . However, moment-to-moment associations of loneliness and cortisol have not been investigated in adults yet. Furthermore, it is still elusive if relationship status and living situation moderate these associations. Lastly, the . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Of the participants in the online survey, 472 showed interest in the EMA with the salivary sampling. Of those 472 participants, 54% (n = 257) took part in the EMA study. After excluding individuals who did not react to our messages and dropouts during data collection (n = 10), the remaining 247 cases were included in the analyses. The participants' mean age was M = 32.6 . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

To measure trait loneliness in the online survey, we employed the German version of the revised 20-item University of California at Los Angeles (UCLA) loneliness scale 70, 71 . Within our study, the scale displayed high internal consistency (Cronbach's α = .91). Participants are asked to answer, how often they felt a certain way during the past two weeks, on a 4-point Likert scale with higher scores indicating more loneliness. Exemplary items are 'I feel isolated from others.' or 'I do not feel alone.' (negatively scored item). In order to assess momentary levels of loneliness in the EMA study, we used a single item measure ("Do you feel lonely at the moment?") with a visual analogue scale (VAS; 0 -not at all, to 100very lonely).

Saliva samples for determination of cortisol concentrations were collected at the same times as EMA. Sampling times were adapted to the individual wake-up time. Samples were taken at six time-points on two consecutive days: directly after awakening, 30 min, 45 min, 2 ½ hours and 8 hours after awakening and immediately before going to sleep. Participants stored the samples in their freezer until collected on dry ice and stored at -80°C until analysis. Analyses were conducted in the biochemical laboratory at Heidelberg University Hospital's Institute of Medical Psychology using commercial enzyme-linked immunosorbent assay (ELISA, Demeditec Diagnostics, Germany) procedures with reported detection limit of 0.019 ng/ml. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

Relationship quality was assessed via the short version of the Partnerschaftsfragebogen (PFB) 72 . It consists of 9 items that can be answered on a 4-point Likert scale. In our sample, the internal consistency of the PFB was very good (Cronbach's α = .85). We used the global PFB score by adding up all items.

As age and sex have been previously shown to influence loneliness during the lockdown 73 , they were included as covariates into the calculations. For the EMA study, control variables (CVs) were assessed on both the momentary level (in case the outcome was cortisol) and the trait level (for both cortisol and loneliness as outcomes). Our decisions on the hormonal CVs were mainly based on expert consensus guidelines 74 . On the momentary level, the following CVs were assessed: sleep duration, sleep quality, sleeping problems, sleep medication, forced awakening, brushing teeth, eating behaviour, drinking behaviour, medication, alcohol consumption, nicotine consumption, caffeine consumption, and physical activity (with respect to the last sample), assessment time-point (1 variable for the rise from time-point 1 to 2, and 1 variable for the fall from time-point 2 to 6), and day (1 vs. 2). Trait level control variables were age, sex, and body mass index (BMI). In order to enable a trade-off between a parsimonious and a sufficiently exhausted model, we decided to include only the significant CVs at level 1.

Significant CVs for cortisol as outcome were: eating, drinking, alcohol consumption, caffeine and physical activity (yes/no). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint

The study was part of a large-scale longitudinal study that aims to investigate long-term consequences of COVID-19 lockdown on psychobiological health. Results within this paper entail data from time-point 1 (first lockdown in Germany). The online survey as well as the EMA were both conducted with the platform soscisurvey.de and participation was completely anonymous. After completing the online survey, participants were asked whether they wanted to take part in the EMA. Those who were interested, were contacted via e-Mail. The responders received Salicap® tubes for saliva collection with additional informational documents via mail and specific instructions via phone. The assessment of the saliva samples took place between April 9 th and June 3 rd 2020. On two consecutive days, the participants received the respective link via SMS to a short online survey including instructions for saliva sampling six times per day. Participants were asked to refrain from food or caffeine before they provided three saliva samples which were stored in the freezer. Then, they were asked to answer further questions about their sleeping behaviour, consumption behaviour, and physical activity. Commitment was constantly monitored online: if the participants have not yet accessed the link 5 minutes after it was sent, they were reminded by phone to do so. After completion of the two sampling days, data were stored on an institute-internal data server and saliva samples remained in the participants' home freezer until collection.

In order to test hypotheses 1 -3, we conducted analyses of covariance (ANCOVA). For hypothesis 1, family status (married/in a romantic relationship vs. single vs. divorced/widowed) served as independent variable (IV) and UCLA loneliness scores as dependent variable (DV).

