Background

Anorexia nervosa (AN), bulimia nervosa (BN) and binge-eating disorder (BED) are three of the most commonly occurring ‘feeding and eating disorders’ for adults (DSM-V; American Psychiatric [1]). Eating disorder (ED) psychopathology in terms of AN, BN and BED involves biological (e.g. weight status), cognitive (e.g. over-evaluation of weight and appearance), affective (e.g. fear of eating in public) and behavioural (e.g. fasting, purging) features [2]. Despite the demonstrated efficacy of transdiagnostic ED interventions such as CBT-E [3, 4] or interpersonal psychotherapy [5, 6], as well as other ED subtype specific interventions such as Maudsley model of AN treatment for adolescents and young adults [7], epidemiological data demonstrates that recovery rates remain at approximately 55–69% [8, 9]. Even when eating behaviours improve, people with EDs can experience both short-term health issues regarding refeeding syndrome as well as long-term negative health outcomes such as cardiovascular conditions and osteoporosis [10,11,12].

Many researchers have attempted to explore how to improve recovery outcomes for this population. There are two main recovery paradigms that define recovery differently; the medical model and the recovery model. In comparison to the medical model, recovery is considered within the recovery model as a journey and a process, whereby the goals of recovery are determined by the individual rather than professionals [13, 14]. Due to the major health risks involved in EDs, professionals working with people with EDs may be at risk of over-emphasising the significance of biological and behavioural symptoms of EDs as the sole indicators of recovery, with a lack of consideration for cognitive and affective symptoms which may persist despite alleviation of the more observable symptoms [15].

ED recovery has been recognised as being difficult to achieve due to the ego-syntonic nature of the disorder [16], meaning that some people with EDs value their disorder and endorse their eating behaviours. A systematic review and qualitative meta-analysis were conducted by de Vos et al. [17], which aimed to collate findings from studies assessing fundamental criteria for ED recovery. The authors demonstrated that resilience was one of six criteria found to have large frequency effect sizes, which accounted for 13.8% of all recovery criteria assessed [17]. The authors concluded that resilience along with other elements of psychological well-being should be considered fundamental aspects of ED recovery. Other authors suggested that a focus on the psychological aspects of recovery would reduce the risk of ‘partial recovery’ occurring, whereby the physical and behavioural symptoms reduce but the psychological aspects remain [18], a term referred to by Keski-Rahkonen & Tozzi [19] as ‘pseudorecovery’ (pp. S83). The role of resilience in recovery is that it assists a person to better adapt to the stress and adversity they are experiencing and to gain control over their difficulties [14]. In this sense, it can be understood why individual factors such as self-efficacy and competence have been identified as “assets” influencing a person’s resilience; however, researchers have also identified that “resources” or external factors such as family support or community services also impact a person’s resilience (pp. 6, [20]).

Resilience, a concept defined within positive psychology perspectives [21] and commonly referred to as the ability to ‘bounce back’ from difficult experiences [22], might help to explain why some people with EDs recover well whilst others do not. According to Windle [20], who references the work of resilience theorists such as Luthar, Rutter, Ungar and Masten, resilience can be underpinned by the following features: (i) Resilience is influenced by factors beyond that of the individual, (ii) resilience is a process, and (iii) resilience occurs in response to adversity, stress or trauma. In relation to the first feature, human development is shaped by many environmental and societal influences [23]. In this sense, it may not be useful to identify personal, family and social factors in isolation, but rather to investigate the factors influencing resilience across personal, family and social levels (i.e. multi-level factors). Many experts in the resilience field acknowledge the importance of looking beyond individual traits [20, 24,25,26] to explain ‘psychosocial resilience’ [27] including within ED research [28]. In relation to the second feature, experts have highlighted the importance of assessing responses to stress and adversity as close as possible to the occurrence of the stressor so as to shift the focus of resilience research to stress reactivity rather than restoration of well-being [29]. This suggests that researchers should be trying to understand the process of resilience during ED recovery as it occurs, rather than when the adult has ‘recovered’. In relation to the final feature, the stress, trauma or adversity preceding resilience for people with EDs might include psychosocial difficulties related to the ED diagnosis [30], subsequent health complications [12], as well as other life adversities (e.g. negative familial comments about eating, weight or appearance [31, 32]).

