Abstract
Objective
A positive sense of self may be a key domain of psychological well-being for people living with dementia and therefore a legitimate target for psychosocial interventions in dementia care. Determining the effectiveness of such interventions often requires valid self-report instruments. This review aimed to investigate what aspects of self have been measured using self-report instruments in evaluating psychosocial interventions for people living with dementia and to explore the effectiveness of these interventions in terms of positive outcomes related to aspects of self.
Method
A systematic search of the literature using five electronic databases and one register (CENTRAL) was conducted. A narrative synthesis and methodological quality assessment was completed for the included studies.
Results
A total of 24 studies were included in the review. Seven aspects of self were measured using a range of self-report instruments, many of which have not been validated for dementia. Aspects of self were; self-esteem, self-efficacy, self-compassion, self-growth, self-acceptance, self-management, and self-identity. Studies included a variety of interventions; however, the effectiveness of these interventions for these aspects of self was mixed.
Conclusion
There is some evidence that psychosocial interventions improve specific aspects of self in dementia but further research to improve this evidence base is needed. Future research should also investigate and confirm the validity and reliability of existing self-report instruments that aim to measure aspects of self in dementia. Limitations and implications of the review are discussed.
Introduction
Feeling uncertain about maintaining one’s sense of self and identity is a key concern for people living with dementia (Caddell & Clare, 2011a; Steeman et al., 2007). Exposure to negative narratives surrounding the condition in terms of progressive ‘loss of self’ (Davis, 2004), competence and self-esteem (Nguyen & Li, 2020) can feed into this and contribute to experiences of stigma (Nguyen & Li, 2020).
Construction of a negative ‘loss of self’ narrative can be understood within Sabat and Harre’s (1992) theory of self in dementia which proposes that a person holds three ‘selves’; (1) one’s point of view and personal identity, (2) the attributes one holds along with beliefs about these, and (3) how one presents themselves to the world (Sabat, 2001). According to this theory, people living with dementia are positioned by others as helpless and confused and their behaviour is then interpreted by others to confirm this (Sabat & Harre, 1992). This can lead to others perceiving a diminished self in people living with dementia even when self remains intact for the person with dementia. Therefore, the preservation of self in dementia is heavily influenced by a complex interplay between interpersonal, social, psychological, and embodied factors (Surr, 2006).
Alternate multi-dimensional theories of self have also been applied to dementia (see Caddell & Clare, 2011b) such as Neisser’s (1988) Five Factor Model of Self, comprising ecological, interpersonal, extended, private, and conceptual selves. More recently, Bomilcar et al. (2021) proposed seven interacting domains of self in dementia; embodied, agentic, implicit, critical, surrogate, extended, and emergent self. Nonetheless, there remains no agreed definition of self in dementia, and no agreement as to whether it is a unitary construct or consists of several different ‘selves’, self domains, or self-evaluations (Caddell & Clare, 2010, 2013a; Klein & Gangi, 2010; McConnell, 2011).
The lack of an agreed definition or model is evident in the range of different qualitative and quantitative methods available to measure and understand aspects of self (Caddell & Clare, 2010). Quantitative measures may be particularly useful in exploring how aspects of self change over time (Caddell & Clare, 2010) and in response to psychosocial interventions in dementia care (Moniz-Cook et al., 2008; Schölzel-Dorenbos et al., 2007). However, a review by Caddell and Clare (2011b) found that for interventions aiming to support self and identity in dementia, very few studies used standardised self-report measures relating to aspects of self and instead relied on observational methods or well-being measures.
The use of observational, or proxy-based, measures risks bias (Schölzel-Dorenbos et al., 2007) and is influenced by the proxies’ own experiences (Logsdon et al., 2002; Sands et al., 2004). e.g., people diagnosed with dementia self-report higher quality of life than carers completing proxy measures (e.g., Griffiths et al., 2020; Hounsome et al., 2011; Logsdon et al., 2002; Moyle et al., 2012; Sands et al., 2004; Sheehan et al., 2012). Subsequently, there has been a growing interest in the use of self-report outcome measures, reflecting the move to person-centred care and recognition of the varied and unique experiences of dementia (Kitwood, 1997, 2019; Kitwood & Bredin, 1992).