Post-hoc contrasts coding was conducted in order to analyse the linear trend of the means. For is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. vs. with others) served as IVs. In this step we were interested in overall loneliness and cortisol in every-day life, thus the aggregated momentary loneliness and cortisol levels were used as DVs. As the distribution of the cortisol data was positively skewed, we natural-log-transformed the data in order to normalize their distribution. In case the assumptions of conducting an ANCOVA were violated, we used bootstrapping estimates (n = 1000) in order to achieve more robust results 75 . In order to test pairwise differences in momentary loneliness scores between the living situation and relationship status groups (in case the main effects were significant), we calculated Tukey Honestly Significant Differences (HSD) with p-values adjusted for multiple comparisons. We further calculated partial η² in order to receive the effect sizes, with η² ≥ 0.01 indicating a small, η² ≥ 0.06 a medium, and η² ≥ 0.14 a large effect.

To test hypotheses 4 and 5, we conducted multilevel modelling (MLM) regression analyses, which enabled us to assess the within-and between-person effects of momentary loneliness on momentary cortisol levels. The individual levels of loneliness were centred on the person's mean in order to test the within-person effect on cortisol levels. In order to assess the betweenperson effects, we centred the individuals' mean loneliness levels on the grand mean. For is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

In the following, we will report results from all hypotheses separately. Descriptive statistics of the outcomes of interest are shown in Table 3 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Tukey's HSD test indicated significant differences for the following pairwise comparisons (see is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Association of momentary loneliness, relationship status, and living situation with cortisol levels (Hypotheses 4 and 5)

The Intraclass Correlation Coefficient (ICC) within the empty MLM was .007, indicating that 0.7% of the variance in cortisol levels was accounted by between-person differences and 99.3% by within-person differences. As 22 cases had missing values on level 2 variables, a total of 225 cases and 1722 data points were included in the analyses. The random slopes model (with level 1-loneliness set as random predictor) showed a better fit to the data compared to the random intercepts model, (χ²(2) = 7.52, p = .020), therefore we report results from this model.

There was a non-significant within-person effect of self-reported loneliness on cortisol levels is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

In the subsample of participants who were in a relationship, multiple regression analysis revealed a significant association between relationship quality and self-reported mean state loneliness levels (t(154) = -2.24, ß = -.71, p = .026). Furthermore, participants who were living alone, showed significantly higher state loneliness levels compared to participants who were living with others (t(154) = -3.13, ß = -.24, p = .002). However, the interaction between relationship quality and living situation was not significant (t(154) = 1.41, ß = .44, p = .162),

indicating that relationship quality did not moderate the association between living situation and loneliness.

This study examined the (separate and joint) associations between structural (relationship status and living situation) and psychological factors (relationship quality) and loneliness and cortisol during COVID-19 lockdown.

All in all, our results provide further evidence for the belongingness-hypothesis, showing that romantic relationships, as a source for meaningful interactions and intimacy, as well as living with others protect against loneliness and neuroendocrine stress-responses, in this case diurnal cortisol levels [36] [37] [38] 54, 59 . Moreover, divorced/widowed participants showed the highest trait loneliness, followed by singles (never-married). Thus, the loss of previously experienced positive relationship aspects such as romantic support, solace, and physical proximity, may be associated with feelings of loneliness. Furthermore, individuals who were in a relationship and living alone ("living apart together"), were lonelier than those who were living with their partner, but did not differ in their momentary loneliness levels compared to singles living alone. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Being in a relationship and living with others was associated with similar levels of loneliness compared to being single and living with others. This indicates that, during extreme physical isolation and contact restrictions, having a partner per se does not protect against loneliness, but rather living with others becomes an increasingly important buffer for loneliness. As during hard lockdown, intimacy and physical closeness are lacking in couples who are living apart, these important stress-buffering factors in the romantic relationship are suddenly missing, which is experienced as aversive 68 . Contrary to this finding, Greenfield and Russel found higher loneliness levels in couples who were living apart but with others 59 . One explanation for these conflicting findings could be that during lockdown, there were no alternatives for direct social interactions outside the apartment and thus the co-habitants became an especially important substitute for any direct contact with the romantic partner. We further found that higher relationship quality predicted lower momentary loneliness levels, which is in line with cognitive approaches to loneliness assuming that quality rather than quantity of social relationships buffers short-term psychological burden. However, relationship quality did not moderate the association between living situation and loneliness. Thus, the protective effect of living together during the COVID-19 lockdown was evident irrespectively of the relationship quality. In the online survey, female participants reported significantly higher trait loneliness levels than male participants. This adds to numerous studies revealing female gender as risk factor for loneliness 76, 77 . Interestingly, however, recent neuroimaging studies indicate that lonelinessassociated neural effects may be more pronounced in high lonely men than women 78, 79 .