Within the ED literature, two studies were identified which directly assessed resilience levels of adults with EDs [33]: the first demonstrated that resilience levels were lower for those recovering from EDs compared to those who had recovered from an ED and the general population [34]; and the second showed that increased resilience among those with EDs was associated with a reduction of ED symptoms, as well as improvements in psychological and social domains of quality of life over time [35]. Furthermore, Las Hayas et al. [36] used qualitative methods to identify the stages involved in the process of developing resilience for adults recovered from EDs, which included 14 stages. Of note, this was the first attempt within the ED literature to investigate resilience as a process, and the authors also identified that resilience facilitated the journey to recovery for people with EDs [36]. However, the authors noted a number of limitations to their study to be considered by future researchers.

This study set out to build on the current understanding of resilience as a process rather than a trait of a person, and to overcome some of the shortcomings of previous resilience research within the ED field by incorporating findings on the external influences associated with resilience alongside the personal influences. The overall aim of the current study was to explore the multi-level process through which adults recovering from EDs develop resilience, from the perspectives of clients and clinicians. The research question was ‘What are the personal, family and social influences on the process of developing resilience for adults recovering from EDs?’. The objective of this research was to outline the stages involved in the process of resilience development, which might help to inform families and services in how best to support adults with EDs during their recovery.

Method

Design

This qualitative study was conducted using a constructivist paradigm and a phenomenological approach, which aimed to address the research question by ascertaining the human experience of the resilience process during ED recovery from client and clinician perspectives. This approach acknowledged that resilience is a socially constructed concept and that the best means in which to explore this concept was via semi-structured interviews, with the view that in-depth qualitative research can help to guide good clinical practice as well as inform policy decisions [37]. Ethical approval for this study was granted by a public healthcare ethics committee in July 2018 (Ref: 130618KG) and by a city-based hospital in April 2019 (Ref: RS19–002).

Participants

Clients

Fifteen participants were recruited to form this sample. Eligibility criteria to be met included that participants were aged 18 years or above; they must have received a diagnosis of AN, BN, BED, Other Specified Feeding or Eating Disorder (OSFED) or Unspecified Feeding or Eating Disorders (UFED) determined through team-based assessment using DSM-V criteria; they must have been attending adult mental health services as either inpatients or outpatients; and they must have been deemed by their treating clinician to be at a stage of recovery which suggests they are making active steps towards their recovery. Prochaska and DiClemente [38] and Prochaska et al. [39] define five stages of recovery within their transtheoretical stages of change model: pre-contemplation, contemplation, preparation, action and maintenance. Regarding the inclusion criterion relating to stage of recovery, participants identified by their treating clinician as being in the final two stages (i.e. action, maintenance) were deemed eligible to participate in this study.

Clinicians

This sample comprised of 15 clinicians who self-identified as working with adults with EDs. Ten worked in a general adult mental health outpatient service, whereas five worked in a specialist ED unit. There was no overlap among clinicians across these two settings. These clinicians had on average 12.10 years (SD = 8.42) of experience working in their respective services.

Procedures

Convenience sampling was used as the heads of two ED services (i.e. general adult mental health service and specialist ED unit) were approached in order to ascertain interest in participating in the research, both of whom agreed to their services taking part. A short research pitch was prepared and delivered by the lead author (KG) whereby service staff members were allowed the opportunity to ask questions about the research before agreeing to participate. Information sheets were provided 2 weeks prior to participant informed consent being retrieved. Interview slots were arranged with clinicians who agreed to take part themselves, and clinicians also disseminated the information sheets to potential clients who met eligibility criteria. All participants were screened using eligibility criteria, and interviews were scheduled and took place in the client’s and clinician’s place of treatment and place of work, respectively.

