Abstract
Given the potentially long-standing emotional impact of breast cancer, theoretical models are needed to identify critical resources and coping strategies that optimize survivors’ long-term adjustment. This study tested a model of well-being recovery with breast cancer survivors at 4 years post-treatment. Structural equation modeling was used to examine relationships between affect, loneliness, self-compassion, self-efficacy for coping with cancer, well-being, and life satisfaction. Two model variations offered good fit to the data, accounting for large portions of the variance in well-being and life satisfaction. Coping efficacy mediated paths from affect and loneliness to emotional well-being and life satisfaction.
One in eight women will experience breast cancer in her lifetime (American Cancer Society, 2015), and the emotional impact of this diagnosis may have long-standing effects. Earle (2006) found that breast cancer survivors were more likely to report depression or anxiety 5 years post-treatment compared to matched controls. A recent meta-analysis supported these long-term consequences for breast cancer survivors’ levels of depression, with a higher prevalence of depressive symptoms persistent more than 5 years after diagnosis (Maas et al., 2015). Social cognitive frameworks have identified factors such as personality characteristics (e.g. optimism), type of coping (Teo et al., 2016; Wen K-Y et al., 2017), and social support as contributors to post-treatment levels of well-being (Hoffman et al., 2013). Given variability in how breast cancer survivors are affected by their cancer experience (Przezdziecki et al., 2013), there is a need for strength-based approaches that identify pathways, critical resources, and coping methods that optimize survivors’ long-term psychosocial adjustment, regardless of their distress level.
A restorative model of well-being
Identifying determinants of positive adjustment is an important step in understanding how best to foster the adaptation of breast cancer survivors (Mols et al., 2009; Wen K-Y et al., 2017). A long-term perspective on understanding this population’s well-being is needed given the ongoing concerns with recurrence, body image, accessing social support after medical treatment (Williams and Jeanetta, 2016), and coping with the long-term side effects of treatment (Hall et al., 2014). Lent (2004, 2007) proposed a unifying framework of well-being recovery that may be applied to the process of psychological adjustment in breast cancer survivors. The model (see Supplementary Figure 1 available at: http://hpq.sagepub.com/) highlights the interplay among trait, social, cognitive, and behavioral variables as they jointly promote well-being after adverse life experiences. Building on the normative model of well-being (Lent, 2004) and stress response theory (Lazarus and Folkman, 1984), the restorative model of well-being proposes that coping is initiated when individuals appraise the level of threat posed by a serious life event (e.g. cancer diagnosis), and the extent to which they believe they have the internal resources to cope with the event. Additionally, environmental supports and resources, such as social relationships, influence the coping strategy utilized (Supplementary Figure 1, path 12 available at: http://hpq.sagepub.com/). Environmental supports (or lack thereof) and selected coping strategies then predict self-efficacy for coping with cancer (paths 13 and 14). Theoretically, those who have adequate levels of support and have identified effective strategies for managing stress inherent in the experience of a breast cancer diagnosis will have increased confidence in their coping ability. Self-efficacy, along with these other factors, is expected to influence the degree to which survivors feel satisfied (or are able to restore their pre-cancer level of satisfaction) in different life domains (e.g. physical, emotional, work, and social; paths 15–17). Experiencing a high level of satisfaction in central life domains then contributes to greater sense of overall life satisfaction (path 20).
Personality and affective variables are also hypothesized to play several roles relative to domain-specific and overall life satisfaction. Along with their direct effect on both domain-specific well-being and overall life satisfaction (paths 8–11), personality and affective factors may affect breast cancer survivors’ assessment of the threat posed by their cancer experience, as well as selection of coping strategies (paths 4 and 5). Personality and affective factors may also influence emotional recovery via other cognitive and behavioral pathways (Lent, 2007). For example, those with higher levels of positive affect are more likely to solicit social support effectively and to view their ability to cope with cancer more favorably (paths 2, 3, 6, and 7). In short, traits interact with more malleable social and cognitive determinants of life satisfaction. The current study assessed personality in terms of trait positive and negative affect (which were assumed to covary, path 1), sufficiency of social support in terms of experienced loneliness, coping strategy in terms of self-compassion, coping efficacy in terms of the perceived ability to negotiate the cancer experience, and domain well-being as level of satisfaction with how well one is coping emotionally with cancer. We briefly review each of these constructs in the context of cancer coping, below.
