Psychosocial factors affect mental health and health-related quality of life (HRQL) in a complex manner, yet gender differences in these interactions remain poorly understood. We investigated whether psychosocial factors such as social support and personal and work-related concerns impact mental health and HRQL differentially in women and men during the first year of the COVID-19 pandemic. Between June and October 2020, the first part of a COVID-19-specific program was conducted within the “Characteristics and Course of Heart Failure Stages A-B and Determinants of Progression (STAAB)” cohort study, a representative age- and gender-stratified sample of the general population of Würzburg, Germany. Using psychometric networks, we first established the complex relations between personal social support, personal and work-related concerns, and their interactions with anxiety, depression, and HRQL. Second, we tested for gender differences by comparing expected influence, edge weight differences, and stability of the networks. The network comparison revealed a significant difference in the overall network structure. The male (N = 1370) but not the female network (N = 1520) showed a positive link between work-related concern and anxiety. In both networks, anxiety was the most central variable. These findings provide further evidence that the complex interplay of psychosocial factors with mental health and HRQL decisively depends on gender. Our results are relevant for the development of gender-specific interventions to increase resilience in times of pandemic crisis.
It is well known that physical and mental health are affected by psychosocial factors1,2,3,4, both in general and in times of crisis, e.g., a pandemic5,6,7. Factors like social support play a protective role1,8,9. In contrast, concerns (e.g., about one’s family or work) are related to deprivations in mental health10,11,12.
Social support has been broadly defined as “support accessible to an individual through social ties to other individuals, groups, and the larger community”13, i.e., having people to rely on who provide care, value, and love14. Social support is associated with lower levels of depression and anxiety1,8, particularly during times of crisis like the COVID-19 pandemic7. Previous studies differentiated between personal support, i.e., provided by family members and friends6,15,16, and work-related support, i.e., provided by supervisors and co-workers17,18. Both forms of social support have been associated with preserving and improving mental health and health-related quality of life (HRQL)19,20,21.
In contrast, concerns are generally associated with higher levels of perceived psychological stress22,23 and deteriorations of mental health and HRQL10,12,24. In one recent US-based study10, concerns about the well-being of one’s children or the health of other relatives were associated with decreased individual well-being during the COVID-19 pandemic25. Work-related concerns like job insecurity and financial concerns were associated with decreased mental health12 and overall well-being in employees11. In students, financial concerns were associated with self-reported anxiousness, nervousness, and sleep problems26. These results imply that social support is generally positively related to well-being and HRQL27,28, whereas financial, work-related, and personal concerns have negative effects26,29,30.
Previous studies have investigated whether social support and concerns have differential effects on mental health in women and men. Prior to discussing these differences, it is necessary to introduce sex and gender as two different concepts that are often conflated. The importance of differentiating between sex and gender in health and clinical research has been underlined repeatedly31,32. Sex, a biological construct, refers to the purely biological differences between males and females enabling sexual reproduction, whereas gender, a social construct, refers to the set of behaviors, interests, social conventions and roles associated with male or female sex33,34. Given the concept’s strong link to the social context31,32, the present study focusses on gender differences in psychosocial factors and mental health. Previous works have shown important gender-related differences in psychosocial and clinical variables. For instance, some studies suggest that women depend more upon personal support35,36,37, whereas men may derive more benefit from work-related support38,39. Women showed lower symptoms of depression and anxiety than men when perceiving social support provided by family members or friends40,41. In men, work-related support from co-workers and supervisions showed protective effects against depressive and anxiety disorders38,39,42. Elsewhere, this association of work-related support to mental health was stronger in men than in women43,44. Regarding concerns, previous studies found stronger deteriorating effects of work-related concerns on men’s health45, e.g., during job insecurity and unemployment46. Addressing personal concerns, women were more prone to psychological distress when neglecting their family due to their work37. Similarly, there were more work-family conflicts in mothers than fathers47, accompanied by factors like higher stress and lower life satisfaction48.
Inconsistent with these findings, other studies revealed comparable effects of social support and concerns in women and men of different ages2,27,28. For example, a study investigating effects of social support on self-reported physical and mental health in Slovak adolescents (N = 2616, mean age = 15) found lower social support and higher (mental) health in boys compared to girls, yet no gender difference in the positive relation between social support and health27. Hann et al.28 found a gender-independent negative relation of social support and depressive symptoms in cancer patients. In a large cross-sectional data set of working European adults (N = 40,089), work-related support mediated the negative association between work-related stress and acute mental well-being and health similarly in women and men2. Work-related concerns like job insecurity, resulting financial concerns, and psychosocial stress at work also had negative effects on mental health in both men and women9,11.
In summary, it remains unclear whether psychosocial variables affect men’s and women’s mental health differently. Moreover, the different dimensions of social support and concerns (personal and work-related) were previously assessed in separate studies or analyses, without accounting for the complexity of their interrelations as well as their associations with mental health and HRQL.
To overcome these limitations, we applied network analyses to quantify the complex associations between important psychosocial variables (personal and work-related support; personal and work-related concerns), mental health (i.e., anxiety and depression), and HRQL in women and men during the COVID-19 pandemic. In comparison to traditional methods such as regression analysis and factor analysis, network analysis allows for the examination of complex relationships between multiple variables simultaneously, providing a more comprehensive understanding of the factors influencing mental health49. This is particularly important given the intricate nature of mental health. Mental health often involves numerous interconnected factors50. By analyzing the interrelations between different psychosocial variables, network analysis can contribute to the identification of the most influential factors affecting mental health51,52. Identifying these central factors can help prioritize interventions and allocate resources more effectively53.
