Rehabilitation agency of older adults in group-based intervention.

Abstract Aims: This study investigated the individual rehabilitation agency of older adults in a one-year group-based gerontological rehabilitation context. Here, rehabilitation agency is understood as being manifested when older adults make choices and decisions regarding their everyday lives, including notions of themselves. Methods: The data were obtained via non-participant observation of the final individual goal evaluation sessions of 38 older adults with their personal counselor. In these sessions, older adults discussed their rehabilitation outcomes, actions, choices and decisions during the rehabilitation year, along with their future in home settings. The data were analyzed using qualitative content analysis and an agency-centred approach. Findings: The findings revealed that older adults made choices and decisions differently concerning their life in and beyond the rehabilitation context. Four forms of rehabilitation agency of older adults were identified: (i) renewable, (ii) widened, (iii) selective and (iv) fractured. These forms of agency were differently connected to older adults’ life courses and to their peer relations in the rehabilitation context. Conclusions: An agency-centred approach could produce new theoretical ideas and practical implications for developing older adults’ rehabilitation to better meet their needs as well as the goals of group-based rehabilitation interventions.


Introduction
Until recently, rehabilitation for older adults has mostly been based on the biomedical paradigm with geriatric and physical approaches [1][2][3][4]. As a consequence of this paradigm, older adults' rehabilitation interventions have widely been focused on specific illnesses (e.g. strokes) or injuries (e.g. hip fractures) with certain physical disabilities or determined functional constraints [5][6][7][8]. However, this paradigm has been criticized as too narrow from the perspective of older adults' everyday activities and social participation. In response to this criticism, client-centred approaches have been introduced in geriatric and rehabilitation institutions [2,[9][10][11][12]. Although these client-centred approaches have emphasized the individual needs of older adults, the implementation of the rehabilitation has been unable to truly recognize older adults as active subjects with their unique life courses.
Overall, both the traditional medical paradigm and the utilization of client-centred approaches have been criticized for the limitations of their rehabilitation outcomes and their short-term practical impacts on older adults' everyday lives [13][14][15]. As a consequence, there is an increasing need to enrich and to redefine both the traditional medical paradigm and clientcentred approaches within rehabilitation for older adults. One attempt to accomplish this has included group-based models [16][17][18][19]. In these models, it has been suggested that older adults' unique individual life course resources can be used with their societal and historical cohort experience as an efficient component in group-based rehabilitation [2,20]. However, the implementations of these group-based rehabilitation interventions have not been unambiguously successful [2,18,21,22].
Following from these critical notions, multifaceted theoretical approaches combined with new practical research are needed so that both a theoretical basis and efficient practices for older adults' rehabilitation could be designed and produced. On the basis of this need, our study investigates the rehabilitation agency of older adults in a one-year group-based rehabilitation. Here, rehabilitation agency is understood as being manifested when older adults influence, make choices and decisions regarding their everyday lives, including notions of themselves as people (see [23,24]). Rehabilitation agency is a broader theoretical concept than traditional rehabilitation concepts such as function, performance or activity, which are widely used in older adults' client-centred rehabilitation in geriatric and physical medicine.
Utilising these theoretical considerations related to agency, we seek to contribute to the discussion and development of efficient gerontological rehabilitation for older adults, a practice that is widely used in Finland. Gerontological rehabilitation is defined as a multidisciplinary approach toward older adults' rehabilitation in which the psychosocial individual and cohort-based life course perspectives are, as significant elements of rehabilitation, combined with geriatric and physical medicine [2,20,25]. We propose that an agency-centred approach could provide a broad perspective in understanding older adults' individual lives and the rehabilitation process as a part of their overall life situations within their life courses. Furthermore, such an approach could encourage older adults to adopt achieved rehabilitation outcomes in their home settings after institutional rehabilitation interventions and in this way improve the continuity and effectiveness of rehabilitation.