A previous scoping review of well-being self-report measures identified six self-related measures and proposed that a ‘positive sense of self’ was a key domain of psychological well-being for people living with dementia (Clarke et al., 2020). In addition, self-report measures of self-efficacy and self-identity have been identified as self-related positive psychology outcome measures in dementia (Stoner et al., 2019). A positive sense of self can be dynamic and maintained whilst living with dementia (Caddell & Clare, 2010; Strikwerda-Brown et al., 2019) and may affect coping with the challenges that follow a dementia diagnosis (Caddell & Clare, 2011b). Therefore, aspects of self that may be quantifiable, such as self-esteem and self-efficacy, can be measured when evaluating psychosocial interventions aiming to improve well-being in dementia (Lamont et al., 2019). Whilst Clarke et al. (2020) and Stoner et al. (2019) did not specifically focus their reviews on measures of disparate aspects of self, their findings suggest that since Caddell and Clare’s (2011b) review a larger pool of self-report measures relating to aspects of self are being used with people living with dementia.
The aim of the current review was therefore to extend previous reviews (Caddell & Clare, 2011b; Clarke et al., 2020; Stoner et al., 2019) to develop a clearer understanding of what aspects of self have, to date, been measured within evaluations of psychosocial interventions for people living with dementia and how effective these interventions have been in relation to these aspects of self.
The specific questions underpinning this review were: What aspects of self have been measured using self-report instruments to evaluate psychosocial interventions for people living with dementia? What is the effectiveness of psychosocial interventions in demonstrating positive outcomes related to aspects of self?
This review aimed to identify what aspects of self have been measured in psychosocial interventions for dementia, and so adopted the position that there are multiple specific domains of self (e.g., Bomilcar et al., 2021) rather than a single unitary self.
Methods
Search strategy
A systematic review of the literature was conducted in January 2023. The following electronic databases were searched via the platform EBSCOHost: Academic Search Premier, PsycINFO, PsycARTICLES, MEDLINE and CINAHL Complete to cover psychology, health, and medicine. The Cochrane Central Register of Controlled Trials (CENTRAL) was also searched to increase the likelihood of identifying all relevant literature relating to randomised control trials (RCTs). Preliminary searches prior to the database review helped to identify key search terms such as aspects of self that had been highlighted or discussed in existing dementia research.
The following search terms were used: (dement a or alzheimera) AND (TI (interventiona or treata or programa or counsela or therapa or activita or groupa or supporta or workshop or course)) AND (“sense of self” or “aspect of self” or selfhood or self-esteem or self-efficacy or self-compassion or self-identity or self-stigma or self-concept or self-worth or self-awarea or self-agency or self-acceptance or self-confidence or self-trust or self-image or self-respect or self-recognition or self-knowledge or self-determination or self-critica or self-control).
Truncations (a) and the operators (OR/AND) were used to broaden the search. To increase the frequency of relevant articles the following search limiters were applied on the EBSCOHost search: academic journals, peer reviewed and English language. A date limiter of 1992–2023 was also used as this was when Sabat and Harre (1992) published their theory of self in dementia.
Article screening
Inclusion and exclusion criteria for the review.

PRISMA (Page et al., 2021) flow diagram of article selection process.
Following the data extraction process for the 24 included studies, the terms ‘self-management’ and ‘self-growth’ were also identified. An additional search using these terms was conducted to identify relevant papers that may have been missed (see Supplemental file 1). Seven full papers were screened; however, all were excluded.