Inspired by previous work6,17,22, we hypothesized that work-related and personal support as well as work-related and personal concerns are core components in a network of psychosocial and psychopathological variables directly connected with individual differences in anxiety, depression, and HRQL, especially during the COVID-19 pandemic. We predicted positive relations between social support and mental health, and negative relations between concerns and mental health10,12,15. These associations might be comparable between women and men2,27,28. Alternatively, inspired by studies showing gender differences35,39,43, work-related concerns and support might be more closely associated with mental health in men, while personal concerns and support should be more influential in women.
As indicated above, we applied a network approach to test these assumptions54,55. The proponents of network theory in clinical psychology assume that mental disorders arise from complex relationships between symptoms and thus conceptualize psychiatric disorders as emergent phenomena56. For instance, when taking a network perspective on psychopathology, depression constitutes a dynamic network of interdependent symptoms57. Analyses based on a network approach allow a graphical representation of all variables as individual nodes. This helps identify variables particularly associated with others, such as links between symptoms of different disorders that could explain comorbidities56,58. Several studies have used this method to address psychiatric research questions, e.g., examining relations between neighborhood social environment and symptoms of paranoia59, the impact of the COVID-19 pandemic on mental health60, or associations between items measuring individual differences in depression and anxiety61,62. We therefore applied a network approach to investigate the hypothesis that during the first year of the COVID-19 pandemic, the complex interplay between concerns, social support, anxiety, depression, and HRQL is modified by gender63,64,65,66.
The Characteristics and Course of Heart Failure Stages A–B and Determinants of Progression (STAAB) study is a cohort of 5010 randomly sampled volunteers of the general population of Würzburg, Germany recruited between 2013 and 2017 (source population 124,297 inhabitants as of 2011 census). Inclusion of STAAB participants is restricted to 30–79 years of age at the baseline measurement and sampling was stratified 1:1 for gender and 10:27:27:27:10 for age groups of 30–39/40–49/50–59/60–69/70–79 years. In order to ascertain a continuously balanced recruitment, invitations to the study were sent out in batches that were iteratively re-adjusted for gender and pre-defined age groups according to the respective response rates. The complete study design and the rationale of the STAAB cohort is described in detail elsewhere67. Here, we analyze data from the STAAB-COVID-One program, which assessed the psychosocial impact of the COVID-19 pandemic of participants of the STAAB cohort during the first year of the pandemic. On-site participation including blood samples was performed in the joint study line of the German Center for Heart Failure (DZHI) and the Institute for Clinical Epidemiology and Biometry (IKE-B) in Würzburg. Afterwards, participants answered the survey items used in the present study at home via online survey or a paper-based questionnaire. Data were collected between June 17th and October 18th 202068. Based on evidence from previous work as reported above, we inserted social variables capturing personal support and concern6,17,22 as well as work-related support and concern35,39,43 into the ongoing study (see “Measures” for more details). All procedures within the STAAB program are subject to a rigorous quality control, follow predefined standard operating procedures, and comply with the Declaration of Helsinki. The study protocol and procedures were approved by the ethics committee of the medical faculty of the University of Würzburg (vote #98/13) and data protection office (#J-117.605-09/13). The final sample size for the present analysis included 2890 participants (1520 female, 1370 male; note that there was no option to identify as non-binary) with a mean age of 59.6 (SD = 11.19, range 34–85). All participants provided written informed consent prior to any study examination.
To test our research question, we applied for access to variables that (A) assess HRQL and important indicators of mental health and (B) variables that address personal and work-related concern as well as support. In the STAAB survey, mental health outcomes were measured using validated scales for depression (Patient Health Questionnaire-9; PHQ-9)69 and generalized anxiety (Generalized Anxiety Disorder Scale, GAD-7)70. For both the PHQ-9 and the GAD-7, a score ≥ 10 indicates clinically relevant levels of depression and anxiety, respectively. HRQL was assessed using the visual analogue scale of the EQ-5D (range 0–100)71. Higher scores indicate higher HRQL.
To document personal support, we used variables assessing perceived social support and emotions associated with family members (“How strongly do you feel supported by your social environment?”, “How do you feel towards the individuals in your family/ in the household at the moment?”, rated on a 5-point Likert scale ranging from 0 = “very bad” to 4 = “very good”; “Did your feelings towards your family member/ individuals in your household change since the COVID-19 pandemic?”, rated on a 3-point scale with 0 = “deteriorated”, 1 = “unchanged”, and 2 = “improved”). For work-related support, we used variables assessing perceived social support and emotions associated with co-workers and supervisors (“How strongly do you feel supported by your colleagues?”, “How strongly do you feel supported by your supervisor?”, “How do you feel towards the individuals at work at the moment?”, rated on a 5-point Likert scale from 0 = “very bad” to 4 = “very good”; “Did your feelings towards the individuals at work change since the COVID-19 pandemic?”, rated on a 3-point scale with 0 = “deteriorated”, 1 = “unchanged”, and 2 = “improved”). To capture personal concern, we used three items related to “burden of caring for children, parents, or other family members”, “having no one to discuss the problems with”, and “burdens due to contact bans with older people (e.g., parents, grandparents, etc.)”. For work-related concern, we used two items targeting “stress at work or school” and “financial problems or worries”. Concern variables were assessed using a three-point Likert-scale (0 = “not impaired”, 1 = “lowly impaired”, 2 = “heavily impaired”). The internal consistencies (Cronbach’s α) of PHQ, GAD, personal support, work-related support, personal concern, and work-related concern were 0.85, 0.87, 0.59, 0.78, 0.49, and 0.44, respectively.
All statistical analyses were conducted using R Version 4.0.4. The structure of associations between the included variables (i.e., personal support, work-related support, personal concern, work-related concern, depression, anxiety, HRQL, age) was analyzed using a regularized partial correlation network. From this network, we identified the most central variables, their predictability, network stability, and inference accuracy72,73. To investigate gender differences, we conducted two separate network analyses for women and men, respectively. In addition, we calculated a joint network of women and men to compare the results with the differential networks.
