https://orcid.org/0000-0001-6540-1284
Figures
Abstract
Background
The Brief Resilience Coping Scale (BRCS) is a brief instrument suitable for epidemiological studies. The aims of this paper were to analyze changes in BRCS depending on time, sex, age group, relationship status, as well as risk of poverty, to test the psychometric properties including test-retest reliability and measurement invariance, and to determine associations with psychosocial stress, depressiveness, anxiety, social support, as well as subjective mental and physical health. As the data from this study was collected during the pandemic, an additional sensitivity analysis was performed with pre-pandemic data.
Methods
A longitudinal study of resilience and distress in a large-sized community sample was performed at one pre-pandemic (T0) and three pandemic time points (T1-3). Resilient coping was assessed by the 4-Item short form of the BRCS, distress by the PHQ-9 and GAD-2.
Results
BRCS decreased between the first and the second and increased at the third pandemic time point. The scale had a good internal consistency. Test-retest correlation coefficients ranged from 0.527 to 0.589. Higher resilient coping was found in younger participants, participants not at-risk-of-poverty and in males. Stability was higher in those with a partner, and at-risk-of-poverty. Significant negative associations with psychosocial stress, loneliness, depressiveness, anxiety, social support, as well as subjective and physical health and SES underscored the construct validity.
Conclusion
Overall, findings underscore that resilient coping is a dynamic construct with considerable stability. The scale showed good psychometric properties including test-retest reliability over four months to two years. We found that it is not only important to describe the level of resilient coping, but also its stability.
Citation: Petersen J, Brähler E, Hettich-Damm N, Schepers M, König J, Lackner K, et al. (2024) Psychometric evaluation of the Brief Resilient Coping Scale (BRCS) over the course of the pandemic in a large German general population sample. PLoS ONE 19(8): e0309587. https://doi.org/10.1371/journal.pone.0309587
Editor: Che Matthew Harris, Johns Hopkins Medicine, UNITED STATES OF AMERICA
Received: March 11, 2024; Accepted: August 14, 2024; Published: August 27, 2024
Copyright: © 2024 Petersen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Access to the data in the local database is possible at any time upon request according to the ethics vote. This concept was developed with the local data protection officer and the ethics committee (local ethics committee of the Rhineland-Palatinate Medical Association, German). Interested scientists can make their requests to the Gutenberg COVID-19 Study Steering Committee (E-Mail: info-gcs@unimedizin-mainz.de).
Funding: The study was funded by the European Regional Development Fund and the Ministry of Science and Health of the State of Rhineland-Palatinate (EFRE/REACT-EU, Grant No. 84007232 and No. 84009735); by the ReALity Initiative of the Life Sciences of the Johannes Gutenberg University Mainz for the establishment of a cell bank. It was also funded by the National University Medicine Research Network on Covid-19 (”NaFoUniMedCovid19”, Grant No. 01KX2021) B-FAST for the topics “poor living conditions” and “working conditions” and their association with COVID-19 in the population.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
The Brief Resilience Coping Scale (BRCS) is a measure comprised of 4 items to assess the ability to cope with stress in an adaptive manner [1]. Given its brevity, it is highly suitable for epidemiological studies. In previous studies on the German general population, individuals with strong resilient coping reported reduced distress and somatoform symptoms, even when having sustained adverse childhood experiences [2, 3]. López-Pina et al. [4] suggested that the measure could identify patients with limited resilience, facilitating the initiation of suitable interventions. In the past, the one-dimensional scale demonstrated good internal consistency, test-retest reliability, psychometric robustness and sensitivity to cognitive-behavioural interventions [1, 5].
Regarding stressful events, results from studies performed during the pandemic in Italy, Australia and Thailand indicate that the BRCS is a useful tool to measure resilience during the pandemic [6–8]. It was also found that, during lockdown, the levels of resilient coping were significantly lower compared to normative data [9]. Results from 26 general population surveys link younger age, female sex, poorer financial situation and the presence of medical conditions with low levels of resilient coping during the COVID-19 pandemic [10]. Individuals with high levels of resilience showed higher emotional stability during lockdown and a reduction in perceived stress after lockdown restrictions had been lifted [11]. Lower levels of resilient coping have also repeatedly been found to be associated with higher levels of depression and anxiety after previous trauma exposure or a multiple sclerosis diagnosis [12–15]. Moreover, resilient coping was also found to have a direct effect on the distress and burnout of caregivers, physicians and teachers during the pandemic [16–19].