Quality assessment
A quality assessment of included studies was conducted using the Mixed Methods Appraisal Tool (MMAT, Hong et al., 2018), which is outlined in Supplemental file 2. The assessment criteria for ‘quantitative descriptive studies’ (4.1–4.5) were removed as these were outside the scope of this review. Qualitative data was excluded from this review; however, for mixed method design studies the qualitative quality assessment items were completed to fulfil the overall mixed method assessment criteria. Quality assessment items were scored using the possible responses; yes, no or can’t tell. The overall percentage score of ‘yes’ criteria is presented alongside descriptive summaries. Therefore, studies may score 20, 40, 60, 80 or 100% (Hong et al., 2020). Percentage scores for mixed methods studies were determined by the percentage of ‘yes’ ratings in the lowest scoring component (Hong et al., 2020). Studies were rated ‘yes’ for the mixed methods criteria 5.5 (adherence to quality criteria for each component) if four or five out of the five assessment criteria for both the quantitative and qualitative components were rated ‘yes’. Regarding complete outcome data and attrition (criteria 2.3), this review used a cut off of 20% for acceptable withdrawal rates (Van Tulder et al., 2003) and 80% for acceptable complete outcome data (Thomas et al., 2004). Studies were not excluded from this review based on quality scores or screening questions, but methodological quality assessment was integrated into the synthesis of findings.
The MMAT advises a minimum of two independent reviewers (Hong et al., 2018), therefore inter-rater reliability was assessed by an independent reviewer who quality assessed five (20.8%) of the included studies; one from each percentage score band. Out of the 45 ratings possible from the five studies, four discrepancies were identified. The discrepancies were discussed before the ratings were mutually agreed.
Data extraction and analysis
Due to the heterogeneity of the included studies (in terms of interventions and outcomes) a meta-analysis was not suitable and therefore a narrative synthesis was used to answer the research questions. Following the narrative synthesis guidance of Popay et al. (2006), a data extraction form was created (Supplemental file 3). A preliminary synthesis of the data was conducted to identify patterns across the studies outcomes and designs, followed by an exploration of relationships between the studies and an assessment of the studies strengths and limitations. For mixed method studies, only the quantitative self-report components completed by people with dementia were synthesised. The term ‘participants’ used in this review therefore refers only to participants in the studies who had a diagnosis of dementia.
Results
Overview of included studies
Summary of included studies characteristics.
aStudies where interventions appeared to explicitly target an aspect of self within the study aims or research questions.
Included studies took place in the UK (n = 8), USA (n = 3), France (n = 3), Spain (n = 2), Netherlands (n = 2), Hong Kong (n = 1), Germany (n = 1), South Korea (n = 1), Australia (n = 1), Canada (n = 1), and one multi-national study took place in the UK, Italy, and Poland.
The majority of studies were RCTs (n = 11), followed by mixed-methods (n = 9), non-randomised repeated measures (n = 8), case series (n = 2), quasi-experimental (n = 2) and one n = 1 design. Evaluated psycho-social interventions varied considerably in their focus and content, encompassing psycho-educational, cognitive, creative/arts and social approaches. Whilst all studies formally assessed an aspect of self as an outcome, only nine explicitly targeted an aspect of self in the evaluated intervention (see Table 2).
The total number of included participants across all studies was 1,893 and the majority of interventions were delivered in a group format. Subtypes of dementia included were Alzheimer’s dementia, Vascular dementia, Mixed dementia, Parkinson’s dementia, Lewy body dementia, and Frontotemporal dementia. Most studies (n = 13) included a range of these subtypes but two included only individuals with Alzheimer’s dementia (Gonzalez et al., 2015; Lee et al., 2008) and four did not report subtypes (Collins et al., 2022; Fitzsimmons & Buettner, 2003; Sprange et al., 2015; Werheid et al., 2021). Six studies included individuals with ‘probable dementia’.
Most studies (n = 11) included mainly participants in the early/mild stages of dementia. Eight studies included participants in the mild to moderate stages (Brooker et al., 2018; Clare et al., 2019; Collins et al., 2022; Cooke et al., 2010; Craig et al., 2018; Dodd et al., 2022; Richards et al., 2019; Werheid et al., 2021), one study included moderate to severe stages (Platel et al., 2021) and three included a range of stages of dementia (Dröes et al., 2019; Hindle et al., 2018; Pérez-Sáez et al., 2020). Foloppe et al., (2018) single participant was described as being in the moderate stage of dementia.