We estimated the network structures via Mixed Graphical Models (MGMs) using the R‐packages mgm74 and bootnet73,75, as MGMs explicitly account for mixed measurement scales of variables. “Nodes” describe the individual variables (e.g., depression symptoms) while the “edges” describe the statistical relationships between them. The so-called “centrality index” determines the importance of each node within the network based on the strength of its connections to other nodes. Nodes and edges were computed with the Fruchterman-Reingold algorithm from the R-package qgraph75. To investigate how support and concern variables were associated with mental health symptoms (depression, anxiety) and HRQL, all seven variables were included into a Gaussian graphical model (GGM)76. Here, a connection between two nodes represents the connection after controlling for all other edges in the network, similar to partial correlations. For the seven variables (i.e., seven nodes), we estimated 21 pairwise parameters for each network. We employed a penalty approach to control for potential spurious associations which would lead to false-positive findings, namely the least absolute shrinkage and selection operator (LASSO)77. LASSO shrinks edge weights, reducing smaller edges to zero and usually resulting in a sparse structure. To choose the most appropriate network, we used the Extended Bayesian Information Criterion78, setting its hyperparameter to default (γ = 0.25). We tested for redundant nodes using the ‘goldbricker’ function from the networktools R-package79. No statistically redundant nodes were indicated. For visualization of edge weights, we consistently used the value of the highest edge weight across a network for the whole (= joint) sample, the female subset, and the male subset (w = 0.678) to ensure a homogenous scaling of edge depictions across all networks.
We report two main parameters for assessing nodes80,81. First, we calculated the expected influence (EI) as centrality index for the female and male networks82. The EI captures the degree of connectivity of a specific node with the network’s other nodes. Its value represents the sum of all edge weight values that are interconnected with a particular node83. Second, predictability estimates were derived for all variables in the networks and implemented visually as pie charts around each node. Predictability quantifies the variance of a node explained by all its neighbors measured as R272. We also conducted difference tests to compare individual centrality estimates as a non‐parametric bootstrapping method with 2000 iterations.
To visually compare female and male participants, we created a difference network based on the difference weights matrix. The matrix was calculated by subtracting the female from the male subsample weight matrix. To compare both networks, we performed the non-parametrical permutation-based Network Comparison Test (NCT)84 with 1000 permutations. As the method compares multiple edges between networks, we applied Bonferroni adjustment for multiple comparison correction.
Network accuracy and stability
For the female and male networks, edge weight stability was tested by bootstrapping 95% confidence intervals (CI) of the edge weights (Fig. S3). The order stability of the centrality measures was tested by subset bootstrap procedures, so that the central node should remain central after dropping random participants (Fig. S4)85. As index, we estimated the centrality stability (CS coefficient), which should be at least 0.25 and preferably above 0.5. Further accuracy and stability analyses for this network, i.e., edge weight significance tests (testing for significant differences for all edges) and centrality difference tests (testing for centrality differences for all nodes), can be found as Supplementary Figs. S4 and S5.
Finally, we tested the split‐half reliability as a confirmatory analysis of the gender-specific networks. Using the psychometrics R-package86, we estimated a GGM in a randomly-drawn half of the female and male dataset, respectively, and then fitted a confirmatory network model to the second half.
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
3034 persons participated in the STAAB-COVID-One program, providing blood sample and questionnaire data. Of those participants, 2892 provided complete questionnaire data, but two were excluded as they did not report their gender. Thus, 2890 (60%) of the 4860 participants remaining in the study provided complete questionnaire data (53% female; mean age women = 59, SD = 11.0; mean age men = 60, SD = 11.4). Descriptive statistics for the analyzed variables are presented in Table 1. The prevalence for depression (PHQ) and anxiety (GAD; cut-off values ≥ 10) were 8% and 4%, respectively.
Analyzing network nodes
Analyzing centrality, i.e., the importance of a node for the connectivity of the network, GAD showed the highest EI (0.61) in the joint network, in the male (0.69), and in the female network (0.56; Fig. 1). In all networks, HRQL had the most negative EI (joint = − 0.20; women = − 0.17; men = − 0.25).
To secure these results, we tested the accuracy of the network and order stability of the centrality measures. According to the width of the CIs, the bootstrapped edge weight analysis suggests that the models were estimated accurately (Fig. S2). The CS‐coefficients of 0.75 suggest a stable order of centrality85. The centrality difference tests indicate that EI significantly differs between most nodes (Fig. S4). For men and women, the node with the largest EI, GAD, was significantly larger than all other nodes except PHQ. Additional edge weight significance tests are illustrated in Fig. S5.
Average node predictabilities in the female and male network were 0.36 and 0.35, respectively. The nodes with the highest predictability in both networks were PHQ (female R2 = 0.69, male R2 = 0.67) and GAD (female R2 = 0.67, male R2 = 0.66). The node with the lowest predictability were age (female R2 = 0.02, male R2 = 0.04) and HRQL (female R2 = 0.31, male R2 = 0.23).
Analyzing network structures
Figure 2 illustrates the network structure of the joint network. The strongest positive associations were found between GAD and PHQ (w = 0.68), work-related and personal support (0.27), and work-related and personal concern (w = 0.24). The strongest negative associations were found between depression and HRQL (w = − 0.29), work-related concern and work-related support (w = − 0.24), personal concern and personal support (w = − 0.21), and GAD and personal support (w = − 0.18). The average node predictability in the joint network was 0.36, i.e., a moderate level of predictability compared with other psychopathology networks72. The nodes with the highest predictability were PHQ (R2 = 0.69) and GAD (R2 = 0.67). The node with the lowest predictability was age (R2 = 0.04).