Positive and negative affect
Emotional disposition may play important roles when facing long-standing adversity such as a cancer diagnosis (Stanton and Low, 2012). For example, positive and negative affect may influence the process through which breast cancer survivors select coping strategies, evaluate the efficacy of the selected strategies, and decide whether to reach out to others for support. In a sample of breast cancer survivors, positive affect was found to relate positively to life satisfaction, happiness, mental health functioning, and physical functioning, whereas negative affect was inversely related to these criteria (Tessier et al., 2012).
Loneliness
Gierveld et al. (2006) observed that cancer survivors reporting adequate levels of social support may still feel lonely; thus, individuals may have a considerable number of relationships in their lives but not feel able to draw on those relationships as a source of support. Although social support is related to well-being in cancer survivors (McDonough et al., 2014), support alone may not convey the extent to which survivors feel connected to others. In a study of 200 breast cancer survivors who were between 2 months and 3 years post-treatment, Jaremka et al. (2013) found that lonelier participants reported significantly more depression, pain, and fatigue compared to those who reported greater feelings of social connection. Loneliness has been identified as a long-term concern of breast cancer survivors after treatment has ended (Rosedale, 2009) and has been associated with depression for breast cancer survivors on average more than 5 years post-diagnosis (Cleary and Stanton, 2015).
Self-compassion
Self-compassion entails the ability to be kind to oneself when experiencing life difficulties (self-kindness), to be mindfully present with painful feelings (mindfulness), and to feel connected to others rather than isolated when experiencing pain (common humanity; Neff, 2003). Self-compassion is assumed to help individuals regulate negative emotional experiences that arise from uncontrollable events (Leary et al., 2007). Sirois et al. (2015) found that in a sample of women with inflammatory bowel disease and arthritis, self-compassion predicted coping styles and coping self-efficacy, which in turn predicted lower stress. Self-compassion also has been associated with health-promoting behaviors, which in turn have been associated with better physical health (Dunne et al., 2018). Among breast cancer survivors diagnosed 5 years previously, higher levels of self-compassion were related to lower levels of body image disturbance, depression, anxiety, and stress (Przezdziecki et al., 2013). Przezdziecki et al. concluded that breast cancer survivors’ levels of self-compassion warrant monitoring by psycho-oncology professionals and that mediating mechanisms between self-compassion and distress merit study.
Self-efficacy for coping with cancer
Coping self-efficacy offers a potential pathway by which self-compassion may promote coping in those dealing with chronic illness (Sirois et al., 2015). Within the context of breast cancer, coping self-efficacy refers to survivors’ confidence in their ability to manage their cancer experience. Coping self-efficacy has been associated with higher relationship satisfaction and self-esteem, less functional impairment (Manne et al., 2006), higher quality of life, and lower levels of negative emotions (Heitzmann et al., 2011). In a study of cancer survivors, an average 9 years post-treatment, coping self-efficacy accounted for 43 percent of the variance in depression scores and partially mediated the relation between cancer symptom impact and depression (Philip et al., 2013). Thus, coping self-efficacy may represent a relatively malleable pathway to improving the lives of cancer survivors long after treatment has ended.
Domain and life satisfaction
Because cancer survivors have indicated greater difficulty managing emotional needs compared to other life domains post-treatment (Wolff, 2007), restoring emotional well-being may be critical to enhancing their life satisfaction. Because emotional well-being has not been examined as a specific domain of well-being that may contribute to an overall sense of life satisfaction among breast cancer survivors, both emotional well-being and life satisfaction could be examined as relevant outcome variables, with emotional well-being also serving to mediate the relation of other predictors to life satisfaction. There is a need to study the interplay among psychosocial variables that predict life satisfaction in the aftermath of cancer treatment, and the restorative model of well-being may be useful in accounting for this process.
