Change in the psychological self in people living with dementia: A scoping review

Alterations in one ’ s sense of self are often considered a significant psychological symptom of dementia. However, the self is not a unified construct; it consists of a set of closely connected, yet substantive, manifestations which might not be equally impacted by dementia. Recognising the multidimensional nature of the self, the current scoping review aimed to explore the nature and scope of the evidence demonstrating change in the psychological self in people living with dementia. Adopting a cognitive psychological framework, a hundred and five (105) quantitative and qualitative studies were reviewed, and findings were organised into three main types of self-manifestations: high-order manifestations, functional aspects of the self, and foundational manifestations. Overall, the results show that although there are alterations in some of these different manifestations of the self, these do not imply a global loss of selfhood. Despite notable cognitive changes during dementia, it seems that preserved aspects of self may be enough to compensate for potential weakening of some self-processes such as autobiographical recall. Better understanding alterations in selfhood is key to addressing psychological symptoms of people living with dementia, such as feelings of disconnection and reduced agency, and may inform new pathways for dementia care interventions.


Introduction
Over 46 million people live with dementia globally, a number expected to triple by 2050 (Mukadam et al., 2019).Although there are different forms of dementia with distinct symptomatology, the pronounced impact on cognition, notably the progressive disintegration of memory, is a common characteristic (Hutmacher, 2021).Changes in individuals' behaviour and psychological experiences are also recognised as being at the core of dementia (D'cruz, 2021).It has been reported that one of the most distressing psychological symptoms of dementia, for people living with dementia and caregivers, is alterations in one's sense of self (D'cruz, 2021).Having access to a life narrative is considered central to the human experience of selfhood (Jetten, Haslam, Pugliese, Tonks, & Haslam, 2010).Given that self and memory are closely linked, it has been suggested that fragmented episodic memories and gaps in autobiographical self-knowledge can result in the 'weakening' of selfhood in people with dementia (Addis & Tippett, 2004;Jetten et al., 2010).However, the self is a complex construct with many different manifestations, among which self-related memory is only one (Hutmacher, 2021).Following a cognitive psychological framework, the human experience of 'selfhood' can be conceptualised here as the knowledge of having a unique identity and intentional physical presence in the world, accompanied by a strong sense of continuity overtime (Ross, Anderson, Campbell, & Collins, 2011).
When exploring dementia's impact on self, terms such as 'total loss', 'impairment' and 'erosion' of selfhood have been historically used in the literature to describe the pronounced changes in the self-experience of people living with dementia, implying the existence of a unitary construct of self (George, 2010;Kontos, 2014).Such descriptions can promote perceptions of complete disintegration of selfhood, exacerbating the stigma as well as fear associated with a dementia diagnosis (Hillman & Latimer, 2017).Consequently, modern opinions caution against viewing the self as a 'monolithic entity', since this conceptualisation could be considered oversimplified, failing to capture the real implications of dementia on the sense of self (Strikwerda-Brown, Grilli, Andrews-Hanna, & Irish, 2019, p.2). Thus, recognising the multidimensional nature of the self, the current scoping review aims to explore the nature and scope of the evidence demonstrating change in multiple aspects of the psychological self in people living with dementia.
Although self-related functions have been extensively explored in the dementia literature, the inherent difficulty of a precise definition of the self has contributed to potential misinterpretations about changes to the sense of self in people with dementia (Tolhurst, Bhattacharyya and Kingston, 2014).Although we experience a unified sense of self, the self might be considered as a set of closely connected, yet substantive, aspects, from which unity emerges (Klein & Gangi, 2010).Accordingly, it has been suggested that the diverse manifestations of the self have a hierarchical relationship, with self-processes developing earlier providing the foundation for the emergence of higher-order processes (Bomilcar, Bertrand, Morris, & Mograbi, 2021;Summa, 2014).Selfrelated functions linked with sophisticated cognitive processes could be characterised as higher-order manifestations of the psychological self; these include autobiographical memory, sense of having a lifenarrative and self-knowledge (Huntley et al., 2021;Mograbi, Huntley, & Critchley, 2021).Nevertheless, there are also aspects of the self that might not be necessarily available to conscious reflection; this is particularly notable in the processes that are based on the awareness of one's body as their own, having a causal influence both in the physical and social world, such as agency and embodiment, which could constitute the foundational manifestations of the self (Huntley et al., 2021;Mograbi et al., 2021).
Although it is theoretically reasonable to distinguish between highorder and foundational manifestations of self, this might not be an unequivocal distinction; different levels dynamically interact with each other to form a network of cross-connections (Summa, 2014).This is what Summa (2014, p. 484) refers to as the 'grey zones' in between the two types of manifestations.The self has a central functional role in human cognition (Cunningham & Turk, 2017), underpinning several cognitive processing biases that occur automatically, without conscious awareness, and involve the cognitive organisation of self-related information.For example, humans preferentially attend to self-related information, which is elaborated and organised within existing semantic and autobiographical self-knowledge, sustaining the self-system (Rathbone et al., 2019).Such processes might be considered to operate in the 'grey zone', uniting high-order manifestations of self-knowledge with foundational aspects of selfhood, such as the agentive experience of an event.Thinking of the different manifestations as existing along a continuum, functional aspects of self might connect the high-order and foundational manifestations together allowing for bidirectional interactions.
The existence of different aspects of the psychological self demonstrates that dementia's impact on selfhood is unlikely to be absolute.The most recent systematic review of the psychological self in dementia, conducted by Caddell and Clare (2010), concluded that self and identity are mostly preserved throughout the course of dementia, at least to some degree, with qualitative aspects of the subjective experience of self remaining undisrupted.However, although it is true that some aspects of the self remain intact in dementia, manifestations of self involving complex high-order processes are susceptible to negative impact (Weiler, Northoff, Damasceno, & Balthazar, 2016).Since Caddell and Clare (2010) review, evidence for this conclusion has continued to grow.For instance, the consequences of declining episodic memory on selfhood in people living with dementia are well-documented; as selfmemories fade and self-knowledge of one's life narrative disintegrates, access to self-related autobiographical information is compromised leading to alterations of perceived sense of self (El Haj, Antoine, Nandrino, Gély-Nargeot, & Raffard, 2015).Additional findings show that self-knowledge, and subsequently self-continuity, can become outdated, often reflecting young adulthood, as people with dementia might not be able to update information about their current self and thus rely on past self-representations (Bomilcar et al., 2021;Klein & Gangi, 2010).Taken together, it has been suggested that only self aspects related to memory might be weakened in dementia while foundational aspects of self might be preserved, including self-related experiences of living at the present (Hutmacher, 2021;see Strikwerda-Brown et al., 2019 for a narrative review).
Overall, alteration in one's sense of self is undoubtedly a significant psychological symptom of dementia, resulting in the experience of fragmented selfhood.Emergent evidence seems to reach agreement concerning the complexity and multidimensional nature of the psychological self, demonstrating that its different aspects might not be equally impacted by dementia.However, existing empirical research on the impact of dementia on selfhood consists of distinct studies employing diverse methods as well as often focusing on individual manifestations of the psychological self.To achieve a more global understanding of the impact of dementia on the psychological self, it is necessary to recognise the multidimensional nature of the self.The cognitive psychological framework set out above (spanning foundational, functional and high-order self-processes) offers a useful framework to bridge this diversity, but has not commonly been employed.Thus, although researchers have answered Caddell and Clare (2010) call, providing further research to better understand how the psychological self is impacted by dementia, this research has not been systematically reviewed and synthesised within a single framework.
To address this theoretical gap, the scoping review method (Arksey & O'Malley, 2005) was used to explore dementia's potential impact on different aspects of selfhood, which have been studied using distinct conceptual approaches and methods across the literature.The aim of this review is to gain greater understanding of the patterns of dementiarelated changes in the psychological self.Exploring the literature on selfhood and dementia over the past 12 years, the current scoping review expands on Caddell and Clare (2010) findings by adopting a rigorous systematised methodology, in accordance with the latest Joanna Briggs Institute guidelines, and offering a nuanced exploration of the self using a cognitive psychological framework.Although research included within Caddell and Clare (2010) review could also be usefully interpreted according to this framework, the sheer volume of dementia research necessitated that we apply time constraints, and we felt a focus on more recent studies would be appropriate.We expand Caddell and Clare (2010) broad search terms to include individual manifestations of selfhood, representing the existence of different levels of selfmanifestations, and a focus on the psychological impact of the condition operationalized within a single framework.Adopting a different approach than Caddell and Clare (2010) in our search and inclusion criteria, rather than focusing on the self as a unitary construct, we apply a cognitive psychological framework, conceiving of the self as a complex system with high-order, functional and foundational aspects.This allows us to offer the first systematic exploration of the impact of dementia on the psychological self, recognising the complexity of the self-system.