A formal network comparison revealed a significant gender effect in the omnibus test on network structure (max. difference in edge weights = 0.255, p = 0.004, 1000 permutations), with comparable overall connectivity strength (global strength difference = 0.04; male sample network strength = 2.34; female sample network strength = 2.38, p = 0.891, 1000 permutations). Work-related concern and support had higher EI scores in the male network (concern: puncorrected = 0.032; pcorrected = > 0.99; support: puncorrected = 0.045; pcorrected = 0.464; Fig. 1).
The network structure of the female and male subsamples and the difference network are illustrated in Fig. 3. They reveal significant gender differences regarding both interconnectivity and relations with mental health outcomes. A gender effect in the connectivity between GAD and work-related concern (Bonferroni corrected p = 0.028; Fig. 3c) reflected a positive association between GAD and work-related concern for men (wmale = 0.19; Figs. 3b and S1b), but not for women (wfemale = 0, Figs. 3a and S1a). Descriptively, the negative association between work-related concern and personal support (wfemale = 0, wmale = 0.12, puncorrected = 0.143, pcorrected > 0.99) was only present in the male, but not in the female network. The negative association between work-related support and PHQ (wfemale = -0.09, wmale = 0, puncorrected = 0.226, pcorrected > 0.99) as well as the positive associations between personal concern and GAD (wfemale = 0.07, wmale = 0, puncorrected = 0.290, pcorrected > 0.99) and personal support and HRQL (wfemale = 0.11, wmale = 0, puncorrected = 0.193, pcorrected > 0.99) were only present in the female, but not the male network.
Apart from these differences, both women and men showed strong positive associations between work-related and personal concern (wfemale = 0.21, wmale = 0.26), GAD and PHQ (wfemale = 0.67 and wmale = 0.66), and work-related and personal support (wfemale = 0.22, wmale = 0.33). The strongest negative associations were found between PHQ and HRQL (wfemale = − 0.28, wmale = − 0.25) and personal concern and personal support (wfemale = − 0.19, wmale = − 0.19).
The split-half reliability analysis suggested a fit with a root mean square error of approximation of 0.05 (95% CI 0–0.097) for women, a fit of 0.08 (95% CI 0.04–0.12) for men, and a comparative fit index of ≥ 0.98 for both women and men87.
Psychosocial factors like social support and financial, work-related, or personal concerns are related to positive and negative states of mental health and quality of life6,20,88. However, data investigating whether these relationships are the same for men and women are lacking. We therefore used network analysis and a large dataset from the general population of Würzburg collected during the first year of the COVID-19 pandemic to investigate the complex associations between social support, concerns, mental health, and health-related quality of life in women and men.
In general, we hypothesized that work-related and personal support as well as work-related and personal concerns would be central nodes in our psychosocial and psychopathological network. We expected positive associations between social support and mental health, and negative associations between concerns and mental health6,17,22. Extending previous research, we further inspected the dependence of these effects and interrelations on gender. Some previous research suggests comparable associations between women and men10,12,15, whereas other findings suggest stronger associations of mental health with work-related concerns and support in men and with personal concerns and support in women35,39,43.
Independent of gender, anxiety was the most central network variable. However, the analysis of structural network connectivity revealed that work-related concerns connect differently with mental health outcomes in women and men. In more detail, increasing anxiety scores were associated with increasing work-related concerns in men, while there was no effect of work-related concerns in women. Instead, increasing anxiety scores were associated with increasing concerns regarding family and friends. Moreover, women showed a positive relation between personal support and HRQL, indicating an increase in quality of life if perceived social support provided by friends and family is high.
The gender-specific effects of work-related concern agree with previous results showing that men perceive higher work-related psychological distress9 and are more affected by job insecurity and unemployment46, while women report higher distress when neglecting their family due to their work37,47,48. The finding of stronger work-related associations in men and stronger family and friend-related associations in women may be traced back to traditional gender norms and roles. Traditionally, the stereotypical male gender role focuses on work and providing resources for the family and the female gender role focuses on caregiving89,90. A review on gender and mental health90 proposes that stressors are more harmful in those domains that are more closely associated with one’s gender role, i.e., work for men and family for women, in line with our findings. In contrast to our findings, some previous research found no gender differences in the effect of work-related concerns on mental health9,11,45, but these studies used a less detailed statistical approach compared to our network analyses. In recent research during the COVID-19 pandemic, female employees perceived the financial consequences of the pandemic as less extreme than male employees91, further supporting a higher relevance of work-related concerns for men’s mental health46. There was no significant gender difference in the negative association between personal concerns and psychopathology10,92. Moreover, although protective effects of social support against mental health deprivations are well known, especially during the pandemic6,17,18, work-related and personal support did not play a central role in our networks. Note also that both men and women showed a negative relation between personal support and mental health, i.e., anxiety. Previous findings have implied a stronger dependency of women compared to men on social support for mental health enhancement35,40. In our study, a gender-specific effect emerged for personal support and HRQL: personal support by family and friends was associated with higher HRQL in women, while this relation was absent in men. This once again aligns with the traditional female family-related role, which includes a higher tendency to engage in close social contacts and seek social support for increased well-being and stress reduction93,94. Moreover, this stronger role of social support in women even emerged beyond the family and friend context, as women showed lower depression with higher work-related support. Thus, with regard to the work context, men’s mental health was negatively associated with work-related concerns, whereas women’s mental health may profit from work-related support.
For both men and women, connectivity analyses showed strong positive association between work-related and personal concerns as well as work-related and personal support. In line with previous studies independent of a pandemic context95,96,97, we found a positive relationship between trait anxiety (GAD) and depression (PHQ), and a negative relationship between depression and HRQL during the COVID-19 pandemic. Moreover, the joint network revealed that personal support was associated with decreasing anxiety, further underlining the mental health-increasing effect of personal support1,7,8.