Sample size was determined based on guidelines from the qualitative methodology literature (e.g. [40,41,42]). The two groups were included to generate more perspectives on the topic which would help to inform clinical practice, similar to previous resilience research [36, 43].

On the day of the interviews, participants from both groups completed short demographic information questionnaires (see Table 1) before taking part in a semi-structured interview conducted by the lead researcher (KG). Prior to the commencement of the interview, the purpose of the research was explained to participants (i.e. to explore the multi-level process through which adults recovering from EDs develop resilience) and limits of confidentiality were outlined. Participants were asked to describe what their understanding of the term resilience was, and this was further clarified with participants by outlining three key features through which resilience is currently conceptualised as per Windle [20].

Table 1 Demographic information on client and clinician groups

Data collection

Data were collected by means of semi-structured interviews using an interview guide. Questions from this interview guide were informed by the resilience literature and reviewed by two clinicians who treat people with EDs. Clinicians were asked to respond based on their experiences of working with adults with EDs thus far in their careers. The average duration of interview was 46.48 min (SD = 11.82). Interviews were conducted between August 2018 to November 2019.

Data analysis

Data were analysed using reflexive thematic analysis (TA) as outlined by Braun and Clarke [41, 44, 45]. Reflexive TA is a version of thematic analysis, which aims to identify patterns of meaning within the data, but with the recognition of the active role of the researchers in the knowledge production process (i.e. reflexivity). The researchers engaged in active reflection throughout the research process, which was aided by the use of a reflective research journal in order to enhance research credibility [46, 47].

The 30 interviews, which were conducted and audio-recorded by KG, were transcribed verbatim by HOD. Transcripts were then reviewed by KG. The lead author engaged in a process of familiarisation with and immersion in the data by listening to the audio-recordings while reading transcripts. Initial ideas and thoughts about potential codes and/or themes were noted before generating codes using NVivo [48], a qualitative data analysis computer software package. Although the research question was considered from a theoretical perspective (i.e. Bronfenbrenner’s ecological theory), data were analysed in an inductive manner whereby themes were constructed based on a data-driven approach using latent coding, representing the researcher’s explicit interpretation of implicit meaning of data. Multiple codes could be given to any specific selection of text. Initial themes were generated, which were later reviewed and refined. Twenty percent of the data were also coded by a second author (HOD). This step was conducted so that authors could discuss ideas and decide on how best to tell the story of the data, rather than for reliability checking, which is not endorsed within this research paradigm [41]. Braun and Clarke’s [45] definition of a theme was used (i.e. stories about particular patterns of shared meaning across the dataset). The entire process was non-linear, requiring movement back and forth between each step. Criteria outlined in two documents, Standards for Reporting Qualitative Research (SRQR [49];) and Braun and Clarke’s [44] 15-point Checklist of Good Criteria for Thematic Analysis, were met to ensure good quality reporting of results.

The component of the research question which aimed to examine the process through which adults recovering from EDs develop resilience was addressed by identifying the temporal sequencing of the stages involved in the resilience process. This was achieved by sequencing themes in the order in which they were described to have occurred for participants in general and in an order that made logical sense. The research question also aimed to identify the multi-level factors influencing the resilience process, which was addressed by distinguishing personal, family and social influences during the coding process. ‘Personal’, ‘family’ and ‘social’ were assigned as prefixes to all codes in order to specify influences of resilience on a personal, family or social level.

Methodological rigour

Various strategies were utilised as per Lincoln and Guba’s [50] four trustworthiness criteria to ensure methodological rigour in the current study (see Additional Information 1 in the Supplementary Materials section). Considering the interpretivist approach within which this study took place, attempts at enhancing the trustworthiness of the study serve to develop more rich, in-depth and reflexive findings, rather than to achieve validity or reliability of results.