Method
The conduct and reporting of this scoping review followed the guidelines recommended by the Joanna Briggs Institute (JBI) and the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist was used (Tricco et al., 2018) (Appendix A).The review protocol has been preregistered on the Open Science Framework (https://osf.io/b4t6n).Deviations from the originally published review protocol are declared and discussed in detail in the following sections.

Review question
Recognising the multidimensional nature of the psychological self, what is the nature and scope of the evidence demonstrating change in the psychological self in people living with dementia?

Eligibility criteria
The inclusion and exclusion criteria were formulated a priori of the A. Mentzou et al. screening of relevant articles (Table 2).The exploratory and broad focus of a scoping review involves an enormous undertaking.For this reason, the systematic review of Caddell and Clare (2010) acted as a reference point for the current scoping review rather than attempt to reproduce it.Given that their database search was conducted on the 16th of March 2009, we made the decision to limit the scoping review on studies published over the last 12 years (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020)(2021) to focus only on the newest developments in the field.Recognising the multidimensional nature of the self, the cognitive aspects of the psychological self closely match the aspects of self conceptualised in recent theoretical papers (e. g., Bomilcar et al., 2021;Klein & Gangi, 2010), which share common elements.As we were interested in exploring only cognitive aspects of the psychological self, while also recognising the inherent first-person subjectivity of selfhood (reflected in criteria 7 and 8), socially and culturally constructed aspects of selfhood (e.g., marital identity and romantic relationships, sexual identity, spirituality etc.) were not considered.Anosognosia (i.e., lack of awareness about disease and cognitive difficulties) was also not considered because it has been conceptualised as a complex cognitive process integrating different monitoring abilities such as self-appraisal and performance evaluation (Lenzoni, Morris, & Mograbi, 2020), which are beyond the focus of this review.Following the psychological definition of selfhood as the knowledge of having a unique identity and intentional physical presence in the world, accompanied by a strong sense of continuity overtime (Ross et al., 2011), we considered studies gathering information contained in this definition (self-concept, self-awareness, self-recognition, self-continuity, self-narrative, self-consciousness, personhood, selfhood, identity, agency, autobiographical memory embodiment, life narrative) to be relevant to our aims.

Search strategy
Comprehensive literature searches were conducted by the research team.In the first stage, an initial limited search was conducted in the databases PubMed and CINAHL to ensure that the keywords in the search string were able to identify relevant papers and necessary adjustments were made before the official searches (see Table 3 for examples of search strings used in different databases).Subsequently, six electronic bibliographic databases were searched to identify relevant papers in the published literature: Scopus, PubMed, ASSIA, CINAHL, Web of Science and PsycINFO.These databases were used as they capture literature on psychological experiences of people with dementia.The initial search period was limited to articles published after March 16th, 2009 until November 25th, 2021.Grey literature was also searched, specifically the Open Grey Repository (which has been transferred to DANS EASY data archive) and Google (first 10 pages).Grey literature that met inclusion criteria was accepted for screening.Reference sections of the selected for inclusion full-text papers were examined for additional relevant studies.After the completion of the pre-registered search strategy as well as of a preliminary title/abstract screening, it was noticed that mostly relevant qualitative articles were identified and well-known quantitative articles on the topic had not been detected.For this reason, the following additional qualifiers of 'change' to the original search string were added aiming to make the search terms more inclusive and facilitate the identification of quantitative literature: greater, less, more, better, worse, higher, lower, fewer, compromise, enhance, impair, improve.These were derived from the abstracts of some quantitative papers not identified in the official database search and were considered as suitable to detect relevant quantitative papers.Follow-up database searches were conducted with the inclusion of the new keywords.Although the initial database search was conducted on 25th November 2021, an updated search was also conducted (using the same keywords and databases) to identify new studies published up to October 14th, 2022, ensuring all the latest studies were included in the review, which resulted in four additional articles eligible for inclusion (Berntsen, Kirk, & Kopelman, 2022;Chapman, Philip, & Komesaroff, 2022;El Haj et al., 2022a;El Haj et al., 2022b).Please see Fig. 1 for the PRISMA flow diagram with overall information of article identification and screening.

