Introduction

Dementia is a neurodegenerative condition that mostly affects older adults. Prevalence increases with age, although a small proportion of cases affect people before the age of 65 and are classified as “early-onset” (Tindall & Manthorpe, 1997). It is estimated that currently at least 35 million people worldwide have a form of dementia, and this is expected to rise to over 115 million people by the year 2050 (Welsh Assembly Government, 2011). There are several types of dementia; Alzheimer’s Disease accounts for approximately 62% of all cases of dementia, vascular dementia for 17%, mixed dementia for 10%, and the remaining 11% includes fronto-temporal dementia and dementia with Lewy Bodies (Dementia UK, 2007). Diagnostic criteria vary across different types of dementia, but memory difficulties are commonly observed through objective memory testing, while other cognitive impairments such as aphasia (problems with speech, such as word-finding) or executive impairment (such as issues with planning or problem-solving) may also be present. Such cognitive difficulties are severe enough to result in impairments in daily living and represent a significant decline from a higher level of functioning. The course of the disease varies according to the type of dementia; some types have a gradual onset and a continuing cognitive decline, such as Alzheimer’s Disease, whereas other types tend to have a more sudden onset and a stepwise deterioration, such as vascular dementia (Groves et al., 2000).

Why is ‘the self’ important in dementia?

Much research has been undertaken to support policy guidelines concerning the care of people with dementia, and these have evolved from concentrating on physical needs to also considering relevant psychological and social needs. The National Institute for Health and Clinical Excellence guidelines (NICE, 2011) suggested that non-pharmacological therapies should be considered for people with dementia who exhibit challenging behavior such as aggression or agitation, or for those who show signs of depression. Policy writers have recently started to recognize the importance of “person-centered” care for people with dementia, which has been heavily influenced by the work of Kitwood (1997) and Fazio (2008), who strongly believed that maintaining the self was key to maximizing the well-being of individuals with dementia. The NICE guidelines (2011) state that the principles of person-centered care underpin good practice in supporting people with dementia, and that these principles reflect the value and individuality of each individual, including his/her unique personality and life history. The guidelines also highlight the importance of relationships and interactions with others, a person’s individual biography, including religious or spiritual beliefs and cultural identity, and other psychosocial factors for good quality person-centered care. A substantial amount of research has focused on “the self” (or “identity”) in people with dementia, and a range of definitions of these terms has been used, as will be discussed in this chapter. Research has also examined the implications of maintaining identity for high-quality care and optimal well-being of the person with dementia (e.g. Clarke, Hanson, & Ross, 2003; George, 1998; Harrison, 1993; Ronch, 1996).

How is ‘the self’ conceptualized with respect to people with dementia?

Researchers have described “the self” in various ways. Some of these definitions overlap while others appear to be based on completely different theoretical frameworks. One of the earliest distinctions to be made was between consciousness and self-awareness, which was addressed by both Mead (1934) and James (1950), who described “the self” according to whether attention is directed outwards through an ongoing stream of consciousness or directed inwards towards traits, characteristics, and preferences (“self-awareness”). Other researchers went on to elaborate on this distinction (e.g. Brown, 1976; Farthing, 1992) by defining various “levels” of consciousness.

More recently, models have been put forward where the complexity of information about the self is more detailed and can be accessed at differing levels of consciousness (Morin, 2006). There has often been a distinction between perceptual information about the self, such as information available directly through the senses, and conceptual information about the self, which is not immediately available through the senses and is therefore represented mentally at an abstract level (e.g. Neisser, 1988; Newen & Vogeley, 2003). Further research has used a level or “component” of the self identified by such models as the basis for measurements of the self for the purpose of quantification. For example, Caddell and Clare (2012, 2013) recently used Neisser’s model of self as a framework for their research with people with dementia. This is constructed from five different components: the “ecological self,” which represents the self with respect to the physical environment; the “interpersonal self,” which represents the self as engaged in social interaction; the “extended self” is the self as it is experienced across time and can involve personal memories; the “private self,” which is based on conscious experiences not available to anyone else (e.g. thoughts, feelings, dreams); and the “conceptual self,” which is made up of abstract representations about oneself such as traits, characteristics and autobiography, and is effectively a drawing together of the other four types of self.

Neisser’s model of self has also been used in research concerning other disorders, and the components of the model may be impacted upon differently according to the features of the disorder. For example, the hallucinations and delusions that can be a feature of schizophrenia may interfere with the “ecological self,” in that the person may believe that they are elsewhere, and the “private self,” because thoughts and feelings may be disturbed. In autism, the “interpersonal self” is likely to be affected, because there are differences in how the person responds to other people. Because the main feature of many types of dementia is memory loss, the “extended self” is likely to be compromised, which in turn would impact upon the “conceptual self” (Neisser, 1997).