Current practices and challenges in rehabilitation interventions
The goal of rehabilitation for older adults has generally been defined as an enhancement of their functional capacity in their everyday activities and support for independent living in their home settings as long as possible [2,11,[26][27][28][29]. At the same time, traditional medical and physical rehabilitation includes the relatively self-evident assumption that people are selfdirected enough to describe their rehabilitation needs and goals. In these goal-setting processes, older adults also ought to be self-guided in identifying and expressing their capacity as well as their perceived problems in everyday tasks and psychosocial activities. However, it is not self-evident that older adults are self-directed actors. Furthermore, it is often challenging to transfer the good practices of rehabilitation to older adults' everyday lives in home settings after institutional interventions [11,19,25,27,30]. As a solution, client-centred rehabilitation has been applied in the past decades [27,31,32]. The client-centred principle has been assumed to enhance older adults' self-directedness in their everyday activities and independent living in home settings [33,34]. However, the concept of self-directedness has been criticized for not acknowledging that all people are interdependent and that they are not able to autonomously control their life courses within unrestricted contextual constraints [35,36]. In addition, the dominant rehabilitation practices are viewed, along with specialized professionals, as reproducing tacit power over individuals [33,37]. What is lacking in client-centred rehabilitation is the life course perspective when identifying older adults' perceived rehabilitation needs and designing meaningful interventions with individually designed durations, methods and contexts. Groupbased rehabilitation can be seen as one solution to these challenges.
Group-based interventions are designed to supplement individual rehabilitation interventions in order to fulfill older adults' social needs to, for example, share their life and cohort experiences, and to process their life courses with their group members [11,19,27]. However, not all group-based activities and interactions are self-evidently supportive for all older participants [17,38]. In many group-based rehabilitations, the focus has not been on group processes, but on the parallel, individually tailored intervention process. In recognition of the challenges described, this study investigates the group-based rehabilitation intervention of older adults within the framework of an agency-centred approach.

Theoretical considerations of agency
There are various disciplines with different theoretical approaches to agency [23,[39][40][41]. These can be divided into personal (individual) and social (contextual) perspectives [42,43]. In the domain of gerontology, Wray [44] suggests that agency is a personal phenomenon with creative, generative, and relational processes. Through this kind of agency, older adults handle issues from their everyday lives, implying that agency with empowerment is directly related to successful ageing. From the perspective of environmental gerontology, Oswald and Wahl [45] emphasize that, at the behavioral level, agency is about reactive and proactive aspects of using, compensating, adapting, retrofitting, creating and sustaining places, meaning the focus is on the contextual or environmental dimension of agency.
In multidisciplinary rehabilitation research, the personal dimension of agency has been applied and studied using more familiar concepts such as activity and participation [46][47][48]. At the same time, in rehabilitation studies [26,[49][50][51] the contextual dimensions of agencysuch as the home, care unit and community [52,53] have been considered to be parallel rehabilitation interventions. However, sociological or educational theories of human agency have rarely been used as a theoretical frame in rehabilitation research. Only a few studies have focused on older patients or their caregivers from a sociological frame of reference [54][55][56]. According to Emirbayer and Mische [41] as well as Berger [56], the disabled aged body can be seen as an essential part of medical and physical rehabilitation, but also as part of the social and institutional experience of rehabilitation through shared age and cohort experiences with peers. Thus, rehabilitation agency and the institutional rehabilitation structure with its interventions are realized through embodied action and experiences. In cases of perceived illness and disability and, moreover, in aging and rehabilitation, older adults may have a unique opportunity to realize, or even to discover, their subjectivity from previous life course stages such as 'who I have been' as well as 'who I might become' [56].
In this study, a subject-centred sociocultural (SCSC) approach [23,24] to agency is utilized to answer the challenges of gerontological rehabilitation for older adults described above. Through this approach, an older adult can be seen as an individual agent and the rehabilitation intervention as a social (institutional and cultural) context, and these are both understood to be analytically separate but mutually constitutive and interdependent.
To be more precise, we utilized features of the SCSC approach to agency [23,24] considered to be significant characteristics of agency in a working context. These include the following: (i) agency is manifested when a subject exerts influence, makes choices, and takes stances in ways that affect his/her life situation and identity; (ii) the practice of agency is closely intertwined with subjects' identities, comprising commitments, motivations, and interests; (iii) subjects' unique experiences, knowledge, and competences function as individual resources for the practice of agency; (iv) agency is always exercised within certain historically formed sociocultural and material circumstances, and it is constrained and resourced by these circumstances; and (v) in the examination of agency, individual and social entities are seen as analytically separate from each other but also mutually constitutive and closely interdependent. In this study, this understanding of agency was applied in the study of the rehabilitation agency of older adults.