There is, however, a lack of published data regarding the BRCS’s test-retest reliability in a German sample during the pandemic and over extended periods. Data from the Gutenberg Health Study (GHS) and Gutenberg COVID Study (GCS), collected both before and during the pandemic, therefore offers a unique chance to study the BRCS’ longitudinal test-retest reliability and temporal stability. Based on this data, this paper aims to
- analyse changes in BRCS over the course of the pandemic,
- test the psychometric properties of the BRCS including test-retest reliability and measurement invariance,
- analyse sex, age group, relationship status, as well as poverty status as predictors of change,
- identify associations between the BRCS and variables such as psychosocial stress, depressiveness, anxiety, social support, and subjective mental and physical health in order to assess construct validity.
2. Methods
2.1. Data
We draw the data from our main sample from the Gutenberg COVID-19 Study (GCS). The GCS was initiated after the outbreak of SARS-CoV-2. Respondents who have already participated in the Gutenberg Health Study (GHS, [20]) prior to the beginning of the pandemic, were invited back in order to investigate the epidemiology of COVID-19 within the population. The participants were recruited in the target area of Mainz/Mainz-Bingen by drawing random samples from the resident’s registration office. The sample was stratified by sex, age, and place of residence.
The baseline examination of the GCS took place from October 2020 to April 2021 (T1). The first GCS follow-up was conducted from March 2021 to June 2021 (T2). Finally, the second follow-up occurred from May 2022 to November 2022 (T3). At the study center, participants underwent a computer-assisted personal interview and sampling of biomaterial. In preparation for their appointments, study participants were sent questionnaires in advance.
In total, the GCS sample comprises of N = 8121 individuals from the GHS and an additional N = 2129 newly recruited younger (aged 25–44) individuals for the GCS. For the present study, we included respondents with available data at all three GCS examination time points and complete and valid BRCS data. This left us with a sample of N = 6009 individuals. Sensitivity analysis was performed on a subsample of N = 3414 with valid BRCS data at all four time points, including additional pre-pandemic BRCS data from the 10-year follow-up of the GHS (2017–2020).
We attended the principles of Good Clinical Practice (GCP), Good Epidemiological Practice (GEP), and the ethical guidelines set forth in the Declaration of Helsinki throughout the entire process of the study design, execution, and analysis. We additionally adhered to the requirements outlines in the Federal Data Protection Act. Both the Ethics Committee of the Rhineland-Palatinate Medical Association and the Data Protection Officer of the Johannes Gutenberg-University Medical Center Mainz thoroughly assessed all pertinent documentation for both the Gutenberg Health Study and the Gutenberg COVID-19 Study, granting their approval. Furthermore, the data protection commissioner of Rhineland-Palatinate approved the selection of the sample through citizens’ registration offices.
2.2. Measures
2.2.1. Brief resilient coping scale (BRCS).
Resilient coping was assessed using the four-item Brief Resilience Coping Scale (BRCS, [1, 5]). Respondents were asked to rate their ability to alter difficult situations, to adjust or control one’s reaction, and to find ways to make up for losses in life on a scale from 1 = does not apply at all to 5 = does apply fully. The sum score of the BRCS, therefore, ranges from 4, indicating the lowest, to 20, indicating the highest level of resilient coping.
2.2.2. Psychological variables.
In addition to the BRCS, we also collected data on depressive symptoms using the Patient Health Questionnaire (PHQ-9, [21, 22]), anxiety symptoms using the Generalized Anxiety Disorder Screener (GAD-2, [23–26]). Psychosocial stress was assessed using the PHQ stress module [27, 28], a 10-item subscale of the PHQ that assesses psychosocial stressors such as concerns about one’s health and weight, stress at work, worries about one’s financial situation, low sexual desire, among others on a 4-point scale. Loneliness using the UCLA Loneliness Scale [29–31], and social support using the Brief Social Support Scale (BS-6, [32]).
2.2.3. Health related variables.
Subjective physical and mental health were inquired using a single item each that asked the respondents to rate their health status on a 4-point scale ranging from 1 = very good to 4 = bad.
2.2.4. Sociodemographic variables.
Additionally, we inquired about age, sex, migration background, education, employment (no, irregular, part-time, fulltime current employment), partnership (yes/no) and socioeconomic status (SES) using the index of Lampert and Kroll [33].
Using the definition of relative poverty by the European Union Statistics on Income and Living Conditions (EU-SILC, [34]), we defined respondents at-risk-of-poverty if their net equalized income was under 60% of the median net equalized income of all households. Net equalized income was calculated by dividing the total monthly household income by a weighted household size: The first adult of a household contributed 1.0 unit to the household weight while every additional adult from an age of 14 increased the household weight by 0.5. Every child under the age of 14 increased the household weight by 0.3. The median net equalized income of 2019 was at 1790€, therefore, a net equalized income of <1074€ was identified to be the threshold of being at-risk-of-poverty.