Two studies explicitly stated that the measure of self was a primary outcome of the intervention (Quinn et al., 2016; Richards et al., 2019) and six studies reported on an aspect of self as a secondary outcome (Berk et al., 2019; Clare et al., 2019; Hindle et al., 2018; Marshall et al., 2015; Mountain et al., 2022; Pongan et al., 2017). The remaining studies did not differentiate between primary or secondary outcomes in relation to the chosen measures.
Quality of included studies
Supplemental file 4 shows study quality ratings. Two studies did not have clear research questions or aims rendering the research purpose unclear (Richards et al., 2019; Werheid et al., 2021). The majority of studies were RCTs, the ‘gold standard’ of research designs (Rennie, 1996; Sibbald & Roland, 1998) although the quality of the studies varied, with only Clare et al. (2019) and Hindle et al. (2018) meeting all quality criteria for RCTs.
Common limitations for RCTs and non-randomised studies were high attrition rates and incomplete data, which can lead to a loss of statistical power and invalid conclusions (Kang, 2013). Lack of consideration of confounding factors (nine studies) may have limited internal validity. External validity may have been limited due to participants not matching target populations (five studies) and generalisability may have been limited in three RCTs due to incomparable control and intervention groups at baseline. A relative strength regarding internal validity was that nine RCTs blinded outcome assessors to intervention groups which reduces the risk of detection bias.
Mixed method studies were relatively of lower quality, often owing to lack of clarity around the rationale for their chosen methodological approach. The quantitative components of these studies were mostly of lower quality compared to qualitative components, although Mountain et al. (2022) and Quinn et al. (2016) met all criteria for both components. Werheid et al. (2021) and Fitzsimmons and Buettner (2003) were both lower quality studies due to not meeting any qualitative quality criteria; however, their quantitative components, which were synthesised in this review, were of higher quality.
Aspects of self and their measurement
Aspects of self measured within the included studies.
Note. Descriptive summary of the rating scale for each measure outlined in supplemental file 5.
Only the DQoL (Brod et al., 1999) has been validated specifically for people living with dementia; however, acceptable internal consistency reliability was found for the RSES (Rosenberg, 1965) in dementia (Burgener et al., 2008) and initial evidence supports the validity and reliability of the IMAGE test (Eustache et al., 2013).
Of the studies included in the review, only three reported on internal consistency reliability of the measures in their respective studies; the RSES (Rosenberg, 1965) demonstrated α = 0.89 to α = 0.92 across three assessment points (Burgener et al., 2008) and the Korean RSES (Jeon, 1974 as cited in Lee et al., 2008) demonstrated α = 0.62 (Lee et al., 2008). Cooke et al. (2010) reported Cronbach’s alpha for the DQoL (Brod et al., 1999) subscales was between 0.62-0.87 but did not report on the exact reliability of the self-esteem subscale separately.
Intervention effectiveness
Outcomes of psychosocial interventions in relation to aspects of self are described below, according to intervention type and grouped by construct.
Self-esteem
Multicomponent Psychoeducational and Social Interventions
Two studies reported an increase in self-esteem following a psychoeducational group (Fitzsimmons & Buettner, 2003) and a psychoeducational psychotherapy group (Marshall et al., 2015) in contrast to one RCT which reported no significant changes in self-esteem following a support group (Young et al., 2014). The participants in all three studies were mostly people diagnosed with dementia in the year (Marshall et al., 2015) or three years (Young et al., 2014) prior to the intervention. Fitzsimmons and Buettner (2003) stated only that participants were ‘newly diagnosed’ and did not conduct statistical analysis, therefore limiting the extent the studies can be compared. However, one key difference was that these studies used different language versions of the RSES within different countries. Therefore, the measures’ translation or possible cultural differences in self-esteem and/or effectiveness of the interventions may have impacted the findings.