Research aim
This study aims to explore the rehabilitation agency of older adults. The research questions are framed as follows: [1] What forms of rehabilitation agency can be identified in the rehabilitation of older adults? [2] How are these forms of agency connected to the individual life courses of older adults and to the peer relations within the group-based rehabilitation?

Research context
The data for this study were derived from a wider research and development project (IKKU) on gerontological rehabilitation funded by Kela, The Social Insurance Institution of Finland [25]. This project aimed to promote independent living in home settings for older adults. The inclusion criteria of the participants consisted of the following: age 74 or older, living at home without any permanent home services, officially diagnosed musculoskeletal problems with symptoms in their lower limbs and body, which created risks for their everyday activities and independence. A total of 369 older adults passed this official application process. Altogether, 40 rehabilitation groups (each consisting of eight older adults mainly from one municipality) were implemented in six different rehabilitation centers.
The structured one-year rehabilitation intervention included four periods in rehabilitation centers (each from three to ten days), two individual home visits (one at the beginning and one during the final part of the one-year rehabilitation intervention), and two group-based intervention days and one individual one in the older adults' local surroundings. During the rehabilitation intervention, older adults had both personal and group-based home tasks (e.g. writing a diary, physical exercises). The periods in the rehabilitation centers comprised mainly group-based interventions (physical training, psychosocial counseling, and health promotion discussions) with some individual rehabilitation sessions (functional assessments, goal setting, and evaluation sessions of rehabilitation outcomes). Ethical approval for the IKKU project was obtained from the Ethics Committee of the Hospital District of Helsinki and Uusimaa (HUS).

Data collection and participants
The data were collected with non-participant observation [57,58]. According to the previous research on older adults' rehabilitation [11,19,27], this particular method enables the 'real' phenomenon of older adults' rehabilitation to be captured along with the benefits for their everyday life (e.g. in comparison with questionnaire studies). All six rehabilitation centers were informed of the available times of data collection, and the rehabilitation groups to be observed were selected at random by the rehabilitation centers' staff members. Altogether 18 rehabilitation groups were involved in the IKKU project.
The data for this study included all the final individual goal evaluation sessions observed by the first author within these rehabilitation groups. Consequently, the data consist of 38 final individual goal evaluation sessions of older adults. Of the 38 participants, 32 were women, 24 were widowed, and 11 were married and living with their spouse at the time of the study. Two were not married and one was divorced. Thirteen rehabilitation professionals (six were physiotherapists, five public nurses and two occupational therapists) conducted the goal evaluation sessions. In the sessions, the older adults discussed the entire rehabilitation intervention with their personal counselor, and evaluated their individual rehabilitation outcomes. All the professionals were instructed to use the Goal Attainment Scaling (GAS) method for documenting and evaluating older adults' goals for the rehabilitation process [59,60]. The method formulated the main schema in the structures of the sessions, which lasted for approximately 40 minutes (range: 30-60 minutes).
The data collection was conducted by the first author, an experienced observer with a background in occupational therapy and gerontology as well as multidisciplinary expertise in working with older adults in various health care and rehabilitation contexts. As a non-participant observer, she located herself behind the older adults so as not to disturb the interaction. She did not interfere with the goal-evaluating discussions except for in the situations where she was directly invited to do so. Observations focused generally on the rehabilitation discussion and particularly on the accounts given by the older adults. The handwritten field notes included rough sentences that followed the schema of the sessions from start to finish, including expressions and issues that were raised and discussed, such as individual decision-making regarding older adults' everyday life, including personal notions of themselves, the negotiation of achieved goals and their estimates, and the summary of rehabilitation outcomes.
The transcribed data resulted in 158 pages of transcribed observation data. Thus, even though the sessions were not audio-recorded (because of the desire for minimum interference), the data were representative. For the purpose of this study, the raw data were organized and summarized by the first author. The overviews generally included older adults' descriptions and evaluations of the rehabilitation process as well as their role in this process. The overviews also encompassed older adults' accounts of their present situation, everyday activities in various environments, social relationships, individual changes in everyday tasks and roles, officially set goals as well as the perceived outcomes of the rehabilitation.