Results

The following section describes the multi-level process through which adults recovering from EDs develop resilience. Samples of supporting quotations from participants can be found in Table 2. ‘Participants’ is the term used to refer to the client and clinician groups collectively, whereas findings specific to one or the other group are indicated by using the terms ‘clients’ or ‘clinicians’.

Table 2 Quotations supporting sub-theme components across personal, family and social levels

The process through which adults recovering from EDs develop resilience

The multi-level process of resilience development for adults recovering from EDs can be described under the overarching theme, ‘Bouncing back to being me’, which is best understood under three separate stages; Stage 1 (‘Who am I without my ED?’) includes three themes, stage 2 (‘My ED does not define me’) also includes three themes, and stage 3 (‘I no longer need my ED’) includes one theme (see Fig. 1). As can be seen in Fig. 2, there are 20 sub-themes involved in this process; seven sub-themes are influenced on an individual level, three are influenced on an individual and family/social level, whereas ten sub-themes are influenced on a personal, family and social level. This demonstrates the significance of the interaction between internal and external factors in the process through which adults recovering from EDs develop resilience. Below is a brief description of the resilience process for adults recovering from EDs. It is highly recommended that readers view Additional Information 2 from the Supplementary Materials section for the expanded interpretation of the results.

Fig. 1
figure 1

The process through which adults with EDs develop resilience during recovery

Fig. 2
figure 2

Thematic map of overarching theme, themes and sub-themes depicting the stages involved in the multi-level process of resilience development for adults recovering from EDs. Colour code: Red = sub-themes influencing resilience on a personal level; Blue = sub-themes influencing resilience on a personal and family/social level; Green = sub-themes influencing resilience on a personal, family and social level

Stage 1 of the resilience process: ‘Who am I without my ED?’

This stage describes a period whereby adults with EDs become defined by their ED, and are unable to separate their own identity to that of the ED (Stage 1: ‘Who am I without my ED?’). This stage generally reflects a personal process, with little involvement at a family or social level, except for sub-themes 1c and 3b.

During this initial stage, participants report that adults with EDs can feel quite dependent on their EDs in order to cope with the demands of the world and possibly to survive (Theme 1). Within this ED dependency theme, participants discuss how ED behaviours assist the adults with EDs to cope with stress or adversity (Sub-theme 1a); clients regularly discuss a point in their lives whereby the ED completely takes over (Sub-theme 1b) and becomes “all-consuming” (Client 4, female, aged 51–60); and how adults with EDs hide or deny their eating problems, or act in an avoidant manner regarding the severity of their eating issues (Sub-theme 1c). Family can also play a role in minimising or denying the ED issues which inevitably leads to the problem persisting rather than the person confronting the issue in a more proactive manner. Taken together, these three sub-themes reflect a stage whereby the adult with an ED becomes quite dependent on their ED in order to cope with the demands of the world and possibly to survive.

Adults with EDs often go through a period of considering other possible ways they might cope instead of depending on their ED (Theme 2). During this period, adults with EDs may pay more attention to their other skills (e.g. communication, problem-solving) which might assist them in coping with their difficulties (Sub-theme 2a); and clients also discuss the advantage of having experienced good coping previously in their lives, resulting in a learning from past experiences (Sub-theme 2b). These two sub-themes involving an acknowledgement of personal skills and evidence of having previously coped with life difficulties allow the adult with an ED to consider what other possible ways they might cope instead of depending on their ED as a core coping strategy.

Participants describe a period of time whereby adults with EDs weigh up the pros and cons of letting go of the ED (Theme 3). As noted previously, EDs assist some people to cope with life adversity, and so they may fear that letting go of their ED will have a negative impact on them (Sub-theme 3a); over time, adults with EDs develop more self-awareness through introspection (as well as through input from their family and service) about why the ED formed for them in the first instance (Sub-theme 3b); and they identify their motivations to let go of the ED and experience a readiness to let go (Sub-theme 3c). These three sub-themes involving initial fears of making change, introspection regarding why the ED formed in the first instance, as well as reflecting on personal motivation for change are all part of a process whereby the adult with an ED is weighing up the pros and cons of letting go of the ED.