There are some limitations to consider when interpreting the results. We conducted a cross‐sectional analysis of a representative age and gender-stratified sample aged 34 to 85 years at baseline investigation drawn from the city of Würzburg located in Southern Germany. Therefore, the findings may not be transferable to populations with a different age context, location, or ethnic background. Also considering that previous research indicated differences in network connectivity in different populations98, the findings may not be generalizable to clinical samples. In addition, the 8% prevalence of depression and 4% prevalence of anxiety are in line with other population-based surveys in Germany99,100,101, but disagree with recent research indicating increases in depression and anxiety scores to 11% during the pandemic102. Note also that our data was obtained during the COVID-19 pandemic, i.e., a context of increased impairments in mental health103,104. Given that the COVID-19 pandemic provided a very specific context, it still needs to be clarified if our results generalize to other pandemic-independent situations. Future studies should validate our findings in more heterogeneous samples, clinical samples, and a non-pandemic context. Moreover, the cross-sectional design does not allow us to test whether the observed associations are affected by the pandemic. Finally, to test our a priori hypotheses, we only analyzed a selection of potentially relevant variables. Statistical models are solely able to inform about variance between items included in the model. As the COVID-19 pandemic resulted in very individual concerns and support strategies105, it was not trivial to decide which variables should be integrated into such a complex system. Therefore, confirmatory research is needed to validate the importance of other factors in psychopathological networks. Finally, future studies should collect socio-economic data, as this would enable a comprehensive understanding of the complex interplay between socio-economic factors, psychosocial variables and psychological well-being, facilitating the development of targeted interventions and policies.
Overall, our study supports the need to consider psychosocial factors and psychopathology as an interrelated network of protective and detrimental influences. As one of the first studies, we point out gender-related differences within these networks while also highlighting the influence of psychosocial factors. Our findings underline the importance of targeting social aspects in therapeutic interventions for the improvement of women’s and men’s mental health.
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to data protection restrictions.
Cohen, S. Social relationships and health. Am. Psychol. 59, 676–684 (2004).
Mensah, A. Job stress and mental well-being among working men and women in Europe: The mediating role of social support. Int. J. Environ. Res. Public Health 18, 2494 (2021).
Suls, J. & Rothman, A. Evolution of the biopsychosocial model: Prospects and challenges for health psychology. Health Psychol. 23, 119–125 (2004).
Taylor, S. E. Social support: A review. In The Oxford Handbook of Health Psychology 189–214 (Oxford University Press, 2011).
Caplan, G., Killilea, M., & Abrahams, R. B. Support Systems and Mutual Help: Multidisciplinary Explorations. (Grune & Stratton, 1976).
Li, S. & Xu, Q. Family support as a protective factor for attitudes toward social distancing and in preserving positive mental health during the COVID-19 pandemic. J. Health Psychol. 27, 858–867 (2022).
Marroquín, B., Vine, V. & Morgan, R. Mental health during the COVID-19 pandemic: Effects of stay-at-home policies, social distancing behavior, and social resources. Psychiatry Res. 293, 113419 (2020).
Thoits, P. A. Mechanisms linking social ties and support to physical and mental health. J. Health Soc. Behav. 52, 145–161 (2011).
Vermeulen, M. & Mustard, C. Gender differences in job strain, social support at work, and psychological distress. J. Occup. Health Psychol. 5, 428–440 (2000).
Chu, K. A., Schwartz, C., Towner, E., Kasparian, N. A. & Callaghan, B. Parenting under pressure: A mixed-methods investigation of the impact of COVID-19 on family life. J. Affect. Disord. Rep. 5, 100161 (2021).
De Witte, H., Vander Elst, T. & De Cuyper, N. Job insecurity, health and well-being. In Sustainable Working Lives: Managing Work Transitions and Health throughout the Life Course (eds Vuori, J. et al.) 109–128 (Springer, 2015).
Wilson, J. M. et al. Job insecurity and financial concern during the COVID-19 pandemic are associated with worse mental health. J. Occup. Environ. Med. 62, 686–691 (2020).
Lin, N., Ensel, W. M., Simeone, R. S. & Kuo, W. Social support, stressful life events, and illness: A model and an empirical test. J. Health Soc. Behav. 20, 108–119 (1979).
Liu, L., Gou, Z. & Zuo, J. Social support mediates loneliness and depression in elderly people. J. Health Psychol. 21, 750–758 (2016).
Hefner, J. & Eisenberg, D. Social Support and mental health among college students. Am. J. Orthopsychiatry 79, 491–499 (2009).
Shilo, G. & Savaya, R. Effects of family and friend support on LGB youths’ mental health and sexual orientation milestones. Fam. Relat. 60, 318–330 (2011).
Geldart, S. et al. Workplace incivility, psychological distress, and the protective effect of co-worker support. Int. J. Workplace Health Manag. 11, 96–110 (2018).
Hämmig, O. Health and well-being at work: The key role of supervisor support. SSM Popul. Health 3, 393–402 (2017).
Brown, L. F., Kroenke, K., Theobald, D. E., Wu, J. & Tu, W. The association of depression and anxiety with health-related quality of life in cancer patients with depression and/or pain. Psychooncology 19, 734–741 (2010).
Kroenke, K. et al. Association between anxiety, health-related quality of life and functional impairment in primary care patients with chronic pain. Gen. Hosp. Psychiatry 35, 359–365 (2013).
Saris, I. M. J., Aghajani, M., van der Werff, S. J. A., van der Wee, N. J. A. & Penninx, B. W. J. H. Social functioning in patients with depressive and anxiety disorders. Acta Psychiatr. Scand. 136, 352–361 (2017).
