Towards conceptual convergence: A systematic review of psychological resilience in family caregivers of persons living with chronic neurological conditions

Abstract Background The demand for family caregiving in persons with chronic neurological conditions (CNCs) is increasing. Psychological resilience may empower and protect caregivers in their role. Thus, a synthesis of resilience evidence within this specific population is warranted. Aim In this systematic review we aimed to: (1) examine the origins and conceptualizations of resilience; (2) summarize current resilience measurement tools; and (3) synthesize correlates, predictors and outcomes of resilience in family caregivers of persons with CNCs. Design We sourced English articles published up to July 2020 across five databases using search terms involving CNCs, family caregivers and resilience. Results A total of 50 studies were retained. Nearly half (44%) of the studies used trait‐based resilience definitions, while about one third (36%) used process‐based definitions. Twelve different resilience scales were used, revealing mostly moderate to high‐resilience levels. Findings confirmed that resilience is related to multiple indicators of healthy functioning (e.g., quality of life, social support, positive coping), as it buffers against negative outcomes of burden and distress. Discordance relating to the interaction between resilience and demographic, sociocultural and environmental factors was apparent. Conclusions Incongruity remains with respect to how resilience is defined and assessed, despite consistent definitional concepts of healthy adaptation and equilibrium. The array of implications of resilience for well‐being confirms the potential for resilience to be leveraged within caregiver health promotion initiatives via policy and practice. Patient or Public Contribution The findings may inform future recommendations for researchers and practitioners to develop high‐quality resilience‐building interventions and programmes to better mobilize and support this vulnerable group.