Study selection
All the results of the bibliographic database searches (official and follow-up) were combined to a unique folder in the referencing software 'EndNote 20'.The PRISMA flow diagram (Fig. 1) gives an overview of the literature search and the screening process.Duplicated articles were initially removed.During level 1 screening, the titles and abstracts of identified studies were screened by the first author, by applying the inclusion and exclusion criteria, to examine their relevance to the current scoping review.To ensure reliability among the research team, a shared 10% of the identified articles of level 1 screening was  A. Mentzou et al. independently reviewed by all the authors.This resulted in 762 articles in the selected pool.During level 2 screening, the full-text copies of the 264 included studies of level 1 screening were screened independently by two authors, by applying the inclusion and exclusion criteria.Another round of reliability checks was performed; a shared 10% of randomly selected full-text articles were independently reviewed by all the authors.This resulted in 26 articles in the selected pool.Potential disagreement or uncertainty at any of the stages of study selection was solved through discussions with the wider research team until consensus was achieved.In the case that a study's full-text was not accessible through University of Dundee's journal subscription, the corresponding author would be contacted to provide an authorised copy.When a study's full-text was not available in the English language, the corresponding author would be contacted to provide a translated copy, if available.All communications were conducted via email and if corresponding authors could not be contacted after two attempts two-weeks apart, the studies were excluded from the screening.All exclusion decisions were recorded and stored electronically.Both included and excluded papers identified in the screening were stored on 'EndNote 20'.
At the end of level 2 screening, 105 articles were included in the review dataset.

Data extraction
Following level 2 screening, the included full-text articles were thoroughly read in the familiarisation process.Data extraction was performed by the first author and checked for accuracy by a second reviewer.A charting table, adapted by the JBI guidelines to mirror the aims of the current scoping review, was used to facilitate extraction of all data relevant to inform the review question, in which the following data was extracted from the full-texts: study aims, sample information, methodology, procedures, measures, results, aspect(s) of self explored and evidence for change in the psychological self.Since this review aimed to explore the nature and scope of evidence demonstrating change in the psychological self in people living with dementia, a thorough descriptive summary of the main extracted results focusing specifically on the aspects of self explored, methods used as well as evidence for change in the psychological self is presented at the 'Results' section below.

Data synthesis
One hundred and five (105) studies met the inclusion criteria for the review (see Table 4 for a summary of key aspects and Appendix B for a more detailed version).Given that different aspects of the self were explored using varied methods and methodologies, the results are organised and presented in three main types of self-manifestations, which conceptually follow recent opinions in the field (e.g., Bomilcar et al., 2021;Summa & Fuchs, 2015) on the existence of different levels of self-manifestations: 1. High-order manifestations (i.e., autobiographical memory, self-knowledge, future self-thinking, perceived change), 2. Functional aspects of the self (i.e., self-reference effect, enactment effect, A. Mentzou et al. pronoun use), and 3. Foundational manifestations (i.e., self-continuity, embodiment, agency).These categories emerged after comprehensively reviewing the included studies together with recent theoretical frameworks of the cognitive psychological self in the literature (e.g., Klein and Gangi, 2010).Clustering together studies exploring the same aspects of self (as identified during data extraction) revealed a clear pattern of the different manifestations of the cognitive self as well as on the evidence for potential changes in dementia, which was indeed in line with recent theoretical frameworks in the field.Details about the time trend of the studies as well as the different self-aspects explored, can be found on Fig. 2.This shows the dominance of autobiographical memory studies over the period of the review, with less sustained activity exploring other aspects of selfhood.