Kämpfen, F. et al. Predictors of mental health during the Covid-19 pandemic in the US: Role of economic concerns, health worries and social distancing. PLoS ONE 15, e0241895 (2020).
Randle, M., Eckersley, R. & Miller, L. Societal and personal concerns, their associations with stress, and the implications for progress and the future. Futures 93, 68–79 (2017).
García, E. L., Banegas, J. R., Pérez-Regadera, A. G., Cabrera, R. H. & Rodríguez-Artalejo, F. Social network and health-related quality of life in older adults: A population-based study in Spain. Qual. Life Res. 14, 511–520 (2005).
Stinesen-Kollberg, K., Thorsteinsdottir, T., Wilderäng, U. & Steineck, G. Worry about one’s own children, psychological well-being, and interest in psychosocial intervention. Psychooncology 22, 2117–2123 (2013).
Cooke, R., Barkham, M., Audin, K., Bradley, M. & Davy, J. Student debt and its relation to student mental health. J. Furth. High. Educ. 28, 53–66 (2004).
Gecková, A., van Dijk, J. P., Stewart, R., Groothoff, J. W. & Post, D. Influence of social support on health among gender and socio-economic groups of adolescents. Eur. J. Pub. Health 13, 44–50 (2003).
Hann, D. et al. The influence of social support on depressive symptoms in cancer patients: Age and gender differences. J. Psychosom. Res. 52, 279–283 (2002).
Harkonmäki, K., Rahkonen, O., Martikainen, P., Silventoinen, K. & Lahelma, E. Associations of SF-36 mental health functioning and work and family related factors with intentions to retire early among employees. Occup. Environ. Med. 63, 558–563 (2006).
Norman, S. B. et al. Moral distress in frontline healthcare workers in the initial epicenter of the COVID-19 pandemic in the United States: Relationship to PTSD symptoms, burnout, and psychosocial functioning. Depress. Anxiety 38, 1007–1017 (2021).
Hartung, C. M. & Lefler, E. K. Sex and gender in psychopathology: DSM–5 and beyond. Psychol. Bull. 145, 390–409 (2019).
Johnson, J. L., Greaves, L. & Repta, R. Better science with sex and gender: Facilitating the use of a sex and gender-based analysis in health research. Int. J. Equity Health 8, 14 (2009).
Doyal, L. Sex and gender: The challenges for epidemiologists. Int. J. Health Serv. 33, 569–579 (2003).
Krieger, N. Genders, sexes, and health: what are the connections—and why does it matter?. Int. J. Epidemiol. 32, 652–657 (2003).
Fiori, K. L. & Denckla, C. A. Social support and mental health in middle-aged men and women: A multidimensional approach. J. Aging Health 24, 407–438 (2012).
Flaherty, J. & Richman, J. Gender differences in the perception and utilization of social support: Theoretical perspectives and an empirical test. Soc. Sci. Med. 28, 1221–1228 (1989).
Viertiö, S. et al. Factors contributing to psychological distress in the working population, with a special reference to gender difference. BMC Public Health 21, 611 (2021).
Hirokawa, K. et al. Associations between occupational status, support at work, and salivary cortisol levels. Int. J. Behav. Med. 29, 299–307 (2022).
Hsieh, C.-M. & Tsai, B.-K. Effects of social support on the stress-health relationship: Gender comparison among military personnel. Int. J. Environ. Res. Public Health 16, 1317 (2019).
Milner, A., Krnjacki, L. & LaMontagne, A. D. Age and gender differences in the influence of social support on mental health: A longitudinal fixed-effects analysis using 13 annual waves of the HILDA cohort. Public Health 140, 172–178 (2016).
Walen, H. R. & Lachman, M. E. Social support and strain from partner, family, and friends: Costs and benefits for men and women in adulthood. J. Soc. Pers. Relat. 17, 5–30 (2000).
Geller, P. A. & Hobfoll, S. E. Gender differences in job stress, tedium and social support in the workplace. J. Soc. Pers. Relat. 11, 555–572 (1994).
Plaisier, I. et al. The contribution of working conditions and social support to the onset of depressive and anxiety disorders among male and female employees. Soc. Sci. Med. 64, 401–410 (2007).
Sinokki, M. et al. The association of social support at work and in private life with mental health and antidepressant use: The Health 2000 Study. J. Affect. Disord. 115, 36–45 (2009).
Kim, T. J. & von dem Knesebeck, O. Is an insecure job better for health than having no job at all? A systematic review of studies investigating the health-related risks of both job insecurity and unemployment. BMC Public Health 15, 985 (2015).
Ferrie, J. E. Labour market status, insecurity and health. J. Health Psychol. 2, 373–397 (1997).
Wharton, A. S. & Blair-Loy, M. Long work hours and family life: A cross-national study of employees’ concerns. J. Fam. Issues 27, 415–436 (2006).
Hill, E. J. Work-family facilitation and conflict, working fathers and mothers, work-family stressors and support. J. Fam. Issues 26, 793–819 (2005).
Lee, J. S. et al. Connecting the dots: a comparison of network analysis and exploratory factor analysis to examine psychosocial syndemic indicators among HIV-negative sexual minority men. J. Behav. Med. 43, 1026–1040 (2020).
de Boer, N. S., de Bruin, L. C., Geurts, J. J. G. & Glas, G. The network theory of psychiatric disorders: A critical assessment of the inclusion of environmental factors. Front. Psychol. 12, 623970 (2021).
Galderisi, S. et al. Interplay among psychopathologic variables, personal resources, context-related factors, and real-life functioning in individuals with schizophrenia: A network analysis. JAMA Psychiat. 75, 396–404 (2018).