Meeting Centre Support Programme’s (MCSP) which provide personalised post diagnostic support were evaluated in two studies (Brooker et al., 2018; Dröes et al., 2019). Significant increases in self-esteem after six months were reported by Brooker et al. (2018). Updating MCSP to include supporting people with dementia to work as volunteers in the community did not show any significant differences in self-esteem when compared to the regular MCSP or those receiving no day services (Dröes et al., 2019). However, findings may have been impacted by demographic differences between the groups such as participants in the volunteering group being younger, more often male and living independently. Furthermore, both studies (Brooker et al., 2018; Dröes et al., 2019) had high attrition rates, possibly impacting non-response bias, and due to the flexible nature of the programmes, participant attendance levels varied.
Reminiscence Interventions
Reminiscence groups using quasi-experimental designs did not demonstrate clear evidence for improving self-esteem; Gonzalez et al. (2015) found no significant improvement in self-esteem and Lee et al. (2008) found the initial significant improvements in self-esteem were not maintained long term at the six month follow up. Participant representativeness is unclear in Lee et al. (2008), who did not fully report on participant demographics, limiting external validity. On the other hand, reminiscence interventions may be more effective when delivered individually, according to the findings of Dodd et al. (2022). However, the evidence in this area is weak due to small sample size, therefore limiting generalisability.
Mindfulness and Third Wave Therapeutic Interventions
Group Tai Chi alongside Cognitive Behavioural Therapy demonstrated a slight improvement in self-esteem 20 weeks post intervention, which stabilised at 40 week follow up (Burgener et al., 2008). However, intervention and control groups showed significant differences in self-esteem at 20 weeks due to a decrease in self-esteem for the control group; in addition, the randomisation procedure was unclear in this study. Considering the progressive nature of dementia, stabilisation (opposed to decline) was perceived as a positive outcome (Burgener et al., 2008), and could suggest that the intervention helped to protect against decline in self-esteem, albeit non-response bias may be an issue due to attrition (>20%).
Creative Interventions
Three RCT’s (Cooke et al., 2010; Pongan et al., 2017; Richards et al., 2019) and one non-randomised design (Pérez-Sáez et al., 2020) reported increases in self-esteem following creative interventions. Pottery workshops (Pérez-Sáez et al., 2020) and visual arts activities such as painting, ceramics, and photography (Richards et al., 2019) may be beneficial for self-esteem regardless of cognitive impairment or stage of dementia. However, the increase in self-esteem was not significant (Richards et al., 2019) and the findings of Pérez-Sáez et al. (2020) should be interpreted with caution due to the exclusion of eight participant data sets as a result of incomplete responses or participants receiving ‘excessive help’ to complete the questionnaire.
Participant attendance may impact the benefits of creative interventions as Cooke et al. (2010) found that only participants who attended over 50% of the music or reading groups demonstrated significant increases in self-esteem. However, approximately half of the participants attended over 50% of sessions, indicating adherence to the intervention was an issue. On the other hand, Pongan et al. (2017) did not find a significant increase in self-esteem for participants with at least 50% attendance but did find that the painting group showed a greater increase in mean self-esteem scores across time. However, the lack of a non-intervention control group as a comparison and insufficient explanation of randomisation procedure is a limitation.
Technologies
Foloppe et al. (2018) investigated the use of virtual reality-based training on autonomy in cooking activities and found no benefits to self-esteem for a 79-year-old woman with probable Alzheimer’s dementia. However, generalisability of these findings is limited by the n = 1 design.
Self-efficacy
Multicomponent psychoeducational and social interventions
Five studies conducting interventions which included social and educational components measured self-efficacy as an outcome (Fitzsimmons & Buettner., 2003; Mountain et al., 2022; Quinn et al., 2016; Sprange et al., 2015; Young et al., 2014), two of which investigated the ‘Journeying Through Dementia’ self-management programme (Mountain et al., 2022; Sprange et al., 2015). Only one study found improvements in self-efficacy, however, with small effect sizes at three- and six-months post intervention compared to treatment as usual (Quinn et al., 2016). Although this study scored highly in quality assessment, statistical analysis of the data was not possible due to lack of power (Quinn et al., 2016), limiting conclusions. Mostly self-efficacy remained stable or showed no significant improvement following the interventions (Fitzsimmons & Buettner., 2003; Mountain et al., 2022; Sprange et al., 2015; Young et al., 2014).