Although the results support our hypotheses about the importance of structural and psychological factors for self-reported loneliness, there are many other potential psychological mediators explaining these associations. It is important to keep in mind that romantic relationships buffer against negative mental and physical health consequences only under certain circumstances, for instance if marital functioning is perceived as positive 33 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Furthermore, social dimensions such as perceived social proximity, knowing that there is someone you can count on, as well as actually perceived support may be important underlying mechanisms influencing psychobiological health 29 .

On a neuroendocrine level, being in a relationship buffered momentary cortisol levels and their association with loneliness. This is in line with theoretical and empirical literature indicating that having a romantic partner serves as a biological zeitgeber, regulating optimal stimulation by modulating arousal levels and attenuating stress 80 . These results show us that romantic relationships have a direct impact on neuroendocrine stress responses, which in a longterm may have a positive effect on health-related outcomes 21, 22 . Contrary to our hypothesis, living arrangements by themselves neither affected cortisol levels nor moderated the association between momentary loneliness and cortisol levels. One reason why we only found these associations with relationship status, could be, that there may be operators that are unique in relationships. For instance, feelings of connectedness 81 , intimacy 41 or affective touch 82 are specific driving factors in romantic relationships. As they are not characteristic for other relationships such as co-habitants, they only come into use when romantic relationships are investigated.

This study adds to previous research on social buffering 16, 26, 27, 29 in the context of enduring stress and extreme physical isolation. As lockdown-related long-term psychological health problems are increasingly revealed, it is important to study structural and psychological factors that might influence those consequences. Furthermore, short-term neuroendocrine responses during lockdown could help unravel the neurobiological mechanisms underlying detrimental effects of loneliness and social isolation for mental health. Using a psychobiological EMA design, we were able to assess not only trait loneliness levels, but also moment-to-moment variations in loneliness and salivary cortisol in a naturalistic setting. The every-day life assessments took place in the individuals' personal environments, which yielded highly . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint ecologically valid data. Furthermore, as the participants' current loneliness levels were directly assessed, reporting errors due to retrospective assessment could be reduced. In order to represent the hierarchical structure of the data, MLM was used, enhancing statistical power of the analyses. Moreover, due to the close supervision of the participants, we were able to keep their commitment high and thus collect high-quality data. Another strength of this study is the wide range of the participants' age, making the sample more representative for every age group. This study has several limitations that need to be addressed. First of all, sample sizes differed between demographic groups. For example, 70 divorced/widowed individuals and 329 singles participated in the online survey. We recruited a convenience sample and widowers/widows and divorced individuals are on average older and less technically involved than singles, which made it more difficult to recruit them in an online survey. Another limitation is the cross-sectional design of the study, which makes it impossible to draw causal conclusions on long-term (mental) health outcomes. Furthermore, there is no baseline assessment of the variables of interest before lockdown, therefore we were not able to control for the participants' pre-lockdown levels of loneliness and cortisol. Thus, our results can only be seen as a "snapshot" of the current situation.

There are several aspects that could be addressed in future research. Although we found main effects of relationship status, living situation, and relationship quality, they only explained a small amount of variance in the outcomes. This indicates that there are additional predictor and moderator variables influencing the outcomes. Furthermore, the stress-buffering effects of close relationships is not restricted to romantic relationships. For example, having meaningful relationships with close friends or relatives 38 could be one protective factor. In addition, . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint longitudinal assessments with repeated within-person measurements of loneliness and cortisol over a longer period of time could be implemented, in order to probe long-term psychological and physiological consequences of COVID-19 and strict lockdowns.

All in all, our study reveals further evidence for romantic relationships as a protective factor against trait and state loneliness, both on a structural level (alone vs. in a relationship) and a psychological level (relationship quality), as well as momentary cortisol levels during the ongoing stress of the pandemic and social isolation. Additionally, living with others during lockdown protects against loneliness in every-day life. The fact that individuals who were living apart from their partner displayed similar levels of loneliness compared to singles, implicates that especially in times of social isolation, the lack of direct physical contact to the partner makes a difference when it comes to psychological burden. This joint role of partnership and living situation should be taken into account when analysing structural factors for negative mental health outcomes, but also identifying resources for resilience. Furthermore, it is especially important to consider not only relationship status, but also relationship quality as an important psychological aspect of romantic relationships and a buffering factor for loneliness in couples, potentially counter-balancing the negative effects of living alone. This is in line with previous epidemiological research suggesting that rather than being married, it is the satisfaction with the relationship (e.g., the amount of support or criticism from a partner), which influences health-related outcomes 83 . All in all, in the context of clinical interventions, the results implicate that especially singles and divorced individuals, women, couples with low relationship quality as well as alone living residents (whether single or in a relationship) should be offered psychosocial support in order to prevent them from long-term negative health consequences. New technical methods such as smartphone apps could provide useful daily interventions or telemedical supervision. More importantly, on the one hand, individuals who are living apart from their partner, could be offered interventions to enhance their perceived . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint relationship quality, on the other hand, alone living single individuals should be offered help in re-establishing meaningful social bonds with their close friends in order to counter-regulate their feelings of loneliness. Finally, public health campaigns should address and sensitize the society towards loneliness and mental health symptoms in those different groups to empower individuals to actively approach social offers and use them as resource.