Wolff, B. et al. Individual-level risk and resilience factors associated with mental health in siblings of individuals with neurodevelopmental conditions: A network analysis. Dev. Neuropsychol. 48, 112–134 (2023).
Sharma, R. & Dale, S. K. Using network analysis to assess the effects of trauma, psychosocial, and socioeconomic factors on health outcomes among black women living with HIV. AIDS Behav. 27, 400–415 (2023).
Girvan, M. & Newman, M. E. J. Community structure in social and biological networks. Proc. Natl. Acad. Sci. 99, 7821–7826 (2002).
Newman, M. E. J. Communities, modules and large-scale structure in networks. Nat. Phys. 8, 25–31 (2012).
McNally, R. J. Can network analysis transform psychopathology?. Behav. Res. Ther. 86, 95–104 (2016).
Borsboom, D. A network theory of mental disorders. World Psychiatry 16, 5–13 (2017).
Beard, C. & Björgvinsson, T. Beyond generalized anxiety disorder: Psychometric properties of the GAD-7 in a heterogeneous psychiatric sample. J. Anxiety Disord. 28, 547–552 (2014).
McElroy, E. et al. Mental health, deprivation, and the neighborhood social environment: A network analysis. Clin. Psychol. Sci. 7, 719–734 (2019).
Zavlis, O. et al. How does the COVID-19 pandemic impact on population mental health? A network analysis of COVID influences on depression, anxiety and traumatic stress in the UK population. Psychol. Med. 52(16), 3825–3833 (2021).
Beard, C. et al. Network analysis of depression and anxiety symptom relations in a psychiatric sample. Psychol. Med. 46, 3359–3369 (2016).
Mullarkey, M. C., Marchetti, I. & Beevers, C. G. Using network analysis to identify central symptoms of adolescent depression. J. Clin. Child Adolesc. Psychol. 48, 656–668 (2019).
Byford, S. The validity and responsiveness of the EQ-5D measure of health-related quality of life in an adolescent population with persistent major depression. J. Ment. Health 22, 101–110 (2013).
König, H.-H. et al. Validity and responsiveness of the EQ-5D in assessing and valuing health status in patients with anxiety disorders. Health Qual. Life Outcomes 8, 47 (2010).
Sobocki, P. et al. Health-related quality of life measured with EQ-5D in patients treated for depression in primary care. Value Health 10, 153–160 (2007).
Supina, A. L., Johnson, J. A., Patten, S. B., Williams, J. V. A. & Maxwell, C. J. The usefulness of the EQ-5D in differentiating among persons with major depressive episode and anxiety. Qual. Life Res. 16, 749–754 (2007).
Wagner, M. et al. Characteristics and course of heart failure stages A-B and determinants of progression—Design and rationale of the STAAB cohort study. Eur. J. Prev. Cardiol. 24, 468–479 (2017).
Eichner, F. A. et al. Seroprävalenz von COVID-19 und psychosoziale Auswirkungen in der Allgemeinbevölkerung: Ergebnisse des STAAB-COVID-One Programms. Gesundheitswesen 83, 965–975 (2021).
Kroenke, K., Spitzer, R. L. & Williams, J. B. W. The PHQ-9: Validity of a brief depression severity measure. J. Gen. Intern. Med. 16, 606–613 (2001).
Spitzer, R. L., Kroenke, K., Williams, J. B. W. & Löwe, B. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch. Intern. Med. 166, 1092–1097 (2006).
Rabin, R. & Charro, F. D. EQ-SD: A measure of health status from the EuroQol Group. Ann. Med. 33, 337–343 (2001).
Haslbeck, J. M. B. & Fried, E. I. How predictable are symptoms in psychopathological networks? A reanalysis of 18 published datasets. Psychol. Med. 47, 2767–2776 (2017).
Haslbeck, J. M. B. & Waldorp, L. J. How well do network models predict observations? On the importance of predictability in network models. Behav. Res. 50, 853–861 (2018).
Haslbeck, J. M. B. & Waldorp, L. J. mgm: Estimating time-varying mixed graphical models in high-dimensional data. J. Stat. Softw. 93, 1–46 (2020).
Epskamp, S., Cramer, A. O. J., Waldorp, L. J., Schmittmann, V. D. & Borsboom, D. qgraph: Network visualizations of relationships in psychometric data. J. Stat. Softw. 48, 1–18 (2012).
Epskamp, S. & Fried, E. I. A tutorial on regularized partial correlation networks. Psychol. Methods 23, 617–634 (2018).
Tibshirani, R. Regression shrinkage and selection via the Lasso. J. R. Stat. Soc. Ser. B (Methodol.) 58, 267–288 (1996).
Foygel, R. & Drton, M. Extended Bayesian information criteria for Gaussian graphical models. In Advances in Neural Information Processing Systems Vol. 23 (Curran Associates, Inc., 2010).
Jones, P. networktools: Tools for Identifying Important Nodes in Networks. R Package, Version 1.5.0. Available online at: https://cran.r-project.org/web/packages/networktools/index.html (2018).
Duek, O., Spiller, T. R., Pietrzak, R. H., Fried, E. I. & Harpaz-Rotem, I. Network analysis of PTSD and depressive symptoms in 158,139 treatment-seeking veterans with PTSD. Depress. Anxiety 38, 554–562 (2021).
Vervaet, M., Puttevils, L., Hoekstra, R. H. A., Fried, E. & Vanderhasselt, M.-A. Transdiagnostic vulnerability factors in eating disorders: A network analysis. Eur. Eat. Disord. Rev. 29, 86–100 (2021).
Opsahl, T., Agneessens, F. & Skvoretz, J. Node centrality in weighted networks: Generalizing degree and shortest paths. Soc. Netw. 32, 245–251 (2010).