Results
The results are presented in three sections of the different levels of self-manifestations: 1. High-order manifestations (with subsections: Autobiographical memory, Self-knowledge, Future self-thinking and Perceived change), 2. Functional aspects of the self (with subsections: Self-reference effect, Enactment effect and Pronoun use), and 3. Foundational manifestations (with subsections: Self-continuity, Embodiment and Agency).For each subsection, a description of the methods used to explore the given self-aspect is firstly presented followed by a description of the key findings concerning potential change in dementia, with comprehensive in-text citations of the corresponding included articles.Notable in their absence to cognitive psychologists were studies of visual self-recognition.Caddell and Clare (2010) review included ten earlier studies demonstrating that people in later stages of dementia may have difficulty recognising their current self-image (e.g., in mirrors, photographs or videos).Our review indicated this methodology has not been employed in recent years.
Although the above studies provide evidence for disruptions in autobiographical memory, it would be difficult to make definite conclusions of a global loss of autobiographical memory in dementia.For instance, when exploring the existence of a temporal gradient in the deterioration of autobiographical recall in dementia, research has shown that remote memories, usually from young adulthood, seem to be better preserved than more recent memories (Barnabe et al., 2012;Berntsen et al., 2022;De Simone et al., 2016;Kirk & Berntsen, 2018a;Kumfor et al., 2016;Leyhe et al., 2009;Muller et al., 2016;Philippi et al., 2012).Rauchs et al. (2013) also demonstrated better performance for remote events as well as very recent ones (i.e., the previous day).An aspect of episodic recall known as personal semantic memory (i.e., knowledge of facts of one's own past) has also been studied in dementia indicating that this seems to remain intact (Addis et al., 2009;Barnabe et al., 2012;Benjamin et al., 2015;Irish et al., 2018;Martinelli et al., 2013;Meulenbroek et al., 2010).However, this evidence is inconclusive as some studies found diminished performance (Caddell & Clare, 2012;De Simone et al., 2016;Kirk & Berntsen, 2018a;Leyhe et al., 2009).
Ecologically valid sensory cues have also given promising results.More specifically, based on the evidence showing that memories from one's distant past are better preserved in dementia than memories from the more recent past, Kirk and Berntsen (2018b) cued autobiographical memories using concrete objects, historically dated to participants' young adulthood.Object cues (a childhood toy; e.g., skipping rope), compared to verbal cues (name of the toy), produced more memories with higher levels of episodic content.In two studies, selfrepresentations were used to elicit autobiographical responses; however, no benefit of self-cues was found over non-self related cues (El Haj & Antoine, 2017a; Rathbone et al., 2019).Films from people's youth have also been used, having a benefit for autobiographical memory eliciting even more memories in people with dementia than control participants (Rasmussen, Salgado, Daustrand, & Berntsen, 2021).Further extending these findings, Miles, Fischer-Mogensen, Nielsen, Hermansen, and Berntsen (2013) examined autobiographical memory retrieval in an everyday setting as well as in a museum setting which recreated the historic and cultural context of participants' youth.Cues of all modalities (e.g., verbal, concrete etc.) were presented in both settings, with notably more autobiographical memories been recalled in the museum setting (Miles et al., 2013).Finally, sensory evoked autobiographical memories seem to be particularly benefited from cues derived from one's young adulthood (e.g., Baird et al., 2018;Baird et al., 2020;Rasmussen et al., 2021).These studies provide evidence that autobiographical memory performance might benefit from ecologically valid methods.

Future self-thinking
Temporally extended self-representations are not only involved in projecting oneself into the past but also into the future.Thus, a key aspect of the temporally extended self is future self-thinking.In addition to past autobiographical memories, few studies identified in this review have also asked participants to imagine future autobiographical events (Addis et al., 2009;El Haj et al., 2022b & Irish, 2021).People with dementia tend to produce fewer past and future autobiographical events (Addis et al., 2009;El Haj et al., 2022b), with notable similarities having been identified between past and future autobiographical events as participants seem to imagine the future based on often fragmented autobiographical information (El Haj, Antoine, & Kapogiannis, 2015c; El Haj, Antoine, Nandrino, & Kapogiannis, 2015).Difficulties in visual imagery have also been explored in relation to future self-thinking (El Haj, Moustafa, et al., 2019), since remembering the past and imagining the future require to visually imagine experienced events.
Findings demonstrate that participants did not report changes to their current self or feelings of discontinuity to their past self, expressing that they feel they are the same person as before the onset of dementia (Borley et al., 2016;Borley & Hardy, 2017;Caddell & Clare, 2011;Caradec & Chamahian, 2017;Hennelly & O'Shea, 2021;MacRae, 2010).The few longitudinal qualitative studies identified (Johannessen et al., 2018;Steeman et al., 2013) did not report substantial alterations to participants' self; minor changes were described, mostly everyday difficulties and cognitive challenges, but people with dementia do not feel that they have changed in a major way.Exploring the self as a narrative, studies show that people with dementia perceive themselves the same across past and present, providing emotional and reflective life stories to demonstrated resilience and self-reliance against the challenges of living with dementia (Angus & Bowen-Osborne, 2014;Karlsson et al., 2014).Nevertheless, there is also some evidence for perceived change; some participants explicitly state that they do not feel the same person to what they used to be in the past (Borley & Hardy, 2017;Caddell & Clare, 2011), even feeling changed in a fundamental level (Scott, 2021).Other studies also report instances of subjective self losses, which were usually specifically related to past capacities (Caradec & Chamahian, 2017), meaningful activities (Larsson et al., 2019) and previous selfrepresentations (MacRae, 2010).Although changes in the sense of self were described, most participants tend to feel that no major changes had occurred to themselves 'as a whole', they were not radically changed (Borley et al., 2016;Borley & Hardy, 2017;Caddell & Clare, 2011;Caradec & Chamahian, 2017;Larsson et al., 2019).

Functional aspects of the self
Dementia research on the sense of self has not only explored its highorder manifestations, but also the functional role of the self in human cognition.These refer mostly to self-processing that occurs automatically, without necessarily conscious awareness and is related to organisational and elaborative properties of the self.

Self-reference effect
Research has shown that self-related information is encoded and retrieved more easily than other-related or non-social information, resulting in self-referent memory biases.This mnemonic advantage of information encoded with reference to the self can be operationalized by the self-reference effect (SRE), which has demonstrated the unique properties of the self in memory (Leblond et al., 2016).By exploring selfreferential processing in dementia, an insight can be gained on whether the sense of self has a preserved functional role in memory.The SRE studies in dementia, identified in this review, have employed the traditional trait paradigm that involves presenting participants with trait words, which they have to process in relation to self and another person (Genon et al., 2014;Kalenzaga & Clarys, 2013;Leblond et al., 2016) or semantically (e.g., 'Provide a short description of this trait') (Kalenzaga, Bugaïska, & Clarys, 2013;Lalanne, Rozenberg, Grolleau, & Piolino, 2013;Leblond et al., 2016).Afterwards, participants are presented with a surprise memory test in which they need to recall whether a trait word was presented in the encoding phase.Although some participants with dementia experience difficulties completing the SRE task (Leblond et al., 2016), findings show a tendency for a self-related mnemonic benefit in the self-encoding condition compared to the semantic processing condition (Kalenzaga et al., 2013;Leblond et al., 2016) as well as a bias in remembering positive words from the selfcondition (Lalanne et al., 2013).Employing an other-related condition can provide evidence on whether mnemonic advantages are self-specific rather than a depth of processing effect arising from social processing.Critically, when comparing performance for self-related and other-related stimuli, studies have not found an SRE, with participants having the same performance in both conditions (Genon et al., 2014) or higher recognition scores for other-related information (Kalenzaga & Clarys, 2013).In a modified SRE paradigm, Bond et al. (2016) presented participants with words presented either with the participants' voice or another person's voice; no self-related advantage was found.However, for people with dementia, memory for words presented with another person's voice showed less impairment compared to the self-condition (Bond et al., 2016).Taken together, these studies confirm that functional aspects of self relating to memory may be negatively impacted in dementia.