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint to all the data in the study and had final responsibility for the decision to submit the manuscript for publication.

The authors declare no competing interests. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Table 1. This table depicts is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

(1) Level 1: lnCortij = ßoi + ß1 C_lonelyij + ß2 eatij + ß3 drinkij + ß4 alcoholij + ß5 caffeineij + ß6 physical activityij + ß7 time_riseij + ß8 time_fallij + ß9 dayij + εij Level 2: ß0j = γ00 + γ10 GC_lonely + υ0i

(2) Level 1: lnCortij = ßoi + ß1 C_lonelyij + ß2 C_lonelyij*relationship.i. + ß3

C_lonely*living.j + ß4 eatij + ß5 drinkij + ß6 alcoholij + ß7 caffeineij + ß8 physicalactivityij + ß9 time_riseij + ß10 time_fallij + ß11 dayij + εij Level 2: ß0j = γ00 + γ10 GC_lonely.j + γ11 age.j + γ12 sex.j + γ13 bmi.j υ0j

Where i denotes the measurement nested in person j, vector C_lonely captures person-meancentered momentary loneliness levels. GC_lonely captures grand-mean-centered loneliness varying on the person level (level 2), and relationship (0 = Single, 1 = In a relationship) and

living (0 = Alone, 1 = With others) characteristics also varying on level 2. The vectors C_lonelyij*relationshipi and C_lonely*living.j represent cross-level interactions, with relationship.j and living.j being level 2 predictors. Finally, εij denotes individual variations, whereas υ0i represents differences between each person's mean from the global mean.

Pseudo R² of the significant predictors was calculated as follows: (σ²0 -σ²1)/ σ²0

Where σ²0 denotes the amount of variance explained before including the predictor, and σ²1 denotes the amount of variance explained after including the predictor. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint Appendix B -Results of the multilevel models Table 7 Results of the multilevel models with loneliness, relationship status and living situation as predictors and cortisol levels (ln-transformed) as outcome.

Fixed Slopes Model Random Slopes Model is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. Notes. Model 1 = Random intercept-only model; model 2 = Random slopes-model with level 1 -loneliness set as random predictor; df = degrees of freedom; AIC = Akaike's information criterion (goodness of fit index); BIC = Schwarz's Bayesian criterion (goodness of fit index); LL = Log likelihood, L Ratio = Likelihood ratio; p = pvalue.

. CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 26, 2022. ; https://doi.org/10.1101/2022.02.25.22271461 doi: medRxiv preprint

Forced social isolation and mental health: A study on 1,006 italians under COVID-19 lockdown

Mental health in the COVID-19 pandemic

The impact of lockdown stress and loneliness during the COVID-19 pandemic on mental health among university students in Germany

Mental health prevalence and predictors among university students in nine countries during the COVID-19 pandemic: a cross-national study

Loneliness and social isolation-a private problem, a public issue

Loneliness in the general population: prevalence, determinants and relations to mental health

Preventing the harmful consequences of severe and persistent loneliness

Social relationships and mortality risk: a meta-analytic review

The need to belong: desire for interpersonal attachments as a fundamental human motivation

Loneliness Across the Life Span

Insula reactivity mediates subjective isolation stress in alexithymia

The effect of loneliness on depression: A meta-analysis

Review on cardiovascular risk prediction

Social support and social networks in COPD: A scoping review

Loneliness, social isolation, their synergistic interaction, and mortality

Loneliness and social isolation as risk factors for mortality: A meta-analytic review

The association between loneliness, social isolation and all-cause mortality in a nationally representative sample of older women and men

Loneliness, health, and mortality in old age: A national longitudinal study

Marital status and mortality in the elderly: a systematic review and meta-analysis

Loneliness and the Social Brain: How Perceived Social Isolation Impairs Human Interactions

Loneliness and neuroendocrine, cardiovascular, and inflammatory stress responses in middle-aged men and women

Day-to-day dynamics of experience-cortisol associations in a population-based sample of older adults