Purpose of the present study
The present study assesses the applicability of the restorative model of well-being with breast cancer survivors, addressing the need for comprehensive models for this population that account for the interplay among social cognitive variables. In adapting the model to the context of coping with breast cancer, we tested the paths described in Supplementary Figure 1 http://hpq.sagepub.com/, using loneliness to index the relative sufficiency or absence of social support, and self-compassion to reflect a coping strategy that has been highlighted recently for its relevance to breast cancer survivors. In addition to the basic structural model, we tested an alternative based on the recent findings of Ezeofor (2015), who observed that model fit could be improved by adding paths from coping strategies and coping efficacy directly to life satisfaction (see the dashed paths in Supplementary Figure 1 available at: http://hpq.sagepub.com/). It is reasonable to assume that when people are faced with especially daunting challenges, having effective coping strategies and confidence in their coping capabilities may contribute directly to their life satisfaction.
Methods
Participants and procedures
The sample included 275 female breast cancer survivors over the age of 18 years who were not currently receiving chemotherapy or radiation. Participants reported a mean age of 47 years at the time of the study (ranging from 24 to 86 years, standard deviation (SD) = 11.12 years) and a mean age of 43 years when receiving their diagnosis (ranging from 20 to 86 years, SD = 10.42 years). They completed the study on average 4 years post-diagnosis (ranging from <1 to 27 years, SD = 4.22 years). Nearly 90 percent of participants indicated that they had completed more than one type of cancer treatment, with 8 percent receiving only chemotherapy and 2 percent reporting only a bilateral mastectomy. Most of the participants were White, non-Latino (87%); 4 percent identified as African American; 1 percent as American Indian; 3 percent as Asian; 1 percent as Multiracial; and 4 percent as Latino. In total, 10 percent of the sample held a high school degree, 51 percent a college degree, and 30 percent a graduate degree (10% reported “other”). In terms of work status, most participants were either employed full-time (51%) or part-time (14%). The majority (62%) reported an income of between US$30,000 and US$100,000. In terms of relationship status, 74 percent indicated that they were engaged, partnered, or married; 10 percent were separated or divorced; 15 percent were single; and 1 percent were widowed.
After institutional review board’s approval was obtained, recruitment occurred in partnership with a US-based breast cancer advocacy group, which used social media to announce the study. After completing an online informed consent form, they were directed to online survey measures. Participants could choose to enter a raffle for 10 US$50 Amazon electronic gift cards using an email address of their choice that was not linked to survey responses.
Instruments
Affect
The Positive and Negative Affect Schedule (PANAS; Watson et al., 1988)—a 20-item measure separated into positive (e.g. excited and interested) and negative affect (e.g. upset and distressed) scales—was used to measure affect. Participants completed each item on a 5-point Likert-type scale assessing affect over the past week. The scale ranges from 1 (“very slightly or not at all”) to 5 (“extremely”). The PANAS has demonstrated adequate psychometric properties in samples drawn from the general population (Crawford and Henry, 2004) and from populations with breast cancer (Hall et al., 2014). The two scales have correlated in the expected direction with measures of anxiety and depression (Crawford and Henry, 2004). Watson et al. (1988) reported Cronbach’s alphas for college student (.84–.90) and adult (.85–.89) samples. Cronbach’s alpha values on the positive and negative affect scales were .92 and .88, respectively, in the current sample.
Loneliness
The 20-item UCLA-R scale (Russell, 1996; Russell et al., 1980), used to measure loneliness, included 10 negatively coded (e.g. “There is no one I can turn to”) and 10 positively coded items (e.g. “I feel in tune with the people around me”). Half of the items reflected satisfaction with social relationships and half reflected dissatisfaction. Items were scored on a 4-point Likert-type scale ranging from 1 (Never) to 4 (Often). Scores have been shown to correlate positively with other measures of loneliness, neuroticism, and burnout, and negatively with social support (Russell, 1996; Russell et al., 1980). Internal consistency estimates have averaged .87 across populations (Vassar and Crosby, 2008) and were reported as .89 in a breast cancer sample (Fogel et al., 2002). The Cronbach’s alpha in the current study was .94.