Stage 2 of the resilience process: ‘My ED does not define me’

During the second stage, the adult with ED begins building a self-identity separate to their ED by tapping into resources on an individual, family and social level (Stage 2: ‘My ED does not define me’). All sub-themes included in this stage are influenced across all three levels, reflecting the multi-level input in developing resilience. Though the adult with an ED may be less reliant on and less consumed by their ED compared to the first stage of the resilience process, the ED is still very much part of their life.

The resilience process is greatly promoted during this second stage when individuals with EDs, their families and other people in their immediate environment see the bigger picture regarding EDs (Theme 4). This theme highlights the importance of the individual, their family and service gaining knowledge and understanding about EDs so the client feels “understood” and “validated” (Clinician 13, female, aged 31–40) (Sub-theme 4a); the recognition by the individual, their family and service that the ED is not just about eating behaviours (i.e. dieting, bingeing etc.) and weight (Sub-theme 4b); the understanding of the individual, their family and service that the recovery journey is long and non-linear (Sub-theme 4c); and the importance of the person, their family and society in being able to manage their anxiety about the disorder (Sub-theme 4d). Collectively, this increased knowledge about EDs, the emphasis of factors beyond weight and eating behaviours in recovery, the knowledge that recovery is a long journey and the ability to manage anxieties about the disorder results in an ability to view the bigger picture of what is involved in ED diagnoses.

Safety and security (Theme 5) are noted as important aspects of the resilience process, which are influenced on an individual, family and social level. Participants emphasise the importance of having a secure base and positive relationships in building resilience (Sub-theme 5a); the need for openness and honesty between the adult with an ED and their family members (both about the ED and about other family issues) as well as the wider public (i.e. friends and service providers) (Sub-theme 5b); and the importance of striking a balance between providing support but also allowing the person to be autonomous and independent, especially regarding ED decisions (Sub-theme 5c). These sub-themes collectively result in a safe and secure environment in which the adult with ED can develop resilience.

Participants discuss the importance of being ready for and aware of potential knocks to their resilience through personal, family and social influences throughout the recovery process (Theme 6). As significant adversity or cumulative life stresses are often related to the onset of EDs for adults with EDs, other potential life stresses during recovery might trigger a worsening of symptoms or relapse (Sub-theme 6a); another potential knock to resilience to be aware of for people with EDs is judgemental environments in terms of pressure, high expectations and comparisons (Sub-theme 6b); and adults with EDs also appear susceptible to any factors that could potentially lead to a disturbed body image, which occurs when they are exposed to an over-emphasis on the importance of food and body image in society (Sub-theme 6c). These sub-themes of general life stress, judgemental environments and an emphasis on food and body image are areas which need to be monitored in order to prevent a set-back in the resilience process.

Stage 3 of the resilience process: ‘I no longer need my ED’

During the final stage of the resilience process, adults with EDs no longer feel they need their ED in order to cope with the demands of the world or to survive due to better developed resilience and the utilisation of a wider set of resources (Stage 3: ‘I no longer need my ED’).

It is recognised that adults with EDs begin to identify resilience in themselves (Theme 7) and become aware that they no longer need the ED to cope and survive. Within ‘the resilient me’ stage, adults with ED are able to reintegrate “more normal experiences” (Client 6, female, aged 21–30) into their lives (Sub-theme 7a); and they may have developed a more positive mindset and future outlook on their life (Sub-theme 7b). This sub-theme includes personal factors that may be more traditionally perceived as ‘traits’ of resilience. Such factors include patience, determination, hope, high self-esteem, self-direction, and self-belief, all of which were noted by the participants as important factors leading to a more generally positive mindset. However, these factors are referred to as goals to work towards or characteristics to be built upon, and not factors that either exist or do not exist within a person. Once this positive mindset is achieved, a person is at a stage within the resilience process which better facilitates recovery (i.e. a reduction of ED symptoms, improved psychological well-being). However, clinicians warn that these personal factors can also work to maintain a person’s ED (i.e. self-determination to remain thin), and although adults with EDs might go on to recover successfully, some may otherwise be at risk of relapse. This highlights the dynamic process involved in developing resilience during ED recovery.