Other researchers have based their work on a range of models, many of which overlap with one or more of the components of Neisser’s model. Examples include the social constructionist model, which posits that language is of fundamental importance in the creation of the self (Fazio & Mitchell, 2009; Li & Orleans, 2002; Sabat & Harre, 1992), and embodied selfhood (Kontos, 2004), which suggests that the body itself is an important source of selfhood, thus challenging the notion that the self is based solely on cognitive abilities. Various interactionist models have been used to study the self in people with dementia, based on the theory that the self is constructed within the context of social interactions and relationships (e.g. Fontana & Smith, 1989; Hubbard, Cook, Tester & Downs, 2002; Saunders, 1998). Other studies have focused on autobiographical memory as the basis for the self (Addis & Tippet, 2004) or the ability to put together a narrative drawing together a combination of significant life experiences (Mills, 1997; Usita, Hyman, & Herman, 1998). A number of qualitative studies have been conducted with people with dementia, with the aim of understanding people’s thoughts and feelings about their own selves using interview techniques (e.g. Beard, 2004; Caddell & Clare, 2011a; Clare, 2003; Gillies & Johnston, 2004; Li & Orleans, 2002). Numerous other studies have asked participants to rate what the researchers believe to be features of the self, such as role identities (e.g. Caddell & Clare, 2012, 2013; Cohen-Mansfield, Golander & Arnheim, 2000), or personality traits (Klein, Cosmides, & Costabile, 2003; Rankin, Baldwin, Pace-Savitsky, Kramer, & Miller, 2005; Ruby et al., 2007), which have often then been compared with parallel ratings from a carer to derive a measure of how accurately a person can describe him/herself. Finally, a number of studies have used self-recognition (for example, in mirrors, photographs, or videos) as a representation of how intact a person’s sense of self is (e.g. Biringer & Anderson, 1992; Bologna & Camp, 1995; Fazio & Mitchell, 2009, Grewal, 1994; Hehman, German, & Klein, 2005).

The persistence of self in dementia

It is challenging to draw firm conclusions regarding the self in individuals with dementia due to the range of theoretical frameworks described above, and also because of the resulting disparity in methods and participant samples used to measure aspects of the self. Thus, to date, there is limited agreement regarding the maintenance of the self in people with dementia.

Much of the older research concluded that participants experienced a “loss of self” (Cohen & Eisdorfer, 1986), an “unbecoming of self” (Fontana & Smith, 1989), or the “loss of all those qualities by which we have come to define our humanness” (Robertson, 1991). The basis for this argument tends to rest on the decline in cognitive abilities, particularly memory functioning, observed in people with dementia. However, a review by Caddell and Clare (2010) included 33 studies of people with a diagnosis of dementia, using the search terms “self,” “identity,” “personhood,” and “selfhood,” and came to several preliminary conclusions. The review suggested that almost all of the qualitative studies provided evidence to suggest that the self was preserved in people with dementia – it concluded that, during speech, participants could use personal pronouns and talk about their physical and mental attributes, as well as demonstrate their multiple personae and construct their identities in social interactions (e.g. Fazio & Mitchell, 2009; Hubbard, Cook, Tester & Downs, 2002; Li & Orleans, 2002; Sabat & Harre, 1992; Saunders, 1998; Small, Geldart, Gutman, & Scott, 1998; Tappen, Williams, Fishman, & Touhy, 1999). The only qualitative study to contradict this finding is that of Fontana and Smith (1989), who controversially contended that the self erodes until only “emptiness” is left. The review also examined the quantitative studies, and concluded that many of these studies presented evidence for the persistence of self, but simultaneously suggested that some results pointed to a deterioration in the self, or proposed aspects of the self, such as role-identities (e.g. Cohen-Mansfield et al., 2000), self-recognition (e.g. Biringer & Anderson, 1992; Grewal, 1994) and self-knowledge (e.g. Klein et al., 2003; Rankin et al., 2005; Ruby et al., 2009). Overall, the review concluded that the majority of existing studies found evidence for the persistence of self, at least to some degree, but that there may be a deterioration in various components of the self, possibly linked to the stage or severity of the illness. However, these conclusions are based on evidence that is limited in several respects and is therefore rather disparate and difficult to interpret.

Despite evidence for the persistence of self in people with dementia, many existing studies also highlight a deterioration in some aspects of self, as mentioned above (e.g. Caddell & Clare, 2010; Fontana & Smith, 1989; Mills, 1997; Small et al., 1998). A recent study (Caddell & Clare, 2012) measured levels of anxiety, depression, and quality of life as well as aspects of identity (self-concept, role-identities, self-knowledge, and autobiographical memory functioning) in their sample of 50 people with dementia. In general, higher scores on measures of identity, indicating a stronger sense of self, were associated with lower levels of anxiety and depression, and a better quality of life.