Data analysis
The written data were analyzed using qualitative content analysis [61]. This method is particularly suited to studying sensitive and unknown phenomena [62], such as the rehabilitation agency of older adults. Our analysis was not comprehensively inductive, since we also applied an SCSC approach [23,24] to investigate the rehabilitation agency.
For the first research question (identifying the forms of older adults' rehabilitation agency), the analysis process was conducted in three phases: preparation (coding), organizing (grouping, categorisation), and abstraction [61]. In the preparation phase, the data were read and re-read to obtain a general overview of the data by making notes and marks. Alongside this phase, the SCSC approach [23,24] to agency was used as a guide to identify and decide the tentative units as codes of analysis. As a consequence, the coding focused on how the older adults influence and make choices concerning their everyday lives, including notes about themselves. Afterwards, all of the identified material was grouped through comparison into four categories illustrating an older adult's activities and choices regarding (a) a stable or unchangeable life situation, (b) a tentatively and selectively changing life situation, (c) a process toward a changed life, and (d) a comprehensively changed life. Finally, these categories were abstracted and, with the aid of researcher triangulation, four rehabilitation agency forms of older adults were identified and named.
To answer the second research question, the analysis focused on how individual life course and peer relations were connected to the identified forms of rehabilitation agency. The selection of these main categories was informed by an SCSC approach [23,24], which suggests that both the subject (examined as individual life course in this study) and social-cultural (examined through the peers in the rehabilitation context) dimensions are important to address when attempting to understand rehabilitation agency. First, the analysis addressed the presence of individual life course for the identified agency forms. In this phase, four subcategories under this main category were identified from the data: (i) the whole life course with past, present and future, (ii) the present life situation, (iii) selected issues of present life situation, and (iv) current single rehabilitation moment or session.
Second, the meaning of peer relations was identified in terms of how other older adults were perceived as supporting individual agency or how they were acknowledged as peers in the group-based rehabilitation or as people with similar life situations. Altogether four subcategories were identified: (i) a peer group with received and given support, (ii) a peer group with received support only, (iii) received support from a single peer and (iv) no meaning of a group or single peer. Consequently, the issues addressed by the second research question (individual life course and peer relations) enriched the understanding of the four rehabilitation agency forms of older adults in an intervention context.

Findings
As an answer to the first research question, four forms of rehabilitation agency were identified: renewable, widened, selective and fractured (Table 1). These forms varied from each other in terms of the choices and decisions that the older adults made regarding their everyday lives. They also differed from each other in terms of the older adult's individual life course perspective and the peer relations within the group-based rehabilitation (research question 2). Next, these four rehabilitation agency forms of older adults are described in more detail as a response to both research questions.

Renewable rehabilitation agency
Renewable agency meant that the older adults made comprehensive changes in their lives ( Table 1). The older adults enacting renewable agency consisted of one male and nine females (n ¼ 10). They described the completed rehabilitation intervention from the whole life course perspective, and how their actions were strongly and mutually influenced by the peer group during the one-year group-based rehabilitation. Their agency was empowered by the shared historical cohort experiences, through which they achieved a new perspective on, or distance from, their current as well as their previous life situations. They also reconfigured their life structures for the future.