2.3. Statistical procedure
We first took a look at the characteristics of our sample by calculating absolute and relative frequencies for categorical variables and means with standard deviations for continuous variables.
To analyze the changes in BRCS over the course of the pandemic, we estimated the means and standard deviations of the BRCS sum score and the individual items at all three measurement time points. We additionally expressed mean differences using Cohen’s d effect sizes and tested for their significance using t-tests.
To complete the next aim of our study and test the psychometric properties, we began by testing the internal stability of the sum score using Cronbach’s alpha (α) and McDonald’s total omega (ωt). The temporal stability of the sum score and the individual items was estimated using Pearson’s correlation coefficient. Discriminatory power coefficients were calculated using the corrected correlation between the individual item and the total score. In order to test factor validity, confirmatory factor analysis (CFA) determined if the pre-defined general factor model with all four items loading onto one factor was appropriate for the BRCS. We tested this for all three time points separately. Deviations from the normal distribution at all three pandemic time points were corrected using maximum likelihood estimation. This is a robust method in dealing with violations of normality [35]. The goodness-of-fit was determined using the following four criteria: the Comparative Fit Index (CFI), the Tucker Lewis Index (TLI), the Standardized Root Mean Squared Residual (SRMR), as well as the Root Mean Square Error of Approximation (RMSEA). A RMSEA and SRMR value of < 0.050 indicates a good fit while a value between 0.050 and 0.080 indicates a reasonable fit. For CFI and TLI, a value of > 0.950 demonstrates a good fit of a model [36, 37]. Subsequently, we tested for measurement invariance across the time points. We used the guidance of Meredith and Teresi’s [38] sequential strategy. Within this framework, we evaluated configural, weak, strong, and strict invariance by introducing more restrictions as the invariance models progress. Configural invariance assumes that there might be variations in the loadings, intercepts, and variances of the latent construct among the different groups or time points. Therefore, this model is completely unrestricted. Loadings are constrained to be equal across groups or time points if weak measurement invariance is assumed. To test for strong measurement invariance, we constrained both factor loadings and item intercepts to be equal across time points. Finally, for strict measurement invariance we assumed factor loadings, item intercepts and residual variances to be equal across time points. All invariance models were then compared to the more stringent model. The most commonly applied method for testing a model fit is the chi-square-test. However, due to its affectability by large sample sizes, which potentially leads to the false rejection of reasonable models, we relied on the abovementioned four fit indices to compare the model fits. Full invariance can be assumed if the CFI difference between all invariance models does not surpass the threshold of ΔCFI ≥ 0.010, as proposed by Chen [39–41].
Next, we looked at differences in changes in resilient coping between different sociodemographic groups by calculating the means of each group and testing the significance of their mean differences compared to the other group using t-tests. We also checked if the temporal stability differed across groups.
The fourth goal of our study was to further test for associations of the BRCS with relevant sociodemographic and psychological factors by using Pearson’s correlation coefficients as well as by testing for mean score differences between groups. Age groups (< 60 vs. ≥ 60) were formed in such a way that they approximately divide the sample in half.
Finally, we performed a sensitivity analysis and checked for significant pandemic-related differences by checking for significant changes in resilient coping since the start of the pandemic compared to a pre-pandemic measurement time point. We additionally assessed test-retest stability between pre-pandemic and pandemic time points using Pearson correlations. All analyses were conducted using R version 4.1.1 (packages: psych [42], lsr [43], lavaan [44], semTools [45]).
3. Results
3.1. Sample characteristics
Of the N = 10250 respondents at the GCS baseline examination (T1), N = 6009 provided valid and complete BRCS data for all three pandemic time points. Sociodemographic characteristics of the sample are shown in Table 1. The sample consisted of N = 3110 (51.8%) female and N = 2899 (48.2%) male respondents with a mean age of 56.57 (SD = 14.70).
3.2. BRCS mean scores and item characteristics
The BRCS score yielded good reliability at all three examination time points with Cronbach’s alpha ranging from 0.76 to 0.78 and McDonald’s omega ranging from 0.79 to 0.81, as can be seen in Table 2. The mean score of the sample decreased slightly from 14.77 (SD = 3.11) to 14.45 (SD = 3.25) between the first and second measurement time point and subsequently increased again at the third time point (M = 14.82, SD = 3.19). These differences had very low effect sizes of d = 0.10 between T1 and T2, d = 0.11 between T2 and T3, and d = 0.01 between T1 and T3, respectively. However, the differences between the time points were statistically significant (p < 0.001) with the exception of the difference between T1 and T3.