| INTRODUCTION
Chronic neurological conditions (CNCs) represent the leading cause of disability and the second most common cause of death worldwide. 1 Globally, it is estimated that approximately one billion people, roughly one in six of the world's total population, are currently living with a CNC. 1 Depending on their origin and aetiology, CNCs are typically divided into four groups: (1) sudden-onset conditions (e.g., acquired brain injury [ABI], spinal cord injury [SCI], traumatic brain injury [TBI]); (2) intermittent conditions (e.g., epilepsy); (3) progressive conditions (e.g., dementia, multiple sclerosis [MS], Parkinson's disease [PD], motor neuron disease [MND] and other neurodegenerative disorders); and (4) stable with/without age-related degeneration (e.g., polio or cerebral palsy). 2 CNCs have an enduring time course and are associated with various complex symptoms, including cognitive impairments, behavioural and psychological problems and marked physical deficits. 2,3 Neurological symptoms and their accompanying disability present challenges for the individual, as independence, functioning and the ability to manage life roles (e.g., employment) are limited. For instance, studies of progressive CNCs (e.g., MS and PD) have reported that the challenges associated with disability management contribute to increased unemployment rates. 4 These findings reflect limitations in the ability to perform occupational and social roles within affected populations. [4][5][6] CNC-related disability results in many persons with these conditions requiring support from others to carry out in-home tasks of everyday living. [7][8][9][10] This role is typically fulfilled by an informal caregiver-an individual responsible for providing unpaid care for family members or close friends. 11 Caregivers often experience role overload, financial strain and are unequipped to provide complex support for their care recipients. 8,12 The extent of this ongoing commitment can culminate in adverse mental and physical health outcomes. 12,13 For family caregivers of persons with CNCs, the caregiving role may contribute to increased stress, depression, anxiety, social isolation and poorer reported quality of life in comparison to the general noncaregiving population. [13][14][15][16] This phenomenon is referred to as caregiver burden. 17 Indeed, depleted caregiver well-being, or burden, impacts the caregiver's ability to provide sufficient support, and is further linked to increased rates of institutionalization of people living with CNCs. 18 Nevertheless, the experience of caring for a loved one with a CNC is broad, dynamic and rarely uniform. 19 Despite facing difficulties, some caregivers experience fewer caregiving consequences, and report rewarding and fulfilling aspects of providing care (e.g., personal growth, strengthening of relationships, enhanced compassion) 20,21 and positive health outcomes (e.g., reduced depressive symptoms). 22,23 Such variability in experience suggests that not all caregivers are harrowed by burden, and that certain caregivers are better equipped to succeed in their role than others. Thus, further exploration of protective strategies that may buffer against the negative effects of burden is needed, and this review seeks to address this gap in knowledge.
To account for this variability, research in the caregiving field is becoming increasingly focused on a protective construct-resiliencewhich, when described briefly, denotes caregivers' ability to adapt to the physical and psychological requirements of their role. 24,25 This transition echoes a paradigm shift in research from a burden-centred caregiving model to a strengths-based model that fixates on healthy development in spite of health risks. 26,27 Still, there remains ample debate in the literature regarding how psychological resilience is defined. Traditionally, trait definitions are used to conceptualize resilience, whereby researchers illustrate resilience as a fixed personal attribute or inherent ability. [28][29][30] This distinction suggests that resilience is stable and unmalleable across the life span. 30 More recently, scholars have investigated the adaptive mechanisms underlying resilience, conceptualizing resilience as a dynamic process. 29 Defining resilience as an adaptive process accepts that resilience may fluctuate in the face of different challenges and stages of the life course and, in turn, is modifiable. 30,31 To further apprise the debate encircling resilience, Windle 29 conducted an extensive review of over 270 resilience-related studies, generating the following definition: 'Resilience is the process of negotiating, managing and adapting to significant sources of stress or trauma. Assets and resources within the individual, their life and environment facilitate this capacity for adaptation and "bouncing back" in the face of adversity. Across the life course, the experience of resilience will vary'.
Evaluation of interventions and policies intended to foster resilience is dependent upon reliable and validated measures. As a reflection of the ambiguity of the resilience construct, a number of resilience measures are available, with minimal progress towards a standardized measure for broad applications. [32][33][34] A methodological review of 15 resilience scales determined that the Connor-Davidson Resilience Scale (CD-RISC), 32 the Resilience Scale for Adults (RSA) 35 and the Brief Resilience Scale (BRS) 36 obtained the highest ratings among authors, despite quality and psychometric deficiencies. 33 Most scales reflect the availability of assets that contribute to resilience (e.g., CD-RISC) 37 or evaluate resilience as an outcome of the capacity to 'bounce back' (e.g., BRS). 33,34 Presently, few measures are available that account for the complexity of resilience from a multilevel and temporal perspective. 33 With limited access to quality scales developed for use in the general adult population, researchers lack robust evidence to inform their choice of resilience measure for differing target populations and contexts. 33 To understand caregiving challenges and the mechanisms by which resilience operates within the caregiving context, multiple studies 25,38,39 have used the Ecological Model of Resilience. 31 This model suggests that resilience operates fluidly across multiple MCKENNA ET AL. | 5 interrelated levels including individual, community and society. 31 This model identifies resources and assets, existent within each of these levels, that may enhance caregiver risk or, alternatively, act to foster resilience. 31 More recently, O'Dwyer et al. 40 proposed a model of resilience in caregivers that conceptualizes resilience as a cyclical process, accounting for the subjective experience of adversity, with varying progressions and magnitudes.
Dissonance persists in the resilience and caregiving literature. A qualitative study among dementia caregivers found that caregivers did not agree on whether resilience was a trait or process, nor could they concur on the factors associated with resilience and its causal pathways. 40 Similarly, a systematic review outlined mainly individual factors as major components of resilience among dementia caregivers; however, the authors acknowledged that there is no single avenue to increase resilience. 41 A recent systematic review determined that resilience was associated with improved caregiver quality of life and alleviated caregiver burden in end-of-life and palliative care contexts, although the authors observed a lack of interest in other psychological aspects that may contribute to resilience. 42 Although the literature supports the notion that a broad range of factors may influence caregiver resilience, the lack of congruence with respect to the conceptualization and measurement of resilience within the literature presents a challenge for future research and practice. 40 Enhancing our understanding of resilience, its measures and associated factors will delineate how resilience capacities may be leveraged and monitored clinically, via intervention, programmes and service development, to better support CNC caregivers in their role. The objective of this systematic review was to synthesize the current scientific literature on the concept of resilience in CNC family caregivers. We aimed to (1) critically examine origins, theoretical conceptualizations and definitions of resilience; (2) summarize current resilience measurement tools; and (3) synthesize correlates predictors, and outcomes of resilience.

| MATERIALS AND METHODS
Our protocol was registered in the PROSPERO database (CRD42020206662). This systematic review was performed in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and reporting guidelines. 43

| Search strategy and selection
A peer-reviewed search strategy 44

| Eligibility criteria
Following a modified PICO (population, intervention, comparison, outcomes) framework, 45 we included quantitative, qualitative or mixedmethods studies that focused on psychological resilience among community-dwelling adult family caregivers (≥18 years old) of adults with CNCs (see Table 1). Articles that were not available in English were excluded. We excluded meta-analyses, dissertations, systematic reviews, case reports, opinion pieces, commentaries and grey literature.