Whole life course in renewable rehabilitation agency
The whole life course of older adults with renewable agency was addressed during the rehabilitation intervention as a meaningful one-year process. These older adults had discussed their previous life as comprising many difficult, transformative events, such as the Second World War or being an immigrant moving to a new living environment during the war and feeling like oppressed foreigners for their entire adulthood. The older women with long marriages described lives with a dominating husband or other relatives, having many children without any family planning, no individual hobbies or opportunities for personal timeeven their lack of a satisfying identity as an individual person and of a vocational career outside of the home. During the rehabilitation year, they had made changes in their appearance and started to rebuild social relationships with old friends and to rediscover old hobbies, such as dancing and handicrafts. They had also started to plan their future by, for example, drawing up their wills for relatives, significant others or society. For these older adults, the rehabilitation had been an opportunity to analyze and reflect on their own life and to express their inner thoughts. As one participant stated: 'I have found the real me, my inner me, which I had lost during these past decades.' These personal reflections provided the energy and motivation to change their previous attitudes and reformulate their whole present view and future orientation to life as an older person. Some older adults recognized their broader physical or psychological capacities, not only the losses of old age. This new attitude to self, and also to becoming an older person, gave them a new kind of freedom to accept, for example, individual help from outsiders, to use official home services and have some amusements for themselves. The illnesses, symptoms and pains were the same, but they perceived themselves as being healthier, sleeping well, and having a happy mood or a lighter attitude toward their present and future lives. At the same time, the use of polypharmacy was evaluated and often reduced.

Peer group as a crucial factor in the rehabilitation context
For these older adults with renewable agency, the process with others was a huge surprise because they had assumed that rehabilitation would be a more physical process, strictly led by professionals, and more like how traditional war veteran rehabilitation has been organized and produced in recent decades. The other older adults were seen as a supportive peer group, not only as a part of an obligatory group-based service. Suddenly, they had found new friends with whom they could socially re-design their everyday lives as older persons. The older adults with renewable agency reported that the peer groups also had an effect on their everyday lives after the rehabilitation intervention, because the groups had decided to begin regularly meeting with each other as a friend group.

Widened rehabilitation agency
The second identified agency form of older adults, with widened agency, was the smallest group of this study, comprising one male and three females (n ¼ 4). In comparison to the renewable agency group, they were part of the path to larger transformative changes in their lives. They had started to change, step by step, their everyday lives during the one-year rehabilitation. For all of them, the rehabilitation intervention changed their attitude toward their severe illnesses, and toward their roles in their everyday activities as older people, not only as older rehabilitees. Re-evaluating and reflecting on these activities with progressive changes was partly obligatory because of the risk to themselves or to others, such as falling during household activities (Table 1).

Individual life of older adults in present life situation
For two older adults in this group, the new reformulated role as a spouse was especially important. They now oriented themselves toward new activities outside of the home without feeling guilt over this change; this orientation was partly due to the renewal of their driving licenses and having a new kind of independence and autonomy in a wider life context. Additionally, one older adult had been a lifelong independent entrepreneur and experienced a difficult adaptive process due to a declining capacity to work alongside increasing financial demands. During the rehabilitation, this person had started to evaluate and pre-design the forthcoming working life with some peers without a decision to get out of business or retire yet.
The older adults with widened agency discussed their current experiences in the rehabilitation session, saying, for example, 'I've realized that I've become old.' During their individual rehabilitation processes, they had divided their declining physical capacities, lower mental energy and demanding everyday activities in different ways and by different routines, listening to their symptoms and avoiding repetitive stress or pain. They came to perceive themselves as older people with a normal aging process, not only as people with chronic diseases. As a consequence, they did not evaluate their situations as negatively as they had previously.

Supportive peer older adults from the same age cohort
For these older adults with widened agency, the peers provided either slightly warning or positively supportive feedback through their own examples as an older person in the same life situation or as a member in the same age cohort. This widened process, along with individual changes in attitude, took time, so that during the rehabilitation year these issues were the focus of discussions in sessions with their individual counselors and occasionally in group settings. These older adults gradually accepted the basis of their situation, but not, on the whole, as concrete actions or decisions on how to continue their everyday life after rehabilitation. They used reflective expressions, describing how, for example, 'these solutions might be temporary'. Particularly in long-lasting marriages, everyday life might also result in crises if the spouse finds new ways to continue his/her everyday life.