Self-compassion
Self-compassion was assessed with Neff’s (2003) 26-item Self Compassion Scale (SCS). The SCS consists of six subscales, including the three dimensions of self-kindness, common humanity, and mindfulness, and the corresponding opposite dimensions of self-judgment, isolation, and over-identification. Participants respond to the items on a 5-point Likert-type scale from 1 (almost never) to 5 (almost always), with reverse-coded items for self-judgment, isolation, and over-identification. As suggested by Neff (2003), subscale means were summed to create a total score. Sample items include “I try to be understanding and patient toward those aspects of my personality I don’t like” and “I try to see my failings as part of the human condition.” The SCS correlated in expected directions with self-criticism, social connectedness, depression, and anxiety (Neff, 2003) and yielded internal consistency reliability estimates of up to .92 in a prior sample of women with breast cancer (Przezdziecki et al., 2013) and .95 in the current sample.
Coping self-efficacy for cancer
Coping self-efficacy was measured with the Cancer Behavior Inventory—Brief Version (CBI-B; Heitzmann et al., 2011), a 12-item unidimensional instrument based on the CBI-Long Version 2.0 (Merluzzi et al., 2001). The CBI-B focuses on perceived capability to cope with areas relevant to cancer survivors. Participants indicate their confidence in performing each behavior now or in the near future using a 9-point scale, from 1 (not at all confident) to 9 (totally confident). Item examples include “Actively participating in treatment decisions,” “Maintaining independence,” and “Expressing feelings about cancer.” Item scores are summed to create a total score. Heitzmann et al. (2011) reported internal consistency reliability values of .84 to .88 with different samples of cancer patients. A Cronbach’s alpha of .91 was obtained in the current sample. Scores on the CBI-B have been positively associated with perceived quality of life, optimism, and life satisfaction, and negatively associated with depression, self-blame, avoidant coping strategies, and sickness (Heitzmann et al., 2011).
Domain well-being
The emotional well-being subscale of Functional Assessment of Cancer Therapy—Breast Cancer Version (FACT-B; Brady et al., 1997) was employed to assess domain life satisfaction. We used the 6-item emotional well-being subscale given its conceptual relevance to satisfaction with emotional functioning in the restorative model of well-being. A sample item is “I am losing hope in the fight against my illness.” Negatively worded items are reverse scored. Items are rated on a 5-point Likert-type scale, from 0 (not at all) to 4 (very much; Cella et al., 1993). Brady et al. (1997) reported that emotional well-being scores differentiated breast cancer survivors who had normal levels of activity from those who spent some of their time during the day in bed due to cancer-related symptoms, reporting a Cronbach’s alpha of .69. An estimate of .93 was obtained in the current sample.
Life satisfaction
Life satisfaction was assessed with the Satisfaction with Life Scale (SWLS; Diener et al., 1985). The SWLS consists of five items (e.g. “In most ways my life is close to ideal”) scored on a 7-point Likert type scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Items are summed to produce a total scale score. The SWLS has demonstrated coefficient alphas ranging from .79 to .89 (Steger et al., 2006); we obtained an estimate of .89. Participants’ self-reported SWLS scores correlated with family, friend, and trained observers’ ratings of participants’ life satisfaction (Pavot and Diener, 1993). SWLS scores were also related to, yet distinct from, positive and negative affect, optimism, and self-esteem (Lucas et al., 1996).
Results
A total of 16 participants had one or more missing values on the survey, representing a low percentage of missing values (under 1% of participant responses). Little’s test was not significant (χ2 = 1652, p = .301), indicating that the data were missing completely at random. Because the percentage of missing data was relatively small, the expectation maximization function in SPSS 18 was used to account for missing values at the item level. Observed variable scores did not display problems with univariate normality, according to Kline’s (2011) criteria. Three participants had Mahalanobis distance values that exceeded the acceptable critical value (χ2[7] = 19.05) and were deleted as multivariate outliers, leaving a total sample of N = 275. One-way analyses of variance (ANOVAs) and independent sample t-tests found no significant differences on the theoretical variables as a function of race, education, and medical history (i.e. prior chemotherapy or radiation treatment). Therefore, we aggregated over demographic and medical history variables in our model testing. Descriptive statistics for the observed variables are presented in Supplementary Table 1 available at: http://hpq.sagepub.com/.
Model testing
To create item parcels for use in model testing with latent variables, each measure was subjected to an exploratory factor analysis with a one-factor solution, and individual items were then assigned to one of three parcels for each latent variable according to their loadings in an iterative fashion. For the seven latent constructs, six were represented by three parcels of two to eight items each. The only exception to this was the life satisfaction latent variable, which was measured by two parcels of two to three items each.