Enactment effect
Expanding on the SRE methodology, few studies identified in this review have also explored the role on self-initiated actions on memory performance.The enactment effect refers to the mnemonic advantage for stimuli that participants have acted upon rather than read or listened (De Lucia, Milan, Conson, Grossi, & Trojano, 2019).The enactment effect task involves presenting participants with verbal stimuli (e.g., action/instruction phrases); in some sessions, participants perform the actions (e.g., 'Open the window') themselves while in others the actions are performed by the experimenter or there is only recall of the stimuli with no enactment (Charlesworth, Allen, Morson, Burn, & Souchay, 2014).A mnemonic advantage was found for subject-performed condition over the traditional verbal task (Charlesworth et al., 2014), but participants with dementia also demonstrated an equal benefit for experimenter-performed condition (De Lucia et al., 2019).Although in both studies participants with dementia had worse memory performance than healthy controls, their memory benefited from enactment but without a self-specific advantage.Similar to the SRE studies, this implies that although social processing still ensures depth of processing relative to other conditions, the strength of self as a cognitive or physical anchor for memories may be negatively impacted in dementia.

Pronoun use
The use of first-person pronouns can be seen as an important linguistic aspect of cognitive self-functioning, which has been conceptualised as an unambiguous reflection of people with dementia's awareness of their self as a distinct and unique entity, reflecting preserved sense of self (Skaalvik, Fjelltun, Normann, & Norberg, 2016).The review found that few qualitative studies have also explored pronoun use in dementia by assessing whether participants would use first-person pronouns during interviews (Batra, Sullivan, Williams, & Geldmacher, 2016;Hedman, Hansebo, Ternestedt, Hellström, & Norberg, 2013;Skaalvik et al., 2016).Findings showed that people with dementia use first-person pronouns frequently in their answers when describing themselves as well as when communicating self-related information (e. g., 'my children'), without experiencing any challenges or confusion (Hedman et al., 2013) and showing no differences compared to control participants (Batra et al., 2016).Successful use of first-person pronouns was interpreted as unequivocal linguistic expressions of the sense of self, expressing one's experiences of being a singular person (Skaalvik et al., 2016).

Self-continuity
Self-continuity, or diachronic unity, refers to the subjective experience of being the same person across the lifespan, despite indisputable developmental changes.It has been suggested that feelings of selfcontinuity are supported by autobiographical memory since recalling the self over time facilitates people to mentally travel back to past events and to relive them in the present, allowing for a coherent and stable sense of self (Tippett, Prebble, & Addis, 2018).The review indicated that studies exploring self-continuity in people with dementia, usually employ a.) questions explicitly enquiring whether participants feel they are the same individuals across time (Clare, Martyr, Morris, & Tippett, 2020;El Haj & Allain, 2020;Tippett et al., 2018), b.) self-statements across past, present and future (Duval et al., 2012) and c.) a self-rated questionnaire (El Haj et al., 2019).The majority of participants reported preserved or slightly altered self-continuity, despite alterations in autobiographical recall and future self-knowledge (Clare et al., 2020;Duval et al., 2012;El Haj & Allain, 2020;El Haj, Boudoukha, et al., 2019;Tippett et al., 2018).Nevertheless, it is important to note that Clare et al. (2020), found that 21% of participants (in a sample of ~1500) reported discontinuity.

Embodiment
Foundational manifestations of selfhood are not only those that the participants explicitly talk about, but also those subjective experiences that one experiences through his/her own body.The concept of embodiment in dementia research suggests that the sense of self in people diagnosed with dementia is preserved at the level of the prereflective body, manifested in habits and purposeful actions (Kontos, 2014).It has been suggested that in dementia, where cognitive processes deteriorate and verbal abilities weaken, the self is manifested through embodiment (Kontos, 2014).This review indicates that the expressions of the self at an embodied level have been qualitatively explored through both interviews (Hellstrom, Eriksson, & Sandberg, 2015;Tatzer, 2019) and observations (Boddington, Featherstone, & Northcott, 2021;Chapman et al., 2022;Robbins, 2019;Tatzer, 2019), which are interpreted by the researchers rather than the participants.
It is suggested that embodiment is manifested through personal objects and their perceived meaning for the sense of self.Handbags and their contents (Buse & Twigg, 2014;Tatzer, 2019) as well as clothes (Boddington et al., 2021;Buse & Twigg, 2016) are seen as extensions of the self, having personal meaning and supporting the 'enactment of self' (Buse & Twigg, 2014, p.14).Personally significant objects constitute ways of self-presentation, reflecting preserved selfhood at an implicit level.Even the maintained participation in household chores (Hellstrom et al., 2015), is seen as an embodied manifestation of the self as people with dementia express who they are through enacted narratives and actions, even when verbal communication is compromised.In the instances when the choice of clothing was not possible, feelings of distress and loss of selfhood were observed, as in the cases of mandatory institutional clothing (Boddington et al., 2021).The corporeal complexity of the self is also reflected in socially embodied ways of expressing the self in people with dementia, through expressive facial expressions and gestures (e.g., expressing disapprovement and social status) and inside jokes (e.g., which might not be obvious to caring staff) (Chapman et al., 2022;Robbins, 2019).Distinctive personality traits (e.g., creativity, adventurousness, resourcefulness) (Frazer, Oyebode, & Cleary, 2012) as well as spontaneous emotionally-motivated comments (Baird, 2019) demonstrate that embodied elements of the self are preserved in dementia.