All items contributed to the decrease between T1 and T2 as well as to the increase between T2 and T3 with effect sizes ranging from d = 0.04 (looking for ways to replace the losses in life) to d = 0.13 (looking for creative ways to alter difficult situations). Additionally, all items contributed to the BRCS total score at all three time points with discrimination power coefficients ranging from 0.44 to 0.68.
The test-retest correlations of the total BRCS score were 0.589 between T1 and T2, 0.539 between T2 and T3, and 0.527 between T1 and T3, while the test-retest correlations of the single items ranged from 0.407 to 0.496.
3.3. Measurement invariance
The results of the confirmatory factor analysis (CFA) and measurement testing are presented in Table 3. When analyzed separately, the BRCS data of all three time points yielded good model fits. The CFI, in particular, indicated a very good fit (T1: 0.988, T2: 0.991, T3: 0.990). We saw that the fits of the model did not change significantly by introducing further restraints. Furthermore, the difference in CFI did not exceed the threshold of a change of more than 0.01, indicating that full measurement invariance was able to be established.
3.4. Associations with sociodemographic and other psychological factors
In Table 4 we compared the means, as well as changes in BRCS between different sociodemographic groups. We found significant mean differences for all time points between people over and under 60. There were also significant mean differences between men and women as well as for being at-risk-of-poverty and not, but only for the first two time-points. We did not find any significant mean difference between people with a partner and single people. In regards to changes in score between the time points T1 and T2, we found a significant difference only for respondents under the age of 60, whose score decreased significantly less between T1 and T2. For the changes in score between T2 and T3, we observed that the score of women and people under the age of 60 increased significantly more than those of men and people over the age of 60. We also found that the temporal stability in BRCS between men and women differs: The temporal stability amongst women was higher than among men (Women T1-T2: 0.601 vs. Men T1-T2: 0.574; Women T2-T3: 0.553 vs. Men T2-T3: 0.524). There were also differences between age groups with the temporal stability among the older group being lower across both tests. For partnership and poverty status, we observed differences for the T1-T2 test-retest only with the group of people with a partner and the group of people at-risk-of-poverty demonstrating higher levels of temporal stability.
Subsequently, we evaluated the associations between the BRCS sum score at all three measurement time points with relevant sociodemographic and psychological factors (see Table 5). Depressiveness, anxiety, psychosocial stress, loneliness, age, subjective mental and physical health were found to have significant negative associations with the BRCS across all three time points with Pearson’s correlation coefficients ranging from -0.084 (PHQ stress) to -0.204 (subjective mental health). Social support and SES consistently demonstrated a significant positive association with the BRCS; Pearson’s correlation coefficients ranged from 0.153 (BS-6) to 0.207 (SES). Being female and being at-risk-of-poverty were found to have smaller significant negative associations with the BRCS at the first two measurement time points, but not at the third. Having a partner or a migration background was not associated with the BRCS at any time point.
3.5. Sensitivity analysis
As a next step, we tested the score’s sensitivity to stressful events such as the onset of a pandemic. For this, we drew additional data from a pre-pandemic measurement time point and tested the mean difference as well as performed a test-retest correlation with the three pandemic time points. The results are shown in Table 6. Compared to the pre-pandemic time point, which yielded a mean score of 15.46 (SD = 3.03), we found significant (p < 0.001) decreases in BRCS at all three pandemic time points with means ranging from 14.48 (SD = 3.31) to 14.80 (SD = 3.17). The testing for test-retest reliability yielded good results with correlation coefficients ranging from 0.471 to 0.528.
4. Discussion
The COVID-19 pandemic and the measures put in place to contain its spread have comprised novel mental and physical health threats, pervasive societal and daily-life restrictions, as well as changes of social and work life on an unprecedented, world-wide basis over an extended period of time. This allows us to gain unique insights into the role of resilience. As a first aim of this study, we, therefore, analyzed the changes of resilient coping over the three pandemic time points. We found a significant decrease between the first and the second time points and a significant increase to the third time point. However, effect sizes of these changes were small (d = 0.10 between T1 and T2, d = 0.11 between T2 and T3). Compared to data from BRCS validation papers, the means of our data (T1: 14.77 (SD = 3.11), T2: 14.45 (SD = 3.25), T3: 14.82 (SD = 3.19) corresponded to other community samples, with Sinclair and Wallston [1] reporting an overall mean of 14.81 (SD = 2.95) and Kocalevent et al. [5] reporting a mean of 14.9 (SD = 3.3) for men and a mean of 14.6 (SD = 3.1) for women. We do, however, find a dip in the BRCS score at the second measurement point. This may be explained by the tightening of COVID restrictions in April of 2021 when the German government invoked an emergency brake by reducing social contacts and by putting a nighttime curfew into place. This happened following a prior loosening of lockdown measures earlier in the year which was highly controversial [46]. Similarly, Gullo et al. [9] reported decreases in BRCS during lockdown in the first half of 2020 compared to normative data. Following this dip of resilient coping, scores went up to the previous levels in 2022.