| Screening process
Retrieved articles were managed using Covidence online systematic review software (Veritas Health Innovation Ltd.). One author (N. L.) ran the initial search, and another (L. P.) merged the results into Covidence, where electronic data could be exported, tracked, deduplicated and managed. A two-stage screening process was used to determine eligibility for inclusion. Articles were first screened for relevance by title and abstract by three reviewers (O. M., K. C., K. J.), with the intention of retaining only articles that involved resilience (i.e., resilience was directly referred to in the title or abstract). Any articles with ambiguous representations of resilience were conservatively retained to the next level of review. In the second stage, full texts were reviewed based on eligibility criteria. Agreement of two reviewers (O. M. and K. C.) was required for article inclusion at this stage, resulting in 100% interreviewer agreement. Discrepancies between reviewers were resolved by the last author (L. P.).

| Data extraction
Data were extracted using an Excel template developed by the research team. The following parameters were extracted: (a) study information (i.e., author, year, country, purpose, design, recruitment setting and sample size); (b) participant characteristics (i.e., age, gender); (c) caregiving context variables (relationship with care recipient, CNC, participant eligibility criteria); and (d) resilience components (operationalized definition of resilience, source of definition, measure of resilience, resilience score, resilience-related results).
Two independent reviewers completed the data extraction (O. M. and K. C.). Once both reviewers completed their respective extractions, the results were compared, and any discrepancies were discussed in detail and clarified in a consensus meeting. If consensus among reviewers was not reached, the final decision was made by the last author (L. P.). The authors of three (6%) studies were contacted for missing resilience score data. [46][47][48] Of those contacted, we received additional data from the authors of one study. 47

| Resilience conceptualizations
Certain (n = 7, 14%) studies minimally or unclearly defined resilience, 75,85 briefly presenting it as a general protective psychological factor, 51,60 or simply in relation to stress 56,57 or positive coping (see Table 3). 47 The remaining articles (n = 43, 86%) offered some type of a theoretical definition of resilience. When broadly discussed, the vast majority of included articles incorporated the idea of healthy adaptation into their conceptualizations of resilience. Further, most referred to preserving some level of well-being, equilibrium or positive functioning in the face of adversity. Many (n = 13, 26%) studies referred to the significance of internal and external resources, protective factors, and relational and situational contexts in facilitating resilience development. 38,48,[52][53][54][55]59,65,67,74,78,82,92 For instance, self-compassion was an internal resource conceptually linked to resilience. 54 Beyond these commonalities, researchers differed in terms of whether they defined resilience as a multidimensional process or a personality trait. On one side of this discordance, some studies (n = 8, 16%) used language that illustrated resilience as a personal quality, skill or attribute enabling caregivers to adapt in the experience of hardship. 49,55,64,67,70,76,78,92 Comparably, resilience was described in articles (n = 7, 14%) as an individual's ability or capacity to adjust successfully and maintain normal functioning despite adverse trauma, 63,65,73,79,80,88,89 alluding to the belief that resilience is a fixed competence. In fact, the main aim of one study was to test the hypothesis that caregiver resilience is a personality trait, after which it was concluded that resilience is, indeed, an individual characteristic. 89  Hybrid (trait-process) A range of thoughts, feelings and behaviours and a dynamic process encompassing positive adaptation within the context of significant adversity; it is also considered a personality characteristic that moderates the negative effects of stress and promotes adaptation The characteristic way in which people approach and cope with life events, described in terms of three related tendencies: commitment, where behaviour is influenced by the meaning and purpose seen in a situation; control, the ability to make one's own choices in a situation; and challenge, the tendency to perceive life events as opportunities for development, rather than threats  64 Hybrid (trait-process) A control-related intrapsychic variable that may promote a more successful adaptation to care demands; personality trait, but broader approaches underline the importance of relational and situational contexts for resilience behaviour Scott (2013) 46 Process A characteristic or developmental process in individuals that, when activated, aids in thwarting the effects of social conditions that can lead to impaired daily functioning Pessotti et al. (2018) 88 Trait One's capacity for successful adaptation when faced with the stress of adversity; not invulnerability to stress, but, rather, the ability to recover from negative events Wilks and Vonk (2008) 79 Trait Implies a track record of successful adaptation in the individual who has been exposed to stressful life events, and an expectation of continued low susceptibility to future stressors; reflects an outcome strength, that is, the ability to recover from the stressor successfully  presented in Table 4. Three studies (6%) used caregiver-specific measurements. 50,85,86 The most commonly used scale was the Resilience Scale (RS) by Wagnild and Young. 37 Twelve studies (24%) 46 One retained study explored conceptual resilience discrepancies between academic definitions of resilience and caregivers' personal conceptualizations, and found that caregivers extended the concept more broadly and emphasized the role of self-compassion. 54 The implications of self-compassion as a protective factor for psychological wellness have been previously documented in family caregivers, 102  The current divergence in resilience conceptualization and assessments may be representative of the relative novelty of this concept in comparison to other long-standing psychological constructs that have had been comprehensively validated across populations and widespread contexts. Equivalently, the differing use of quantitative resilience scales supports the notion that there is no gold standard of resilience assessment. 33 The absence of homogeneity in resilience measurement undermines the ability of researchers, clinicians and community members to reliably monitor and evaluate the efficacy of resilience-building programmes. 103 This lack of standardization further prevents resilience levels from being compared across different caregiver subpopulations. 27 Similarly, with few developed and validated condition-specific scales within the resilience field, it is difficult to reliably verify and contrast resilience levels within distinct caregiving populations. It is recommended that future resilience and caregiving research draw on contemporary views of resilience from broad literature to formulate context-specific measures, while attending to the evolving theory and research. This will surely elevate the quality of resilience-based research moving forward, while preventing further dispersion within the field.
The underrepresentation of caregivers of persons with less common CNCs in the included studies was apparent and presents a challenge for advancing disease-specific resilience applications. Few studies examined sudden-onset conditions, there was an absence of stable and intermittent types of conditions and progressive conditions consisted of mostly dementia. As we approach a saturation of research in resilience in dementia caregiving, resilience in caregiving populations of other CNCs (e.g., MS, MND, epilepsy, cerebral palsy) remains understudied. This is problematic because it has been empirically proven that caregiver demographics and health outcomes vary as a function of the specific CNC encountered. 104 It is important that such overlooked populations garner further interest in the field of resilience investigation, allowing their respective resilience processes to be equitably understood, measured and harnessed. Finally, we observed resilience exploration in connection to caregiver burden in one quarter of the included studies. This prevalence suggests that resilience research in CNC family caregivers is inconsistent with emergent resilience research, as broader disciplines now favour a strengths-and competence-based approach. 26,27 We acknowledge that a proportion of the reviewed studies (30.5%) excluded burden from their design and depicted the positive link between resilience and well-being. However, it appears that many researchers continue to examine the negative consequences of caregiving, and how resilience protects caregivers from impending risk innate to their role. This, in turn, fails to abandon the outmoded deficit-based model of caregiver resilience and mental health. 27 Despite the expected inverse relationship between resilience and burden, we caution against the assumption that caregiver burden and resilience can coexist, such that one determines the other. Instead, it is advised to position each concept as mutually exclusive, as the exact causal mechanisms responsible for their association remain unknown. 105,107 Arguably, through conflation of dimensions of flourishing and languishing, our understanding of resilience becomes obscured by the pathologies and dysfunction denoted by caregiver burden. 107 It is suggested that scholars adopt a more proactive or preventative approach that prioritizes building strengths, 27 while simultaneously no longer assuming that CNC caregiving is uniformly burdensome.