Cognitive interventions
Cognitive interventions showed some evidence for improving self-efficacy, specifically Cognitive Stimulation Therapy, (CST; Collins et al., 2022; Werheid et al., 2021) which may be influenced by participants recognising a stabilisation in their cognitive abilities (Werheid et al., 2021). However, studies had small sample sizes and Collins et al. (2022) did not conduct statistical analysis, rendering evidence for this assertion relatively weak.
Two RCTs evaluated Cognitive Rehabilitation (CR) interventions and found no significant changes in self-efficacy (Clare et al., 2019; Hindle et al., 2018). Whilst Hindle et al. (2018) did find a significant difference in self-efficacy between the CR and relaxation control group at two months post-intervention, there was no significant difference between the groups at the six month follow up or with the TAU group at any time point. Both studies were rated as high quality due to meeting all quality assessment criteria.
Self-compassion
Mindfulness and third wave therapeutic interventions
Two studies showed contrasting findings for self-compassion following a mindfulness-based intervention (Berk et al., 2019) and a Compassion Focussed Therapy (CFT) intervention (Craig et al., 2018). Both studies were limited due to no statistical analysis. Although Berk et al. (2019) found a reduction in self-compassion post intervention, 71% of participants received help completing the self-compassion measure and therefore the findings should be interpreted with caution. Furthermore, generalisability was low for the self-compassion improvements found by Craig et al. (2018) due to the case series design.
Self-acceptance and self-growth
Reminiscence interventions
Two reminiscence interventions showed positive outcomes in relation to self-growth when delivered in group (Gonzalez et al., 2015) or couples format (Dodd et al., 2022). However, Dodd et al.’s (2022) case series design limits generalisability. Gonzalez et al. (2015) also found a significant time and group interaction for self-acceptance but due to lack of a follow up it was unclear whether the significant increase in self-acceptance post intervention was maintained longer term (Gonzalez et al., 2015).
Self-identity
Reminiscence intervention
Only one study (Platel et al., 2021) measured changes in self-identity, in the context of an evaluation of a musical reminiscence programme using songs to promote autobiographical memory retrieval. Participant inclusion and exclusion criteria was unclear; therefore, sample representativeness was uncertain, limiting external validity. Overall, no significant differences were detected using either the IMAGE test or I-AM test (Eustache et al., 2013), suggesting that musical reminiscence did not affect self-identity.
Self-management
Multicomponent intervention
Mountain et al. (2022) was the only study to measure self-management in their intervention, which comprised self-management elements and engagement in meaningful activity, even though an additional two studies were described as ‘self-management’ interventions (Quinn et al., 2016; Sprange et al., 2015). Reflecting findings for self-efficacy, at eight months post intervention, self-management outcomes were in favour of the intervention group compared to the control but differences between the groups were not significant and self-management remained stable pre and post intervention (Mountain et al., 2022).
Discussion
The aims of this review were to (1) explore what aspects of self have been measured in the context of evaluations of psychosocial interventions for people living with dementia and (2) to evaluate the effectiveness of existing interventions in achieving positive outcomes for aspects of self in dementia. The included studies demonstrated considerable heterogeneity in relation to study design, country, and intervention type and study outcomes were also therefore heterogeneous. Nevertheless, this review offers a constructive synthesis of the relationships and patterns between these studies based on aspects of self and various psychosocial interventions.