As the next step we tested the psychometric properties of the BRCS. The scale was found to have a good internal consistency with Cronbach’s alpha ranging from 0.76 to 0.78 and McDonald’s omega ranging from 0.79 to 0.81. These values are slightly higher than those reported by Sinclair and Wallston [1] as well as Kocalevent et al. [5], whose alphas ranged from 0.64 to 0.76. The test-retest correlation coefficients of our study ranged from 0.527 to 0.589, which are lower than the values reported by Sinclair and Wallston [14] (r = 0.71 after 5 weeks and r = 0.68 after three months), but are similar to those of another study of a German sample during the same time period with r = 0.60 [47]. As more time elapsed in our study (from four months to two years), consistent with previous studies we would presume that temporal stability decreases as time between measurement points increases [48, 49]. We additionally tested the measurement invariance across all three pandemic time points, which–to our knowledge–has not yet been done with the BRCS, and were able to establish full invariance. This indicates that the instrument is being interpreted the same way at all three time points, allowing for a fair comparison of the means over time.
When we associated the BRCS differences with sex, age group (< 60 vs. ≥ 60), partnership status (with vs. without a partner), as well as poverty (being vs. not being at-risk-of-poverty), we found significant mean differences for all time points between people over and under 60 with higher resilient coping in the younger age group. Significant effects were found for sex and risk-of-poverty for the first two time points. Corresponding to Wong et al. [10], female sex and risk-of-poverty were associated with lower resilient coping. In Wong et al.’s study, however, it was found that younger age was associated with lower levels of resilient coping, which contradicts our findings. This might be due to the fact that our sample is relatively old with a mean age of 56.6 (SD = 14.7) years; younger people, who were more affected by distress (and a negative impact to their resilient coping skills) during the pandemic [50, 51], were underrepresented in our sample. If our sample had included more younger people with symptoms of distress, we might have come to different results.
The temporal stability among women and participants under the age of 60 was higher than among men and people over 60 years. Furthermore, we found higher temporal stability (T1-T2 test-retest coefficients) for participants with (vs. without) a partner and those at-risk-of-poverty. The lowest temporal stability, however, was found for men (rtt t3-t2 = 0.524). These findings show that not only is it important to describe the level of resilient coping, but also its stability. For example, high stability in those at-risk-of-poverty could mean that it is hard for economically disadvantaged to sustain resilient coping under stressful conditions. Previously, we were able to find that those at-risk-of-poverty indeed suffered more COVID-19 related economic disadvantages compared to those who were better off [52]. High stability during stressful times might make it harder for these population groups to ‘bounce back’ after major economic setbacks. On the other hand, a partnership may also enhance stability of resilient coping, regardless of its level.
As we had expected, depressiveness, anxiety, psychosocial stress, loneliness, subjective mental and physical health had significant negative associations with the BRCS across all three time points, while social support and SES consistently demonstrated a significant positive association with the BRCS. These findings indicate construct validity and are in line with results from previous studies [10, 12–19].
Subgroup analysis confirmed significantly lower BRCS levels (means ranged from 14.48 to 14.80) during all three pandemic time points compared to a pre-pandemic measurement point (M = 15.46, SD = 3.03). This finding is in accordance with Gullo et al. [9], who found BRCS levels during the pandemic to be lower compared to (non-pandemic) normative data. This finding indicates that the BRCS in the general is sensitive to large-scale external stressors. Though, it must be noted that our pre-pandemic data yields signficantly higher BRCS mean scores than those reported in previous validation papers. Our BRCS score from our pandemic data, however, as mentioned above, were more aligned with those reported in these papers. This raises the question if the GHS and GCS samples are generally more resilient than other samples.
4.1. Limitations
While we have a unique data base over an extended time period before and during the pandemic, the community sample was relatively old with a mean age of 56.6 years and financially better off with only 3.4% people being at-risk-of-poverty compared to the 14.7% reported by the German Federal Statistical Office [53], which limits the representativeness of our results and, therefore, the mean scores should be interpreted with caution.