Selective rehabilitation agency
Nine women and one man (n ¼ 10) were identified as exhibiting selective rehabilitation agency. These older adults seemed to be partially formal rehabilitees and partially individualistic actors in their lives. They shared only select parts of their everyday lives and of the recently completed rehabilitation. For example, if there was a plan for hip surgery, they talked as if they had taken in all the information concerning that procedure and they ignored other advice (Table 1).

Individual life as separated from rehabilitation intervention
The older adults with selective agency chose which issues were beneficial to express to professionals. Following these schemas, they formulated their rehabilitation evaluations either silently or critically, and expressed how they would continue with their everyday lives. Because of this approach, they were not disappointed if the official rehabilitation goals were not achievedthese outcomes did not seem to be relevant for their private everyday life. Some of these older adults also changed their rehabilitation goals many times during the rehabilitation year.
Before and also during the rehabilitation process, they received a range of advice or support from their counselors, but their reactions were always the same: 'I have heard this already … I know this and I have received these gym leaflets and instructions before.' They were not dependent on other people's advice. During the rehabilitation process, they were not truly engaged in the process, for example they had been late to some sessions because of their personal appointments or other interests, such as swimming, handicrafts and phone calls.
Neither the previous life courses of these older adults nor their present period of older age seemed to be in focus during the rehabilitation intervention. They preferred to express their present situation in life and in rehabilitation with this sufficient, conditional approach. Despite that preference, some of them rediscovered their old hobbies or found new ones, such as writing to the local newspaper, working out at fitness clubs or using new technology, but they carefully evaluated these before accepting them.

Older adult peers as a one-sided resource
For the older adults with selective agency, some of their peers were occasionally an important resource, helping them to continue the rehabilitation through moments of decreased motivation. These peers served as an extra positive dimension, one they would never have had without this rehabilitation year. However, the group sessions with shared tasks were described as not affecting their everyday life. They did make comments such as 'this rehabilitation group has been the best', but this only applied during the official rehabilitation year, not after. They intended to continue their everyday lives in their previous relationships and networks within their home environments. At the same time, they seemed to be surprisingly eager to bring up issues regarding the health, physical or social situations of their rehabilitation peers with their counselors.

Fractured rehabilitation agency
The mode of the older adults with fractured rehabilitation agency was the narrowest in comparison with the other forms of rehabilitation agency. This form applied to three males and eleven females (n ¼ 14). These older adults with fractured agency were confined to an official role, just as if they were polite, humble guests in the structured, scheduled sessions carried out by professionals. The sessions did not cohere into any individual or group-based process for them. Instead, they were experienced more as a series of single, unrelated sessions, one after another (Table 1). Their agency was like a fractured entity under an authorized rehabilitation structure, which is why the individual and contextual perspectives will be described next as the same entity.
Individual life and the peer group as a non-significant entity The sessions of older adults with fractured agency consisted of traditional medical and physical rehabilitation in clinical practice, concerning their health situation and physical deficits. They described their current diseases and functional deficits, comparing them to their previous situations in health care institutions where they had received rehabilitation or care in previous years. In this way, they seemed to express how they have learned and adapted to use services in these institutional contexts.
During the rehabilitation, the older adults' heath situation or physical condition had either declined or remained the same in comparison to the results of the first measurements. The participants themselves used expressions such as 'everything [physical condition] is like it has been, life will be the same, and everything has been good [in the rehabilitation]'. However, in some cases there had been momentary recovery in their physical or functional condition during the rehabilitation, but these changes were emphasized mainly by their counselors.
These older adults with fractured agency had also lost family members or friends and they possessed deteriorated social networks. They seemed to be exhausted or not to know even how to react to these changes. They considered these issues as not being part of the official appointments, saying, 'I am sorry that I'm using your time for this … What was it that you were supposed to do with me?' For them, the group-based rehabilitation process with other older adults was concretely visible only when they participated as individuals in the same sessions, not as a possibility for social participation and new kinds of relationships. The social relationships were a part of rehabilitation because of the obligatory, official homework or tasks counselors had asked them to do in their home settings. Some of them expressed disappointment or contradictory feelings if they felt some other group members were too eager to make contact or had become too inquisitive.
Individual previous life courses and individually perceived old age were not brought up in the final discussions. On the contrary, these older adults expressed how rehabilitation sessions were official tasks in their daily programs, scheduled into their timetables by professionals. They followed these official programs as obedient patients, clients or rehabilitees. In this concrete but convenient way, they seemed to confirm themselves as passive objects in a special institutional structure or context more than as initiative subjects.