Aspects and measures of self
An increase in the use of self-report instruments to measure aspects of self in interventions was evident due to the majority of studies included in this review being published following that of Caddell and Clare (2011b). Whilst current conceptualisations highlight the existence of multiple domains of self in dementia (Bomilcar et al., 2021; Neisser; 1988), this review highlights how the majority of included studies have measured self-esteem or self-efficacy. One reason for this may be that at present there are limited validated self-report instruments that allow assessment of different aspects of self for people living with dementia. This is also reflected in the wider literature outside the scope of this review, for example; self-stigma scales have tentatively demonstrated validity in dementia (Bhatt et al., 2021; Burgener & Berger, 2008), and the Tennessee Self-Concept Scale (TSCS-II; Fitts & Warren, 1996) and Self-Identity in Dementia Questionnaire (SID-Q; Cohen-Mansfield et al., 2000) have been used with people with dementia in non-intervention studies (e.g., Addis & Tippett, 2004; Caddell & Clare, 2013b). However, all of these measures have limited psychometric validity regarding use in dementia. This highlights the need for future research to investigate and confirm the validity of a range of self-related measures for people living with dementia.
There may be accessibility issues for some measures, such as the SCS-SF (Raes et al., 2011) and the Spanish RSES (Martín-Albo et al., 2007), considering the additional support that some participants required to complete the scales in different studies. Further concerns regarding validity and possible bias against non-western cultures are indicated by translation issues of the RSES (Rosenberg, 1965) into Chinese (Leung & Wong, 2008) and the Korean RSES (Jeon, 1974 as cited in Lee et al., 2008), which shows limited validity with older adults (Lee, 2022). This may be due to differences in how self-esteem is self-evaluated in individualistic or collectivist cultures (Cai et al., 2007; Schmitt & Allik, 2005). Given these issues, the use of measures of self that lack demonstrated validity with people with dementia is concerning if they are being used as evidence to determine effectiveness of psychosocial interventions across cultures.
Effectiveness of interventions
The effectiveness of interventions on aspects of self varied in relation to the type of intervention and aspect of self which was measured. Multicomponent interventions that incorporate psychoeducational elements, followed by reminiscence interventions, were the most frequent interventions conducted and tested. Of note, included studies varied in the extent to which interventions explicitly targeted an aspect of self. There was no discernible pattern as to whether such interventions led to better or different outcomes for people with dementia, suggesting that positive outcomes can be achieved using different therapeutic approaches. However, only interventions that explicitly targeted an aspect of self within the study aims or research questions were considered as having targeted an aspect of self. It was not within the scope of this review to explore the conceptual underpinnings of interventions, therefore, interventions that may have strong conceptual ties with an aspect of self, such as self-compassion within CFT, but did not target an aspect of self within the study aims/research questions were not defined as explicitly targeting an aspect of self. Future intervention research should consider whether accordance between the conceptual underpinnings of a psycho-social intervention and targeted aspects of self influences outcomes for people with dementia (see Clarke et al., 2020).
Whilst CST showed consistent evidence for improving self-efficacy, which may be influenced by a perceived improvement in memory and cognitive abilities (Hall et al., 2013), other psychosocial interventions did not have significant positive outcomes in relation to self-efficacy or self-management in dementia. The wider chronic disease literature, suggests that such mixed findings in relation to self-efficacy outcomes following interventions such as self-management programmes, may be influenced by barriers such as feeling overwhelmed by the amount of information presented about a condition in a short period of time (Farley, 2020). Considering this, interventions which include elements of psychoeducation may not demonstrate positive outcomes regarding self-efficacy for people with dementia if these elements are experienced as overwhelming and not sufficiently adapted.
Findings overall appear stronger for self-esteem though also mixed. Creative interventions demonstrated the most consistent improvements in self-esteem across differing stages of dementia, reflective of the wider literature showing that art therapy programmes improve self-esteem for older adults (Ching-Teng et al., 2019; Kim, 2013). Improvements seen for self-esteem may relate to the increased sense of control, mastery and accomplishment that follow creative interventions, alongside providing an opportunity for self-exploration (Ching-Teng et al., 2019; Kim, 2013; Richards et al., 2019). A sense of accomplishment relating to the possible perceived improvements in cognitive abilities (Hall et al., 2013) may also explain the improvements in self-efficacy following CST.