Additionally, even though 10250 people participated at the GCS baseline examination (T1), only N = 6009 participated in all three measurements with valid data. This raises the question of who drops out of a longitudinally study. It cannot be precluded that unwell people with low resilience scores become unfit for these types of examinations, which might also explain why our BRCS score are so high compared to normative data. While effect sizes of changes of resilient coping were small, these may still be meaningful in the general population as we could show that resilience varies with sociodemographic characteristics. Finally, some of the study took place during the roll-out of vaccinations. The access to vaccines and the subsequent loosening of lockdown measures might have had an effect on the BRCS.
5. Conclusion
Resilience is a dynamic construct. Indeed, the BRCS demonstrated a good temporal stability and proved to be a reliable instrument for longitudinal studies. Furthermore, we found that the score is sufficiently sensitive to external stressors such as a global pandemic, rendering it to be a valuable instrument for implementation in future studies on resilient coping. Finally, sociodemographic factors such as sex, age, and poverty were significantly associated with changes in BRCS, therefore providing more insight on factors influencing changes over time.
Acknowledgments
We are indebted to all study participants of the Gutenberg COVID-19 study and the Gutenberg Health Study for their contribution. We thank all study staff and researchers involved in the planning and conduct of the study.
References
- 1. Sinclair VG, Wallston KA. The Development and Psychometric Evaluation of the Brief Resilient Coping Scale. Assessment 2004;11:94–101. pmid:14994958
- 2. Beutel ME, Tibubos AN, Klein EM, Schmutzer G, Reiner I, Kocalevent R-D, et al. Childhood adversities and distress—The role of resilience in a representative sample. PLOS ONE 2017;12:e0173826. pmid:28296938
- 3. Häuser W, Brähler E, Schmutzer G, Glaesmer H. The association of adverse childhood experiences and of resilience with chronic noncancer pain in the German adult population—A cross-sectional survey. Eur J Pain 2019;23:555–64. pmid:30318676
- 4. López-Pina J-A, Meseguer-Henarejos A-B, Gascón-Cánovas J-J, Navarro-Villalba D-J, Sinclair VG, Wallston KA. Measurement properties of the brief resilient coping scale in patients with systemic lupus erythematosus using rasch analysis. Health Qual Life Outcomes 2016;14:128. pmid:27619635
- 5. Kocalevent R-D, Zenger M, Hinz A, Klapp B, Brähler E. Resilient coping in the general population: standardization of the brief resilient coping scale (BRCS). Health Qual Life Outcomes 2017;15:251. pmid:29282066
- 6. Murphy M, Lami A, Moret-Tatay C. An Italian Adaptation of the Brief Resilient Coping Scale (BRCS) and Attitudes During the Covid-19 Outbreak. Front Psychol 2021;12:641213. pmid:34276468
- 7. Nochaiwong S, Ruengorn C, Awiphan R, Phosuya C, Ruanta Y, Kanjanarat P, et al. Transcultural adaptation and psychometric validation of the Thai-Brief Resilient Coping Scale: a cross-sectional study during the coronavirus disease 2019 pandemic in Thailand. Sci Rep 2022;12:21521. pmid:36513707
- 8. Rahman MA, Salehin M, Islam SMS, Alif SM, Sultana F, Sharif A, et al. Reliability of the tools used to examine psychological distress, fear of COVID‐19 and coping amongst migrants and non‐migrants in Australia. Int J Ment Health Nurs 2021;30:747–58. pmid:33555128
- 9. Gullo S, Misici I, Teti A, Liuzzi M, Chiara E. Going through the lockdown: a longitudinal study on the psychological consequences of the coronavirus pandemic. Res Psychother Psychopathol Process Outcome 2021;23. pmid:33585300
- 10. Wong MCS, Huang J, Wang HHX, Yuan J, Xu W, Zheng Z-J, et al. Resilience level and its association with maladaptive coping behaviours in the COVID-19 pandemic: a global survey of the general populations. Glob Health 2023;19:1. pmid:36597129
- 11. Osimo SA, Aiello M, Gentili C, Ionta S, Cecchetto C. The Influence of Personality, Resilience, and Alexithymia on Mental Health During COVID-19 Pandemic. Front Psychol 2021;12:630751. pmid:33716896
- 12. Ng R, Ang RP, Ho M-HR. Coping with Anxiety, Depression, Anger and Aggression: The Mediational Role of Resilience in Adolescents. Child Youth Care Forum 2012;41:529–46. https://doi.org/10.1007/s10566-012-9182-x.