Structural equation modeling was performed with the Mplus 7.2 (Muthén and Muthén, 2012) statistical package. Since Mardia’s coefficient was greater than 5, suggesting multivariate non-normality (Bentler, 2005), robust maximum likelihood (MLM) estimation was used. Model fit was assessed with the comparative fit index (CFI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). As recommended by Hu and Bentler (1999), the two-index rule was utilized, with model fit assumed to be adequate when SRMR < .08, along with CFI > .95, or RMSEA < .06. Competing models were compared with both Satorra–Bentler’s χ2 difference test (Bryant and Satorra, 2012), a statistical criterion, and by ΔCFI values, an index of practical significance (ΔCFI > .01 suggests meaningful differences in model fit; Cheung and Rensvold, 2002).
The measurement model was first tested to confirm that the parcels loaded as expected on their respective factors. Results suggested that this model provided adequate fit to the data (see Supplementary Table 2 available at: http://hpq.sagepub.com/); the factor loadings were all above .82, and all factors correlated significantly with one another. The target structural model also produced adequate fit to the data. Though it fit the data less well than the measurement model on the statistical criterion (ΔS-BX2 [3, N = 275] = 19.77, p < .001), this difference in fit was not practically significant (CFI < .01). As noted earlier, the alternative structural model added direct paths from coping efficacy and self-compassion to life satisfaction. This model improved statistically upon the fit of the target structural model (ΔS-BX2 [2, N = 275] = 13.27, p < .01) though this difference did not meet the criterion for practical significance. In sum, the target and alternative models fit the data comparably in a practical sense. While the target model is slightly more parsimonious, one of the two added paths of the alternative model (the path from coping efficacy to life satisfaction) did yield a significant coefficient. Because this additional path may have useful implications for practice, we present its structural coefficient in Supplementary Figure 2 available at: http://hpq.sagepub.com/. This model explained 35 percent of the variance in loneliness, 59 percent in self-compassion, 54 percent in coping self-efficacy, 75 percent in emotional well-being, and 53 percent in life satisfaction.
Significance of indirect effects
To test indirect effects associated with emotional well-being and life satisfaction, we used bias-corrected bootstrapping with 95 percent confidence intervals; a significant mediation effect is assumed when the confidence interval does not include zero. Because of the non-significant paths from self-compassion to coping efficacy, emotional well-being, and life satisfaction, self-compassion was not included in the bootstrapping tests for indirect effects. Coping efficacy was a consistent mediator in the path sequences from positive affect, negative affect, and loneliness to emotional well-being and life satisfaction. In addition, emotional well-being mediated the path between coping efficacy and life satisfaction. Only three sets of indirect effects were non-significant: the paths from positive affect to life satisfaction through loneliness and emotional well-being; from negative affect to emotional well-being via loneliness; and from loneliness to life satisfaction via emotional well-being. 1
Discussion
This study was the first to test the restorative model of well-being (Lent, 2004, 2007) in a sample of breast cancer survivors. We found that the model fit the data well and the predictors accounted for large portions of the variance in emotional well-being and life satisfaction. The hypothesized paths were generally significant, except for the paths from self-compassion to coping efficacy, emotional well-being, and life satisfaction, and the path from negative affect to life satisfaction. With the exception of the latter path, there was good support for the posited direct and indirect links from positive and negative affect to the criterion variables via perceptions of one’s coping efficacy and loneliness (i.e. insufficiency of social support). It was also noteworthy that coping efficacy produced a unique direct path to life satisfaction, consistent with recent findings in a different coping context (Ezeofor, 2015).
The findings involving the predictive utility of loneliness and coping efficacy relative to coping outcomes were generally consistent with prior findings with cancer survivors years 18 months post-treatment (Jaremka et al., 2013). When survivors’ social needs are met, they may be better able to practice self-compassion, to feel confident about managing the demands of life with cancer, and to experience emotional well-being. By contrast, the finding that self-compassion did not yield significant paths to other variables in the model is contrary to prior research linking self-compassion to breast cancer survivors’ psychological functioning (Przezdziecki et al., 2013). Self-compassion did produce strong bivariate correlations with all variables in the model, though it failed to uniquely predict coping efficacy and the two indicators of functioning after controlling for the other predictors.