Agency
Sense of agency can be cognitively defined as the experience of having a causal impact on the world accompanied by a feeling of having control over one's actions.This review shows that different studies have employed varied methods to assess sense of agency in people with dementia.
Some quantitative studies explored affective reactions to selfregulated music (i.e., participants were able to monitor music with their movements) versus random-regulated music (i.e., music was played automatically) as implicit manifestations of agency (Lancioni et al., 2013a;Lancioni et al., 2013b;Lancioni et al., 2015).Results A. Mentzou et al. showed that participants tended to have increased positive affective states and behaviours (e.g., smiling, singing etc.) in the self-initiated condition where they were able to enjoy their causal influence on music, demonstrating preserved sense of agency.Another manifestation of cognitive agency that has been assessed via tracking participants' ability to differentiate self-generated touch (remotely controlling the strokes of a paintbrush on the hand) from non-self-generated touch (when the brush strokes were remotely controlled by the experimenter) (Bond et al., 2016;Downey, Mahoney, Rossor, Crutch, & Warren, 2012).Downey et al. (2012) found that although all participants with dementia had flawless attribution of self-generated actions, there were potential difficulties with self-other differentiation as average performance of the non-self conditions was reduced compared to healthy participants.This implies that participants with dementia were more prone to erroneously perceive another's touch as their own.However, this set up lacks ecological validity, and Bond et al. (2016) did not find any differences in performance of people with dementia with control participants.
In contrast to quantitative research, qualitative studies have explored sense of agency almost in an all-encompassing way demonstrated by a range of behaviours and subjective experiences.For instance, through interviews, agency has been conceptualised from a decision-making perspective (Boyle, 2014) as well as self-mastery and past achievements (Hedman, Hansebo, Ternestedt, Hellström, & Norberg, 2016).Boyle (2014) explored agency as a form of intact decisionmaking demonstrated when people act in a deliberate way in order to have an impact on their personal circumstances, either explicitly (e.g., expressing opinions) or implicitly (e.g., emotional reactions and expressing preferences non-verbally).Hedman et al. (2016) explored manifestations of agency in the past and present, which was reflected in intact self-mastery (e.g., mastering the consequences of dementia and deliberately preserving one's autonomy) and achievements (e.g., pride in past accomplishments and taking responsibility for one's health), as well as in the future, which was reflected in perceptions of potential loss of agency by becoming more dependent.