Discussion
This study revealed renewable, widened, selective and fractured rehabilitation agencies of older adults.
These four forms of rehabilitation agency illustrate how the older adults influenced and made different choices and decisions regarding their everyday lives, which also encompassed notions of themselves. The older adults with renewable rehabilitation agency processed their previous life, made choices and decisions concerning their present life and actively constructed their future after rehabilitation. These agentic actions were mutually influenced by other older adults. In comparison with renewable agency, the older adults with widened rehabilitation agency were merely halfway to this kind of transformative change, concerned mainly with their present situation. They actively received support for this widening perspective from other older adults as peers with a similar life or health situation.
The older adults with selective rehabilitation agency focused on certain current personal life events (e.g. an illness, functional constraints or living conditions) with their limited actions in cooperation with a single, chosen peer. Finally, in the case of fractured agency, the presence of one's individual life course and the meaning of peers were most modestly manifested. For them, the rehabilitation process was a series of single events and actions, and there was no peer group or a single older adult to receive or give support.
These findings contribute to the discussion on developing gerontological rehabilitation services for older adults. On the basis of our study, we suggest a new kind of agency-centred gerontological rehabilitation. Through this approach, client-centred and group-based rehabilitation interventions could be formulated more precisely and efficiently to respond not only to older adults' various needs in medical and physical rehabilitation, but also to their psychosocial needs in their individual life course situations as older persons.

Older adults' individual life course agency in client-centred rehabilitation
The first basic assumption of client-centred rehabilitation is that the rehabilitees are naturally committed to sharing all necessary information with professionals concerning their health, functioning and everyday demands. Furthermore, they are expected to be motivated and committedand to do their best during the structured intervention. Through this normalized and causal process, they should achieve the goals which have been set together with the professionals at the beginning of the rehabilitation [20,63,64].
The findings presented here show that the rehabilitation intervention is not a continuing, progressive, productive, or even causal intervention process for all older adults. Instead, it is a personal, temporally and socially divergent experience. For the older adults with fractured agency, the rehabilitation seemed to be a series of single and random sessions carried out by professionals. It seemed that they did not have personal resources or capacity to respond either to the demands or possibilities of rehabilitation, but they accepted the institutional rehabilitation as a selfevident entity. It has previously been found [65] that older adults with mild psychosocial demands, memory problems or physical frailty with exhaustion faced particular challenges in constructing a life continuum through which they could achieve their rehabilitation goals in an institutional rehabilitation context. However, the other older adults with renewable, widened and selective agencies could be assumed to have better physical, mental and cognitive resources in utilizing the resources offered by institutional rehabilitation. A further assumption is that their sense of agency has been strong during their life course, or at least in some periods and phases, with positive ramifications for their new situations (see [66,67]). Additionally, older adults with selective agency chose to address medical issues within the rehabilitation intervention, which could be called 'health agency' [67] or 'physical agency', and it could be seen as suitable for a traditional medical and physical rehabilitation context.
It has also been found that older adults, in comparison to younger adults, prefer a stronger process focus over an outcome focus (compare [68][69][70]). In this study, single outcomes were perceived as being the most important for older adults with selective agency. Older adults with renewable and widened rehabilitation agency utilized various kinds of rehabilitation events for reflecting on their life course as a whole. For them, both the rehabilitation processes and the outcomes were meaningful. On the contrary, older adults with fractured agency seemed to manage the formal rehabilitation program with official requirements without any perceived process or outcome effects.
In terms of an agency-centred approach, we suggest that the focus of older adults' client-centred rehabilitation should be not only on the physical outcomes or psychosocial benefits of the intervention, but moreover on the supportive reflection of old age as a part of one's life course and agentic background. In these reflective processes, older adults could learn from others in the same age and life situation, share their experiences, and gain good examples as well as to consider the future demands.