Group interventions appeared to demonstrate more positive outcomes for aspects of self compared to interventions delivered to participants individually. This supports previous research that has identified social connectedness as a key facilitator of self-efficacy during interventions for people with chronic diseases (Farley, 2020), as well as suggestions that for people with dementia, social connections are a key factor for improving experiences of self within interventions (Baird & Thompson, 2018). Group interventions provide opportunities for people with dementia to share experiences, information, and understandings of dementia to help others, and these can directly improve self-esteem (Mason et al., 2005). However, it is important to note that the majority of included studies in the current review were group interventions and direct comparisons with one-to-one interventions were limited. Future research may seek to further explore the differences in outcomes for aspects of self between group versus individual interventions.
Differences in intervention outcomes according to demographic characteristics were not necessarily a focus of included studies but may be a possible explanation for the varied findings. For example, differences in the proportion of male and female participants may have impacted the contrasting findings regarding the impact of mindfulness/CFT on self-compassion as previous research suggests that masculine stereotypes are associated with lower self-compassion (Reilly et al., 2014). On the other hand, amongst older adults, self-compassion may be greater in males compared to females (Bratt & Fagerström, 2020). An alternate explanation may be that CFT interventions support all three self-compassion components (Neff, 2003), self-kindness, common humanity, and mindfulness, whereas mindfulness-based interventions only align with the mindfulness component. This raises questions as to what specific components of interventions may positively impact aspects of self in dementia. Future research should examine what other factors (e.g., attendance rates, social connections, dementia subtypes and level of cognitive impairment) may be associated with differential outcomes for aspects of self specifically for people living with dementia.
Limitations
Whilst a range of databases and search terms were used in this review, as well as the additional search for terms that arose post-data extraction, it is possible that other measures of self were missed due to complexities surrounding definitions of self. For example, Clarke et al. (2020) included ‘dignity’ as a measure of ‘positive sense of self’ but was not considered as an aspect of self in this review. Excluding specific task focussed measures of self and only including peer reviewed studies may have further limited the present review. Similarly, only studies in the English language were included; therefore, this review may not capture all relevant evidence. Whilst two independent reviewers were involved in the quality rating process, only one reviewer was involved in abstract/title screening and data extraction; therefore, the increased risk of researcher bias should be acknowledged. Lastly, heterogeneity of included studies resulted in challenges synthesising and comparing findings. However, the MMAT (Hong et al., 2018) allowed the quality of these studies to be evaluated effectively and the range of methodological designs may be considered a strength in providing converging levels and sources of evidence.
Implications and recommendations
The issues raised regarding the validity of the measures of self for dementia included in this review have important implications for how these instruments are used in research and clinical settings with people living with dementia. Clinicians and researchers should take caution when interpreting unvalidated measures of self for people living with dementia and further research is needed to investigate the psychometric properties of these measures in dementia. Future research should explore potential differences in outcomes for aspects of self following psychosocial interventions in relation to demographic variables such as gender, culture, stage, and subtype of dementia. Due to the heterogeneity of the studies, it was not possible to draw conclusions regarding the effectiveness of interventions on the different aspects of self. Future psychosocial interventions that are conceptually grounded and measure corresponding self-reported outcomes relating to aspects of self using high quality methodological designs, such as RCTs, are needed and will help to add to the quality and expansion of the existing evidence base in this area.
Conclusions
Research with people living with dementia is increasingly considering the effectiveness of psychosocial interventions at providing positive outcomes related to aspects of self. However, the effectiveness of these interventions appears mixed and the paucity of validated measures of self in dementia is a major limitation. This review highlights the need for further research in order to improve the evidence base supporting the use of self-report measures of aspects of self in dementia.
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Supplemental Material
Supplemental Material - Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions
Supplemental Material for Self-report instruments measuring aspects of self for people living with dementia: A systematic literature review of psychosocial interventions by Jessica Baggaley, Emma Wolverson, and Chris Clarke in Dementia
Footnotes
Declarations of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