Theoretical or methodological considerations might account for the limited unique predictive utility of self-compassion in this study. At a theoretical level, it is possible that self-compassion facilitates (and operates through) the use of other coping strategies. For example, studying women with non-cancer-related chronic illness, Sirois et al. (2015) found that self-compassion served as an emotion regulation strategy that predicted the use of adaptive coping (e.g. positive reframing) and decreased the use of maladaptive coping (e.g. self-blame), which in turn predicted higher levels of coping efficacy. Dunne et al. (2018) reported that the direct effect of self-compassion on physical health was partially mediated through health-promoting behaviors. The utility of self-compassion in relation to well-being outcomes could be moderated by such factors as the stage of cancer, recency of diagnosis, or intensity or duration of treatment. Additional research is needed to explore the mechanisms that may mediate (e.g. health-promoting behaviors) or moderate (e.g. stage of cancer) the effects of self-compassion.
While this study suggests the predictive utility of the model of restorative well-being for breast cancer survivors, the current results must be considered within the context of the study’s limitations. First, the study’s cross-sectional design does not permit inferences regarding causality (Maxwell and Cole, 2007). Longitudinal tests are needed to assess the temporal precedence of the predictors relative to the dependent variables, and experimental studies are needed to test the causal sequence implied by the model. Future research could include coping strategies other than self-compassion that may have a stronger association with well-being for cancer survivors. Second, because all measures used in the study relied on self-report, the potential for common method bias should be considered.
Third, given that 87 percent of the current sample identified as White, non-Latino, and over 80 percent reported at least a college education, it would be premature to generalize the findings to more diverse groups of cancer survivors. Research on the range of the model’s applicability to survivors differing in race/ethnicity, educational backgrounds, socioeconomic status, and other individual difference dimensions is needed. Factors such as health literacy and comfort using the Internet to access cancer-related support may be important to account for in future research designs as well. Fourth, conclusions cannot be drawn regarding the applicability of the restorative model of well-being for other cancer types and stages of treatment as the current study focused only on breast cancer survivors no longer receiving treatment.
This study’s results suggest that interventions addressing the social needs of breast cancer survivors that promote self-efficacy for coping with challenges posed by cancer could be especially valuable. In a qualitative study with breast cancer survivors’ post-treatment, survivors described the confidence and reassurance they gained from interactions with medical oncologists and the sense of loss when returning to daily life without the support offered by their oncology team (Powers et al., 2016). Increasing survivors’ self-efficacy for coping and offering them additional sources of support, post-treatment may be key for promoting their long-term well-being.
Fukui et al. (2003) presented a 6-week intervention directed at improving health education, stress management, and coping skills for Japanese breast cancer survivors, finding that the intervention group reported lower levels of loneliness compared to a control group, both immediately after the intervention and 6 months later. Cleary and Stanton (2015) found that breast cancer survivors who completed a 3-hour personal website creation workshop reported decreased depression and increased positive mood and life appreciation, with these changes maintained for 6 months. Coping self-confidence and decreased loneliness mediated the effects of the intervention on decreased depressive symptoms. Thus, interventions that combine a focus on self-efficacy for coping with cancer and enhancing social connectedness may have a greater impact than interventions that focus on only one of these components. Additionally, interventions that address both positive coping behaviors (e.g. benefit finding; Wen K-Y et al., 2017) and maladaptive coping behaviors (Teo et al., 2016) may be warranted. In conclusion, the current findings offer preliminary support for the utility of the restorative model of well-being and for the relevance of particular affective and social cognitive factors in understanding the domain well-being and life satisfaction of breast cancer survivors’ post-treatment.
Supplemental Material
Restorative_Well-being_JHP_Supplemental_Materials_10.30.16 – Supplemental material for Test of a social cognitive model of well-being among breast cancer survivors
Supplemental material, Restorative_Well-being_JHP_Supplemental_Materials_10.30.16 for Test of a social cognitive model of well-being among breast cancer survivors by Trisha L Raque-Bogdan, Robert W Lent and Brooke Lamphere in Journal of Health Psychology
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