Peer relations in group-based rehabilitation
The second basic assumption of rehabilitation is that rehabilitees are self-conscious and self-directed in group-based interventions with other rehabilitees. However, rehabilitees often do not respond to these interventions in an expected or uniform way [14,16]. In this study, other older adults gave each other various perspectives on being an older person in a group-based intervention when they perceived their individual rehabilitation process differently (see [70,71]). However, the meaning of peers differed depending on the individual agentic perspective on one's own life course and capacity to respond to group-based interventions. Different forms of older adults' agency ought to be recognized by rehabilitation professionals, so that group-based intervention could be designed to be as suitable and efficient as possible according to individual needs. At the same time, a group-based approach for the rehabilitation of older adults could be developed into an acknowledged intervention once an individual rehabilitation intervention has already been formed.
Agency-centred gerontological group-based rehabilitation could provide opportunities for accommodating or modifying the typical health-or function-oriented goals of traditional geriatric rehabilitation to other, more individually meaningful and tenacious life course goals on the activity and participation level. This kind of approach would not be so based in the poor health condition or declining functions of older adults and could also motivate them to pursue an individually meaningful everyday life in a home context after the rehabilitation interventions throughout the later years of their lives. However, the individual client-centred and group-based rehabilitation as a parallel intervention need to be further studied in order to identify the best and most suitable practices for all kinds of older adults with differing life course agencies as well as their needs to change or reflect them.
At its best, group-based gerontological rehabilitation, as an identified social structure, could be seen as a collective, networking intervention influencing group members' lives in an efficient and multifaceted wayan effect that professionals, as younger adults, could never anticipate or provide [72,73]. In future studies, this collaboration between older adults and their personal counselors should be studied in more detail with an agency-centred approach. A further issue to examine in interventions would be how peers in the rehabilitation context, as a supplemental intervention dimension, are related to this collaboration.
Overall, if rehabilitation is designed as a part of older adults' individual life courses with their agentic identities, we could enhance both individual-and group-based processes as part of interventions that are effective mutually as well as in parallel. Furthermore, we could ensure the continuity of positive outcomes in everyday environments after rehabilitation as a part of these individual life projects. The agency-centred approach within the field of rehabilitation research also offers possibilities to broaden the scientific basis of multidisciplinary rehabilitation (compare [14][15]).

Conclusions and limitations
This study relied on observation research and content analysis, an approach that has been considered fruitful when researching an unfamiliar area, yet it still contains a number of potential limitations [74]. First, observations by two researchers could have enriched the data, but such a process would have been timeconsuming and required more arrangements and further disruption of the rehabilitation sessions. Second, even though individual rehabilitation sessions of only 38 older adults were observed, the amount of the data was sufficient enough to capture the main and essential dimensions of the phenomenon being examined [75]. Third, despite the fact that the collected observation data were handwritten without any audio recording, the data were multifaceted. To enhance the trustworthiness of this study, the process has been made as transparent as possible, with the first author cooperating with authors from different scientific backgrounds (adult education). Additionally, if focus group discussions had been used as the data collection method, older adults' expressions concerning themselves, their life course and peer relations in groupbased intervention would have been constructed differently.
Because each of the identified agency formsrenewable, widened, selective and fracturedseemed to be logical for the older adults themselves, they ought to be elaborated and identified in rehabilitation practices. In this way rehabilitation could be designed to include suitable collaborations and interventions with meaningful durations, schedules and contexts for these older adults, ones in which they are the subjects of their own lives.
Even though more research is required, the findings are innovative enough to broaden the interdisciplinary approaches in rehabilitation contexts [13][14][15]. The agency-centred approach could create an understanding of how efficient rehabilitation interventions can be dependent on each older adult's whole life course, and how that person is never only a patient, a client or a rehabilitee in a structural, institutional context. Instead, each person is an individual agent of their life and a member of their own cohort.