Discussion

Summary of research aims and findings

This qualitative study aimed to describe the multi-level process through which adults recovering from EDs develop resilience, from the perspectives of clients and clinicians working in the field. The overarching theme which described the process of developing resilience was ‘Bouncing back to being me’, which involved three stages: ‘Who am I without my ED?’, followed by ‘My ED does not define me’, and finally ‘I no longer need my ED’. This study demonstrated that self-identity was recognised as an important feature in this resilience process, whereby the resilience process involved a regaining of personal identity separate to that of the ED identity. Within ED research, Bowlby et al. [51] showed that recovery involves an understanding that the ED is separate from one’s identity as a person. This study built on these findings by identifying the various stages involved in this multi-level resilience process through which a person recovering from an ED develops a sense of self that is separate to that of the ED. This process might explain how EDs become less ego-syntonic, as the thoughts and behaviours associated with EDs are no longer as valuable to the person’s sense of self. As long as a person’s sense of self is intertwined with the ED, it is difficult for thoughts and behaviours related to the ED to become ego-dystonic. The separation between sense of self with that of the ED might make it easier to challenge disordered thoughts and behaviours which may be driving the EDs during treatment.

Findings in relation to previous research

This study built on the previous research conducted by Las Hayas et al. [36], demonstrating many similarities across study findings. Of the 14 themes identified in their study, all themes overlapped conceptually with findings from the present study. However, the current study identified a number of themes which were not identified by Las Hayas et al. [36]. Firstly, this study referred to sub-themes occurring early on in the resilience process such as ‘ED is a source of coping’ and ‘Secrecy, denial and avoidance’. It is likely that this difference reflects the inclusion of participants in recovery in the current study which differed to Las Hayas et al.’s [36] inclusion of ‘recovered’ individuals only. As suggested previously, it is important to investigate responses to stress and adversity as close as possible to the occurrence of the stressor so as to shift the focus of resilience research to stress reactivity rather than restoration of well-being [29]. Therefore, it was valuable to include these sub-themes which reflect important influences relevant to the early stages of the resilience process, as specified by participants. Another unique sub-theme identified in the current study was the importance of ‘Learning from the past’. Participants discussed how the ability to reference a previous time in their life that they showed good coping, particularly during past difficult experiences, facilitated the resilience process. Exposure to adversity in moderation can initiate previously untapped resources, allowing a person to benefit from support they were not previously utilising, and this leads to mastery for future adversities experienced [52]. Finally, there were various sub-themes identified in stage two which were not demonstrated by Las Hayas et al. [36] including the acknowledgement of ‘The difficult road ahead’, the importance of ‘Managing emotions’, the significance of ‘Communication and honesty’ and the acknowledgement of the potential impacts of ‘Judgemental environments’ as well as ‘Food and body image emphasis’ on the resilience process. It is important to note that these sub-themes were identified as multi-level influences, meaning that they were significant across personal, family and social levels of influence. Factors beyond the individual were not addressed by Las Hayas et al. [36]. These findings support the postulation of experts in the field that resilience is a multi-level, psychosocial and ecological concept [24,25,26,27, 53], which was similarly emphasized by ED experts [28, 36].

Clinical implications

Findings from the current study are relevant to clinicians who are working with adults recovering from EDs specifically. First, this division of the resilience process into three stages allows for the identification of the resilience stage a person is at during recovery. The identification of the three stages assists clinicians to pace the expectations of the client, their families and themselves. For instance, the process suggests that it would be unrealistic for clients to want to fully let go of their eating behaviours before the identification of other skills and coping abilities has occurred. This knowledge will help clinicians and clients to set realistic goals during treatment. The caveat of this resilience process, just like other psychological processes outlined in previous research, for example, the transtheoretical stages of change [38] or the five stages of grief [54], is that it is not as simplistic as it may appear [55]. This process, though outlined as a linear process for ease of access, is more dynamic and complex, and it is likely that some adults may skip steps, or steps may occur at different stages for different individuals. Without this consideration, clinicians may be at risk of over-simplifying a complex process [20, 56].