Robinaugh, D. J., Millner, A. J. & McNally, R. J. Identifying highly influential nodes in the complicated grief network. J. Abnorm. Psychol. 125, 747–757 (2016).
van Borkulo, C. D. et al. Comparing network structures on three aspects: A permutation test. Psychol. Methods https://doi.org/10.1037/met0000476 (2022).
Epskamp, S., Borsboom, D. & Fried, E. I. Estimating psychological networks and their accuracy: A tutorial paper. Behav Res 50, 195–212 (2018).
Epskamp, S. Psychonetrics: Structural equation modeling andconfirmatory network analysis. R Package, Version 0.11. Available online at: https://cran.r-project.org/web/packages/psychonetrics/index.html (2019).
Hu, L. T. & Bentler, P. M. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equ. Model. 6, 1–55 (1999).
Wang, J., Mann, F., Lloyd-Evans, B., Ma, R. & Johnson, S. Associations between loneliness and perceived social support and outcomes of mental health problems: A systematic review. BMC Psychiatry 18, 156 (2018).
Eby, L. T., Casper, W. J., Lockwood, A., Bordeaux, C. & Brinley, A. Work and family research in IO/OB: Content analysis and review of the literature (1980–2002). J. Vocat. Behav. 66, 124–197 (2005).
Simon, R. W. Twenty years of the sociology of mental health: the continued significance of gender and marital status for emotional well-being. In Sociology of Mental Health: Selected Topics from Forty Years 1970s–2010s (eds Johnson, R. J. et al.) 21–51 (Springer International Publishing, 2014).
Alsharawy, A., Spoon, R., Smith, A. & Ball, S. Gender differences in fear and risk perception during the COVID-19 pandemic. Front. Psychol. 12, 689467 (2021).
Garcia-Larrea, L. & Peyron, R. Pain matrices and neuropathic pain matrices: A review. In Pain Vol. 154 S29–S43 (Elsevier B.V., 2013).
Bedrov, A. & Gable, S. L. Thriving together: The benefits of women’s social ties for physical, psychological and relationship health. Philos. Trans. R. Soc. B Biol. Sci. 378, 20210441 (2022).
Taylor, S. E. et al. Biobehavioral responses to stress in females: Tend-and-befriend, not fight-or-flight. Psychol. Rev. 107, 411–429 (2000).
Ali, S. et al. The association between depression and health-related quality of life in people with type 2 diabetes: A systematic literature review. Diabetes Metab. Res. Rev. 26, 75–89 (2010).
Johnson, E. K., Jones, J. E., Seidenberg, M. & Hermann, B. P. The relative impact of anxiety, depression, and clinical seizure features on health-related quality of life in epilepsy. Epilepsia 45, 544–550 (2004).
Bouzari, Z., Shenas, M. H., Zeinalzadeh, M. & Barat, S. Depression and health related quality of life in breast cancer patients. Am. J. Cancer Res. 4, 43–46 (2011).
Santos, H. Jr., Fried, E. I., Asafu-Adjei, J. & Ruiz, R. J. Network structure of perinatal depressive symptoms in Latinas: Relationship to stress and reproductive biomarkers. Res. Nurs. Health 40, 218–228 (2017).
Bretschneider, J. et al. Time trends in depression prevalence and health-related correlates: Results from population-based surveys in Germany 1997–1999 vs. 2009–2012. BMC Psychiatry 18, 394 (2018).
Erhardt, A. et al. Generalised anxiety and panic symptoms in the German National Cohort (NAKO). World J. Biol. Psychiatry 5, 1–16 (2021).
Streit, F. et al. Lifetime and current depression in the German National Cohort (NAKO). World J. Biol. Psychiatry 0, 1–16 (2021).
Beutel, M. E. et al. Mental health and loneliness in the German general population during the COVID-19 pandemic compared to a representative pre-pandemic assessment. Sci. Rep. 11, 14946 (2021).
Hawes, M. T., Szenczy, A. K., Klein, D. N., Hajcak, G. & Nelson, B. D. Increases in depression and anxiety symptoms in adolescents and young adults during the COVID-19 pandemic. Psychol. Med. 52, 3222–3230 (2022).
Kwong, A. S. F. et al. Mental health before and during the COVID-19 pandemic in two longitudinal UK population cohorts. Br. J. Psychiatry 218, 334–343 (2021).
Ogueji, I. A., Okoloba, M. M. & Demoko Ceccaldi, B. M. Coping strategies of individuals in the United Kingdom during the COVID-19 pandemic. Curr. Psychol. 41, 7493–7499 (2022).
We thank the participants of the STAAB-COVID Program and the city of Würzburg for their support. In addition, we thank the members of the STAAB-COVID study group for their contribution to the study.
Open Access funding enabled and organized by Projekt DEAL. The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article. The STAAB-COVID Program was supported by the Bavarian Ministry of Research and Art [U.7-H4001.1.7/24/2]. The STAAB cohort was supported by the German Ministry of Research and Education within the Comprehensive Heart Failure Center Würzburg (BMBF 01EO1004 and 01EO1504). GH was supported by the German Research Foundation (HE 4566/6-1). The funding source was not involved in the collection, analysis, and interpretation of the data, in the writing of the report, or in the decision to submit the paper. This publication was supported by the Open Access Publication Fund of the University of Würzburg.
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Jürgen Deckert is Investigator in an EU-grant to P1Vital and corecipient of a grant of the Bavarian Secretary of Commerce to BioVariance. The remaining authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Weiß, M., Gründahl, M., Deckert, J. et al. Differential network interactions between psychosocial factors, mental health, and health-related quality of life in women and men. Sci Rep 13, 11642 (2023). https://doi.org/10.1038/s41598-023-38525-8