- 13. Roma P, Monaro M, Colasanti M, Ricci E, Biondi S, Di Domenico A, et al. A 2-Month Follow-Up Study of Psychological Distress among Italian People during the COVID-19 Lockdown. Int J Environ Res Public Health 2020;17:8180. pmid:33167479
- 14. Sinclair VG, Wallston KA, Strachan E. Resilient Coping Moderates the Effect of Trauma Exposure on Depression. Res Nurs Health 2016;39:244–52. pmid:27176758
- 15. Tan-Kristanto S, Kiropoulos LA. Resilience, self-efficacy, coping styles and depressive and anxiety symptoms in those newly diagnosed with multiple sclerosis. Psychol Health Med 2015;20:635–45. pmid:25588098
- 16. Friesen KA, Weiss JA, Howe SJ, Kerns CM, McMorris CA. Mental Health and Resilient Coping in Caregivers of Autistic Individuals during the COVID-19 Pandemic: Findings from the Families Facing COVID Study. J Autism Dev Disord 2022;52:3027–37. pmid:34240291
- 17. Keim R, Pfitscher G, Leitner S, Burger K, Giacomoni F, Wiedermann CJ. Teachers’ emotional well-being during the SARS-CoV-2 pandemic with long school closures: a large-scale cross-sectional survey in Northern Italy. Public Health 2022;208:1–8. pmid:35659680
- 18. Khalaf OO, Khalil MA, Abdelmaksoud R. Coping with depression and anxiety in Egyptian physicians during COVID-19 pandemic. Middle East Curr Psychiatry 2020;27:63. https://doi.org/10.1186/s43045-020-00070-9.
- 19. Vilovic T, Bozic J, Zuzic Furlan S, Vilovic M, Kumric M, Martinovic D, et al. Mental Health Well-Being and Attitudes on Mental Health Disorders among Family Physicians during COVID-19 Pandemic: A Connection with Resilience and Healthy Lifestyle. J Clin Med 2022;11:438. pmid:35054130
- 20. Wild PS, Zeller T, Beutel M, Blettner M, Dugi KA, Lackner KJ, et al. Die Gutenberg Gesundheitsstudie. Bundesgesundheitsblatt—Gesundheitsforschung—Gesundheitsschutz 2012;55:824–30. pmid:22736163
- 21. Kocalevent R-D, Hinz A, Brähler E. Standardization of the depression screener Patient Health Questionnaire (PHQ-9) in the general population. Gen Hosp Psychiatry 2013;35:551–5. pmid:23664569
- 22. Löwe B, Unützer J, Callahan CM, Perkins AJ, Kroenke K. Monitoring Depression Treatment Outcomes With the Patient Health Questionnaire-9: Med Care 2004;42:1194–201. https://doi.org/10.1097/00005650-200412000-00006.
- 23. Kroenke K, Spitzer RL, Williams JBW, Monahan PO, Löwe B. Anxiety Disorders in Primary Care: Prevalence, Impairment, Comorbidity, and Detection. Ann Intern Med 2007;146:317. pmid:17339617
- 24. Löwe B, Wahl I, Rose M, Spitzer C, Glaesmer H, Wingenfeld K, et al. A 4-item measure of depression and anxiety: Validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population. J Affect Disord 2010;122:86–95. pmid:19616305
- 25. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Arch Intern Med 2006;166:1092. pmid:16717171
- 26. Wicke FS, Krakau L, Löwe B, Beutel ME, Brähler E. Update of the standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population. J Affect Disord 2022;312:310–4. pmid:35760191
- 27. Gräfe K, Zipfel S, Herzog W, Löwe B. Screening psychischer Störungen mit dem “Gesundheitsfragebogen für Patienten (PHQ-D)“. Diagnostica 2004;50:171–81. https://doi.org/10.1026/0012-1924.50.4.171.
- 28. Petrowski K, Schmalbach B, Tibubos A, Brähler E, Löwe B. Psychometric evaluation of the patient health questionnaire stress scale. J Affect Disord 2024;357:37–41. pmid:38657765
- 29. Hughes ME, Waite LJ, Hawkley LC, Cacioppo JT. A Short Scale for Measuring Loneliness in Large Surveys: Results From Two Population-Based Studies. Res Aging 2004;26:655–72. pmid:18504506
- 30. Klein EM, Zenger M, Tibubos AN, Ernst M, Reiner I, Schmalbach B, et al. Loneliness and its relation to mental health in the general population: Validation and norm values of a brief measure. J Affect Disord Rep 2021;4:100120. https://doi.org/10.1016/j.jadr.2021.100120.