Second, this framework identified the multi-level influences on the development of resilience during ED recovery, which were most significant during the second stage of the resilience process. This multi-level process reduces the level of expectation and the burden put upon people in recovery to be solely responsible for their ‘bounce-backability’. Researchers in other areas of resilience research (e.g. disability) previously reported the potential danger in implying individual responsibility without consideration of external factors [57]. Results demonstrated that families, friends and services have a direct role in influencing the resilience process by understanding the bigger picture of what is involved with EDs, by creating a safe and secure environment within which the person can recover, and through a reduction in behaviours and attitudes that may result in potential knocks to the person’s resilience. Family-based interventions are often recommended in working with clients with EDs, especially considering family factors may inadvertently influence the recovery of adults with EDs (e.g. enabling behaviours, expressed emotions, psychological distress [58]).

Finally, the findings pertaining to this resilience process fit with positive psychology perspectives that aim to restore hope in individuals overcoming adversity by identifying what makes life worth living rather than to focus solely on alleviating symptoms [29, 59]. This supports personal recovery perspectives which aim to allow a person to function their best despite ongoing persistence of symptoms [60]. This multi-level resilience process similarly suggests that by building up one’s resilience (rather than focusing on symptom reduction), adults with EDs arrive at a point whereby they no longer depend on their ED to function or survive, suggesting that positive psychology interventions such as life coaching and resilience training may be useful in working therapeutically with adults with EDs. There has been some evidence of the efficacy of positive psychology interventions in the forms of ‘positivity groups’ [61] and mindfulness-based intervention [62] for adults with EDs, but research on such interventions appears to be far less frequent compared to the more commonly researched interventions such as CBT-E [3]. It must be recognised that though various psychological or psychotherapeutic interventions might serve as effective treatments for EDs, clinicians might be ethically obliged to focus on the ED behaviours if they pose a significant risk to the person (e.g. risk of starvation, cardiac problems).

Research implications

From a research perspective, this study is one of four studies identified on the topic of resilience in the ED literature [34,35,36], and is the only study, to our knowledge, that focused on resilience from a multi-level perspective. This study provides support for a multi-level perspective of resilience, and sheds light on how future resilience research can look at an integrative approach by identifying factors influencing resilience across personal, family and social levels to establish the role of multi-level influences on resilience development.

Strengths and limitations

As well as the strengths previously mentioned, this qualitative study offered unique contributions by allowing for the perspectives of adults with EDs as well as their treating clinicians to be captured through semi-structured interview, which has been suggested to be an effective means for guiding good clinical practice [37], particularly for a concept as complex as resilience [63]. As findings are based on client and clinician experiences, this increases the usability of the framework, which is more likely to be endorsed by clients and clinicians as it was informed by them. However, there were some limitations. Although we attempted to include male participants to avoid exclusion of this under-represented group, the client sample was not balanced for gender with only two males included. Furthermore, the framework might be more fully informed with the inclusion of data from family members, considering the role they appear to play in ED recovery [58].

Conclusions

Despite these limitations, this qualitative study provides a rich description of the multi-level process of resilience development for adults recovering from EDs. This framework provides empirical evidence that resilience is an ecological process involving an interaction between internal and external influences, which involves a dynamic interplay between adults with ED and their most immediate environments (i.e. family and social). The inductive and qualitative nature in which this process was identified should result in a high level of acceptability of the framework in clinical practice by clients and clinicians, so that this framework can be used to inform treatment choices and recovery plans for adults recovering from EDs from a positive psychology perspective.