Discussion
Recognising the multidimensional nature of the self, the current scoping review aimed to explore the nature and scope of the evidence demonstrating change in the psychological self in people living with dementia.The self was not perceived as a unitary construct, but rather as consisting of closely connected, yet distinctive, manifestations (Klein & Gangi, 2010).Expanding the findings of Caddell and Clare (2010), which offered a broad review of early literature, we focused on giving a theoretically comprehensive account of the different types of the cognitive self explored in recent literature, some of which have not been previously considered (notably with regards to the functional aspects of the self), allowing for a better understanding of self-related alterations in dementia.Thus, the findings were presented into three main types of self-functions: high-order manifestations, functional aspects of the self, and foundational manifestations.
Overall, the findings demonstrated that not all aspects of the self are equally impacted by dementia, providing support for recent conceptual frameworks of the self as a multifaceted construct (e.g., Bomilcar et al., 2021;Huntley et al., 2021).For different aspects of the self, there seems to be some evidence that high-order complex manifestations are more likely to be altered (e.g., autobiographical memory, self-knowledge) while foundational manifestations are more likely to remain intact (i.e., agency, embodiment, self-continuity).It is important to note that these foundational manifestations should not necessarily be considered as less sophisticated; they have inner depth and potentially interact with each other as well as with the functional aspects of the self, becoming integrated to produce emergent high-order cognitive properties, possibly explaining the relative perseverance of self in dementia (Bomilcar et al., 2021;Hutmacher, 2021).These emergent properties might be enough to counteract the progressive self-alterations, sustaining subjective feelings of a stable and unified sense of self despite the notable changes (Mograbi et al., 2021).Self-manifestations linked with embodiment and agency expand beyond primitive aspects of the self, incorporating different non-verbal, yet sophisticated processes at a sensory and affective level to express the temporally extended history of one's life in nuanced and concrete ways (Hutmacher, 2021;Summa & Fuchs, 2015).This idea is reflected in the empirical evidence showing that, even though people with dementia acknowledge diminished cognitive functioning, they still feel they are essentially the same person as they have always been, having intact experiences of selfhood.Consequently, it would be erroneous to believe that having deteriorated memory undoubtedly results in total fragmentation of every aspect of the self.Additionally, as results showed, even fragmented episodic information might be sufficient to provide access to one's life story and subsequently maintain strong feelings of a coherent sense of self across time (Tippett et al., 2018).Only when this connection to one's past is lost, fundamental changes to the core of self may become more prevalent (Tippett et al., 2018).
Although people with dementia encounter progressive cognitive difficulties, empirical evidence does not demonstrate that the sense of self becomes totally lost, at least for participants with dementia who still had functioning capacity to complete the tasks.It might be possible that self-alterations are more absolute in later stages of the disease, nonetheless, the hierarchical nature of the self renders less likely the prospect of total loss.Accordingly, an absolute change was not found in this review for any aspect of the self, with most findings showing an inconclusive pattern of self-alterations (i.e., with regards to autobiographical memory, self-knowledge, SRE, enactment effect and perceived change).Given the heterogeneity of results, it is challenging to make any definite conclusions about the exact pattern of self-alterations since changes in the sense of self in dementia are not a binary phenomenon, and focus should be placed on the strength of changes rather than the mere existence of changes.For instance, although a large number of studies has provided evidence that autobiographical memory performance is particularly diminished in dementia, sensory cueing methods show that episodic memory is not totally diminished but, rather, retrieval difficulties might be more prominent (Kirk & Berntsen, 2018b).Furthermore, preserved self-continuity shows that not all aspects of temporally extended self-representations are lost as well as that diachronic unity is not only supported by autobiographical memory, but, evidently, also from additional manifestations of the self which enable the creation of a coherent life narrative.For example, it has been argued that embodiment may be a key aspect of the self, potentially providing insights to the perceived continuity of one's self, despite memory deficits (Bomilcar et al., 2021).
Focusing on the high-order manifestations, the exact pattern of selfchanges has not yet been established, potentially attributed to the fact that research exploring cognitive self-processes has predominantly focused on autobiographical memory and self-knowledge, reflecting a potential bias towards mnemonic components of selfhood (please also see Fig. 2 that visualises this bias).This domination might further contribute to erroneous assumptions of global change in the sense of self since making definite conclusions from aspects of self showing the most obvious signs of change in dementia is problematic and could potentially lead into a vicious cycle of reasoning.These aspects of self might have been traditionally considered as the core aspects of the self because they are discursively used in explicit ways when presenting oneself (Hutmacher, 2021;McLean, Pasupathi, & Pals, 2007).However, they only constitute two among several aspects of self that may contribute towards maintaining a coherent sense of self, which are often manifested reflexively (Hutmacher, 2021).While the close relationship between self and memory is indisputable, the suggested centrality of memory as the ultimate manifestation of selfhood has been challenged by the presented findings of this review.As research on self-continuity has shown, people with dementia often feel the same individuals across time, providing strong support for the role of less studied manifestations of the self, especially foundational aspects of self, to maintain subjective feelings of preserved selfhood.Future research of the sense of self in dementia should adopt a more encompassing approach, trying to explore the interaction between different manifestations of self in a more holistic way.
This variability of results could also be attributed to the predominant use of verbal tasks when assessing self-processes in people with dementia.These, often quantitative, methods require sophisticated selfreflection abilities which are known to be compromised in dementia (Strikwerda-Brown et al., 2019).Additionally, it is difficult to make define conclusions about what performance on these tasks mean; alternative accounts of responses could easily exist, maybe reflecting intellectual disengagement with the complexity of the questions rather than a loss of self-processing abilities (Tippett et al., 2018).Nevertheless, it is important to mention that these tasks do not take into account preexisting individual differences.Better insights into the pattern of selfalterations could be achieved with the use of sensitive methods that do not place additional cognitive demands to participants with dementia.Employing more naturalistic and ecologically valid tasks, aiming to capture self-processing abilities as used in everyday life, which might be implicitly manifested, could results in a more accurate picture of selfalterations in dementia (Mentzou & Ross, 2022).Only few identified studies used ecologically valid and sensitive methods (e.g., Miles et al., 2013;Rasmussen et al., 2021), which not only facilitated recall but also enhanced motivation, since participants with dementia reported even more memories compared to control participants.Immersing participants into multisensory experiences as well as historically authentic and personally meaningful settings have revealed a different picture of selfalterations, showing promising results of preserved episodic recall in dementia.
Qualitative research exploring the lived experiences, feelings of perceived change and embodied manifestations in dementia has many advantages.It allows people with dementia to express their own feelings and interpretations of progressive self-alterations after a dementia diagnosis, providing insights into how changes in the psychological self are perceived.This reflects the idea that, despite the undisputable alterations of cognitive self-processes, these might not always be perceived as having a significant impact in the sense of self of people with dementia.Research on embodiment, which was mostly based on observations and everyday encounters, provides new methodological opportunities to explore foundational manifestations of self using sensitive research designs.Findings also showed that although the majority of studies reported preserved subjective experiences of selfhood, there have been instances of reported changes.Given the inherent subjectivity of the self-experience, these studies demonstrate that potential changes are perceived differently, having a greater impact for some people than others (Caradec & Chamahian, 2017).Notably, perceived changes might not be stable; as dementia progresses, challenges to the sense of self might become more pronounced (MacRae, 2010).An important limitation of the qualitative approaches reviewed here is that they often perceive the self as a singular entity, as it was also identified in Caddell and Clare (2011).Terms such as identity, sense of self and personhood are often used interchangeably, without the provision of a specific definition, even though such terms can be defined differently with subsequent different methodological implications.
Research on the functional aspects of the self further demonstrates that self-processes can be successfully studied implicitly using intelligently designed cognitive tasks, that might not place strong linguistic demands on participants.These studies have provided further support for the complexity of the different presentations of the sense of self.There is no binary distinction between high-order and foundational manifestations, rather different self-functions exist along a continuum with functional aspects allowing for cross-connections, demonstrating that self-related changes in dementia are more complex than originally thought, potentially without a dualistic distinction of alterations.Although these, mostly implicit, manifestations of self can provide a promising new direction for the study of self-related changes in dementia, the existing evidence is scarce and characterised by methodological limitations.For instance, agency seems to be preserved in dementia; both quantitative and qualitative studies have shown intact agency using different approaches.However, agency is conceptualised differently in the different studies reviewed here, which is particularly true for qualitative studies that seem to explore agency in an allencompassing way without a clear conceptual explanation.This is also the case with research identified in this review exploring pronoun use; perceiving instances of successful use of first person indexical as solely a manifestation of intact sense of self is flawed since pronoun use might also be considered as a socially acquired ability that does not necessarily constitute an aspect of the self per se, but rather a linguistic by-product of higher-order cognitive self-processes.
The current scoping review has certain limitations.Notably, given the multidimensional nature of the self as well as our inclusive criteria, only preliminary conclusions can be made as not all manifestations of selfhood have been adequately explored in the dementia literature.Furthermore, considering that cognitive processing in individuals in early-stage dementia differs fundamentally from individuals in latestage dementia, conclusions about developmental trajectories could not be made due to the lack of evidence from people with advanced stages of dementia (please see Table 4).This is an important dimension that future research ought to explore in order to acquire a more nuanced understanding of the exact pattern of self-alterations in dementia.The tentative nature of this review could be further attributed to the limited focus of the scoping review method in producing a critically appraised and synthesised answer to a given question within the literature.The lack of a quality assessment of the included studies can be viewed as an additional limitation.Nonetheless, this is in accordance with current guidelines, since the scoping review method aims to provide a broad overview of the nature and scope of evidence in the literature, irrespective of study quality, and therefore grey literature is also searched (Tricco et al., 2018).
Overall, the results of the current scoping review showed that although there are some alterations in different manifestations of the self, these do not imply unified change.Given that some elements of autobiographical memory and self-knowledge seem to still be preserved in dementia, together with the findings of intact embodiment and agency, the self-system still functions in dementia, albeit a progressive 'break down' might be occurring.Despite the notable cognitive changes during dementia, it seems that those aspects of self that are still preserved could be enough to compensate for potential weakening of some self-processes.In accordance with Bomilcar et al. (2021), the current review reveals that foundational manifestations of the self are complex and more sophisticated than it might have been initially assumed as they interact with each other to potentially sustain a subjective experience of stable and coherent sense of self across time, despite notable alterations in high-order manifestations.Better understanding of alterations in selfhood, ensuring that people's experiences are not undermined by perceptions of total loss, is key to address psychological symptoms of people living with dementia as well as to inform new pathways for dementia care interventions.Person-centered psychological care can only be achieved by recognising the centrality of selfhood in people's wellbeing and quality of life (Huntley et al., 2021).