- 31. Russell D, Peplau LA, Cutrona CE. The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. J Pers Soc Psychol 1980;39:472–80. pmid:7431205
- 32. Beutel ME, Brähler E, Wiltink J, Michal M, Klein EM, Jünger C, et al. Emotional and tangible social support in a German population-based sample: Development and validation of the Brief Social Support Scale (BS6). PLOS ONE 2017;12:e0186516. pmid:29023540
- 33. Lampert T, Kroll LE, Müters S, Stolzenberg H. Messung des sozioökonomischen Status in der Studie „Gesundheit in Deutschland aktuell”(GEDA). Bundesgesundheitsblatt—Gesundheitsforschung—Gesundheitsschutz 2013;56:131–43. https://doi.org/10.1007/s00103-012-1583-3.
- 34.
Eurostat. EU statistics on income and living conditions (EU-SILC) methodology—monetary poverty n.d. https://ec.europa.eu/eurostat/statistics-explained/index.php?title=EU_statistics_on_income_and_living_conditions_(EU-SILC)_methodology_-_monetary_poverty#Description (accessed May 25, 2022).
- 35. Chou C, Bentler PM, Satorra A. Scaled test statistics and robust standard errors for non‐normal data in covariance structure analysis: A Monte Carlo study. Br J Math Stat Psychol 1991;44:347–57. pmid:1772802
- 36. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model Multidiscip J 1999;6:1–55. https://doi.org/10.1080/10705519909540118.
- 37. Van De Schoot R, Lugtig P, Hox J. A checklist for testing measurement invariance. Eur J Dev Psychol 2012;9:486–92. https://doi.org/10.1080/17405629.2012.686740.
- 38. Meredith W, Teresi JA. An Essay on Measurement and Factorial Invariance: Med Care 2006;44:S69–77. https://doi.org/10.1097/01.mlr.0000245438.73837.89.
- 39. Chen FF. Sensitivity of Goodness of Fit Indexes to Lack of Measurement Invariance. Struct Equ Model Multidiscip J 2007;14:464–504. https://doi.org/10.1080/10705510701301834.
- 40. Cheung GW, Rensvold RB. Evaluating Goodness-of-Fit Indexes for Testing Measurement Invariance. Struct Equ Model Multidiscip J 2002;9:233–55. https://doi.org/10.1207/S15328007SEM0902_5.
- 41. Hirschfeld G, von Brachel R. Improving Multiple-Group confirmatory factor analysis in R–A tutorial in measurement invariance with continuous and ordinal indicators 2014. https://doi.org/10.7275/QAZY-2946.
- 42. Revelle W. psych: Procedures for Psychological, Psychometric, and Personality Research. R package version 2.3.9 2023.
- 43. Navarro D. Learning statistics with R: A tutorial for psychology students and other beginners. (Version 0.6). R package version 0.5.1 2015.
- 44. Rosseel Y. lavaan: An R Package for Structural Equation Modeling. J Stat Softw 2012;48. https://doi.org/10.18637/jss.v048.i02.
- 45. Jorgensen TD, Pornprasertmanit S, Schoemann AS, Rosseel Y. semTools: Useful tools for structural equation modeling. R package version 0.5–6. 2022.
- 46. German Federal Institute for Risk Assessment. Nighttime curfew met with mixed acceptance. 2021.
- 47. Köhne S, Engert V, Rosendahl J. Stability of resilience in times of the COVID‐19 pandemic. Personal Ment Health 2023;17:55–66. pmid:35905979
- 48. Hinz A, Esser P, Friedrich M, Glaesmer H, Mehnert-Theuerkauf A, Schroeter ML, et al. Changes in anxiety in the general population over a six-year period. PLOS ONE 2023;18:e0291206. pmid:37699040
- 49. Struijs SY, Lamers F, Verdam MGE, Van Ballegooijen W, Spinhoven P, Van Der Does W, et al. Temporal stability of symptoms of affective disorders, cognitive vulnerability and personality over time. J Affect Disord 2020;260:77–83. pmid:31493643
- 50. Bu F, Steptoe A, Fancourt D. Who is lonely in lockdown? Cross-cohort analyses of predictors of loneliness before and during the COVID-19 pandemic. Public Health 2020;186:31–4. pmid:32768621
- 51. Wang Y, Kala MP, Jafar TH. Factors associated with psychological distress during the coronavirus disease 2019 (COVID-19) pandemic on the predominantly general population: A systematic review and meta-analysis. PLOS ONE 2020;15:e0244630. pmid:33370404
- 52. Petersen J, Hettich N, Baumkötter R, Wild PS, Pfeiffer N, Münzel T, et al. The burdens of poverty during the COVID-19 pandemic. Front Sociol 2022;7:995318. pmid:36505762
- 53. Federal Statistical Office. Gut ein Fünftel der Bevölkerung Deutschlands von Armut oder sozialer Ausgrenzung bedroht 2023.