| Limitations
This study had some limitations. First, resilience is a wide-ranging and nuanced concept that parallels with other psychological topics (e.g., hardiness, adaptation, coping) across psychosocial disciplines. To address this concern, we adopted a broad approach in our initial search strategy to acquire as much relevant literature as possible; still, it is possible that we missed relevant literature because of our resilience-specific focus. Second, due to the nature of many variables often examined in close association with resilience, it is possible that there was potential methodological bias among the retained studies in describing these relationships, including issues of simultaneity and reverse causality.
There is scarcity of research in family caregivers of more un-

| CONCLUSION
This review synthesized existing knowledge of resilience in family caregiving for persons living with a CNC. Findings revealed an insufficient level of agreement among researchers with respect to how resilience is theorized, conceptualized and assessed. This emphasizes the fact that resilience is a complex, multifaceted phenomenon that merits further clarification within the caregiving sphere with respect to whether it is a trait, process or a hybrid of the two. Collective findings demonstrate that resilience is associated with better overall health and psychological well-being, and contributes to optimal stress management among CNC family caregivers. The ideal context in which resilience develops and how that process varies cross-culturally has yet to be determined, though this represents a useful direction for future research and complements newfound socio-ecological resilience theories. Furthermore, while a strengths-based approach does not currently unanimously prevail across the reviewed literature, there is room for evolution to dissociate weakness-, risk-and deficit-focused models of caregiving from resilience and to cultivate approaches rooted in caregiver empowerment. With limited representation of intervention studies, there is a need to develop targeted interventions for informal CNC caregivers aimed at promoting resilience and increase awareness of the positive aspects of caregiving.