Role of funding sources
Aikaterini Mentzou is currently undertaking a PhD at the University of Dundee, awarded by the Scottish Graduate School of Social Sciences (SGSSS) and funded by the Economic and Social Research Council (ESRC).SGSSS and ESRC had no role in the study design, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

3-5
Objectives 4 Provide an explicit statement of the questions and objectives being addressed with reference to their key elements (e.g., population or participants, concepts, and context) or other relevant key elements used to conceptualize the review questions and/or objectives.
2 & 5 METHODS Protocol and registration 5 Indicate whether a review protocol exists; state if and where it can be accessed (e.g., a Web address); and if available, provide registration information, including the registration number.

5
Eligibility criteria 6 Specify characteristics of the sources of evidence used as eligibility criteria (e.g., years considered, language, and publication status), and provide a rationale.

6-7
Information sources* 7 Describe all information sources in the search (e.g., databases with dates of coverage and contact with authors to identify additional sources), as well as the date the most recent search was executed.

7-8
Search 8 Present the full electronic search strategy for at least 1 database, including any limits used, such that it could be repeated.

8
Selection of sources of evidence † 9 State the process for selecting sources of evidence (i.e., screening and eligibility) included in the scoping review.

8-9
Data charting process ‡ 10 Describe the methods of charting data from the included sources of evidence (e.g., calibrated forms or forms that have been tested by the team before their use, and whether data charting was done independently or in duplicate) and any processes for obtaining and confirming data from investigators.

10-11
Data items 11 List and define all variables for which data were sought and any assumptions and simplifications made.9-11 Critical appraisal of individual sources of evidence § 12 If done, provide a rationale for conducting a critical appraisal of included sources of evidence; describe the methods used and how this information was used in any data synthesis (if appropriate).
N/A Synthesis of results 13 Describe the methods of handling and summarizing the data that were charted.10-14 RESULTS Selection of sources of evidence 14 Give numbers of sources of evidence screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally using a flow diagram.

10
Characteristics of sources of evidence 15 For each source of evidence, present characteristics for which data were charted and provide the citations.

13-24
Critical appraisal within sources of evidence 16 If done, present data on critical appraisal of included sources of evidence (see item 12).N/A

Results of individual sources of evidence 17
For each included source of evidence, present the relevant data that were charted that relate to the review questions and objectives.

13-24
Synthesis of results 18 Summarize and/or present the charting results as they relate to the review questions and objectives.11-24 DISCUSSION Summary of evidence 19 Summarize the main results (including an overview of concepts, themes, and types of evidence available), link to the review questions and objectives, and consider the relevance to key groups.

24-29
Limitations 20 Discuss the limitations of the scoping review process.6-8 & 29 Conclusions 21 Provide a general interpretation of the results with respect to the review questions and objectives, as well as potential implications and/or next steps.

28-29
FUNDING Funding 22 Describe sources of funding for the included sources of evidence, as well as sources of funding for the scoping review.Describe the role of the funders of the scoping review.

Fig. 1 .
Fig. 1.PRISMA flow diagram illustrating the process of our article identification and screening.

Fig. 2 .
Fig. 2. Bubble plot illustrating the time trend and different self-aspects explored across the reviewed articles.The size of each 'bubble' is representative of the number of papers published in each year.
for the review in the context of what is already known.Explain why the review questions/objectives lend themselves to a scoping review approach.

Table 2
Inclusion and exclusion criteria of the current scoping review.
g., studies focused on the impact of interventions on the self in dementia)

Table 3
Examples of search strings expressed using Boolean operators that were implemented to different databases.
PUBMED (("self"[Title/Abstract] OR "identity" [Title/Abstract] OR "personhood" [Title/Abstract] OR "selfhood" [Title/Abstract] OR "agency" [Title/ Abstract] OR "autobiographical" [Title/Abstract] OR "embodiment" [Title/Abstract] OR "continuity" [Title/Abstract] OR "narrative"[Title/ Abstract]) AND ("dementia" [Title/Abstract] OR "alzheimer's"[Title/ Abstract])) AND ("change" [Title/Abstract] OR "difference" [Title/ Abstract] OR "impact" [Title/Abstract] OR "transition"[Title/Abstract]) PsycINFO TI ((self OR identity OR personhood OR selfhood OR agency OR autobiographical OR embodiment OR continuity OR narrative)) AND AB ((self OR identity OR personhood OR selfhood OR agency OR autobiographical OR embodiment OR continuity OR narrative)) AND TI ((dementia OR alzheimer's)) AND AB ((dementia OR alzheimer's)) AND TI ((change OR difference OR impact OR transition)) AND AB ((change OR difference OR impact OR transition OR greater OR less OR more OR better OR worse OR higher OR lower OR fewer OR compromise OR enhance OR impair OR improve))

Table 4
Dementia Stage, Methodology and Levels of Self for the included studies.
Barnabe, Whitehead, Pilon, Arsenault-Lapierre, & Chertkow, 2012).Accordingly, it has been widely explored in people with dementia as it has been argued that disruptions in autobiographical memory could result in a compromised sense of self.