Loneliness and Diurnal Salivary Cortisol in Emerging Adults

Loneliness and cortisol: Momentary, day-to-day, and trait associations

Handbook of Neuroendocrinology

in Handbook of Psychology and Health (Volume IV) Ch

Stress, social support and cardiovascular activity over the working day

Does emotional support influence survival? Findings from a longitudinal study of patients with advanced cancer

Psychobiology of social support: The social dimension of stress buffering

Handbook of Social Support and the Family

Marriage, divorce, and the immune system

Marriage and health: His and hers

The physiology of marriage: pathways to health

Marital quality and health: a meta-analytic review

Marital status, close social network and loneliness of older adults in the czech republic

Loneliness among older European adults: results from the survey of health, aging and retirement in Europe

Loneliness in married, widowed, divorced, and never-married older adults

Age differences in loneliness from late adolescence to oldest old age

Marital status as a predictor of diurnal salivary cortisol levels and slopes in a community sample of healthy adults

Intimacy as Related to Cortisol Reactivity and Recovery in Couples Undergoing Psychosocial Stress

Positive couple interactions and daily cortisol: on the stress-protecting role of intimacy

Increased risk of acute cardiovascular events after partner bereavement: a matched cohort study

A systematic review of longitudinal risk factors for loneliness in older adults

Divorce in Europe -New Insights in Trends, Causes and Consequences of Relation Break-ups

Loneliness, optimism, and well-being among married, divorced, and widowed individuals

Neuroendocrine mechanisms of grief and bereavement: A systematic review and implications for future interventions

Developing and testing a model of loneliness

From social structural factors to perceptions of relationship quality and loneliness: The chicago health, aging, and social relations study

Quality of Marriages in Later Life and Emotional and Social Loneliness

Marital quality and loneliness in later life: A dyadic analysis of older married couples in Ireland

Two-Wave Dyadic Analysis of Marital Quality and Loneliness in Later Life: Results From the Irish Longitudinal Study on Ageing

Lonely You: Loneliness and the Longitudinal Course of Relationship Satisfaction

Social Isolation, Loneliness, and Living Alone: Identifying the Risks for Public Health

Living alone and positive mental health: a systematic review

Feelings of Loneliness and Living Alone as Predictors of Mortality in the Elderly: The PAQUID Study

Marital status, living arrangements, and mortality in middle and older age in Europe

Identifying Living Arrangements That Heighten Risk for Loneliness in Later Life: Evidence From the U.S. National Social Life, Health, and Aging Project

Loneliness in the UK during the COVID-19 pandemic: Crosssectional results from the COVID-19 psychological wellbeing study

Widowhood, widowhood duration, and loneliness among older adults in China

Loneliness during a strict lockdown: Trajectories and predictors during the COVID-19 pandemic in 38,217 United Kingdom adults

The trajectory and determinants of loneliness during the early months of the COVID-19 pandemic in the United States

Term Impact of Economic Downturn and Loneliness on Psychological Distress: Triple Crises

Social isolation and diurnal cortisol patterns in an ageing cohort

Developing the Yearning in Situations of Loss Scale: Convergent and Discriminant Validity for Bereavement, Romantic Breakup, and Homesickness. Death Stud

Heartbreak and Bereavement¡ -Romantic Breakups

Every time you go away: Changes in affect, behavior, and physiology associated with travel-related separations from romantic partners

Impact of loneliness on diurnal cortisol in youth

The Revised UCLA Loneliness Scale: Concurrent and discriminate validity evidence

Bericht über den 13. Kongreß für Angewandte Psychologie

Fragebogen zur Partnerschaftsdiagnostik (FPD). 2. Auflage. (Hogrefe

Loneliness around the world: Age, gender, and cultural differences in loneliness

How to assess and interpret everyday life salivary cortisol measures: A tutorial on practical and statistical considerations

Discovering Statistics Using R

Loneliness in the COVID-19 pandemic: Associations with age, gender and their interaction

Gender-related differences in the psychological impact of confinement as a consequence of COVID-19 in Spain

000 social brains: Sex differentiation in human brain anatomy

Lonely in the dark: trauma memory and sex-specific dysregulation of amygdala reactivity to fear signals

Coregulation, dysregulation, self-regulation: an integrative analysis and empirical agenda for understanding adult attachment, separation, loss, and recovery

The association between relationship strain and emotional well-being among older adult couples: the moderating role of social connectedness

The social buffering of pain by affective touch: a laser-evoked potential study in romantic couples

We are deeply grateful for the indescribable support of C. Gäbel, who provided us her programming code for the ecological momentary assessment. We further want to thank our