The role of self for well-being in individuals with dementia remains to be examined. Further, there are very few research studies on the whether interventions that target the self in people with dementia improve well-being in such populations (also see a review by Caddell & Clare, 2011b).

Interventions developed to support the self in people with dementia

This section will review the interventions conducted as part of research studies where the researchers were specific about the aim of the study (or one of them) being to support self or identity in participants who had a diagnosis of dementia, or the intervention is at least discussed with reference to its impact upon self or identity. These interventions have been broken down into several groups according to their features. It is important to note that in much of this literature, the terms “self,” “identity,” and “personhood” have been used interchangeably.

Firstly, there have been three group interventions that aimed to support self or identity in participants but failed to use outcome measures (Harlan, 1993; Jensen & Wheaton, 1997; Johnson, Lahey, & Shore, 1992). These three studies were all based around the use of art materials. One of the studies also included a variety of music and movements at the beginning of each class before the participants began using the art materials, and often allowed time for discussion at the end of each class (Jensen & Wheaton, 1997). Of these studies, only Johnson et al. referred to a specific framework or concept of self, by referring to memory as playing an important role in the maintenance of identity. There was little information available about the participants in these interventions, such as the number of people involved or the severity of the dementia of those who participated, which again made it challenging to draw very specific conclusions from these studies. This difficulty is exacerbated by a lack of detail about the intervention, such as the time spent participating by each person. However, the researchers were able to draw some conclusions from these studies. Jensen and Wheaton (1997) reported improvements in activity level and self-esteem during weekly 90-minute sessions in a nursing home, which included music, movement, and visual art. They suggested that the classes acted as a reminder of participants’ healthier, more productive selves such that remote memory was stimulated. Harlan (1993) noted that the intervention provided an opportunity for communication and enhanced motivation, confidence, and feelings of identity, and suggested that this might partly be due to the group nature of the sessions. Johnson et al. provided regular creative art therapy sessions and focused on remaining strengths and abilities to increase self-esteem, as well as offering opportunities for expression and spontaneity. Although there were no comparison groups or outcome data for these three studies, therapists reported a variety of positive results, and no drawbacks were noted.

A further three intervention studies have been conducted where supporting self or identity was not specified as a direct aim of the study, but the researchers explained why the intervention might be expected to be supportive of self or identity (Irish et al., 2006; Sherratt, Thornton & Hatton, 2004; Yasuda, Kuwabara, Kuwahara, Abe, & Tetsutani, 2009). Yasuda et al. and Irish et al. believed that using reminders of a person’s life to stimulate autobiographical memory (i.e., photos or music) could lead to an improved sense of identity, while Sherratt et al. also believed that listening to music has the potential to maintain personhood. Thus, these studies all involved participants listening to music, but each study involved more than one condition. One intervention involved participants watching a personalized reminiscence photo video, with music and narration added, before or after which they would also watch two other types of TV show, so that the researchers could compare responses across conditions (Yasuda et al., 2009). Yasuda et al. noted that the participants’ concentration scores were highest while their personalized reminiscence photo video was being shown. Irish et al. (2006) used music as a background stimulus while 10 participants completed measures of anxiety and autobiographical memory, which were also completed by each participant in a silent condition at another time. They found that in the music condition people with dementia (but not control participants) reported lower levels of anxiety and a significantly better score on the measure of autobiographical memory. Sherratt et al. (2004) compared the effects of live music and taped music on well-being in 24 people in the moderate to severe stages of dementia, and found improved lengths of responding and higher levels of well-being during the live music condition compared to the other conditions. Thus, the researchers from all interventions noted some benefits of one specific condition over others, highlighting the benefits of certain types of activity over others. However, no outcomes specifically related to the self or identity were reported.

There were four further studies that were very specific about the aim of the intervention being to support self or identity, and which used outcome measures for at least some variables, which enabled more precise conclusions to be drawn from each study. As the studies were rather diverse, they will be discussed separately. An intervention based on the concept of “self-maintenance therapy” was developed by Romero and Wenz (2001), who conceptualized the self as a cognitive schema which can store and update information about the person and his/her environment. The intervention consisted of a four-week residential programme that was attended by both the person with dementia and their primary caregiver. The intervention attempted, among other objectives, to maintain personal identity and continuity in people with dementia. The aim was to achieve this through four different components of the programme: psychotherapy, training in self-knowledge, the facilitation of everyday activities, and communication in caregiving. A range of measures were completed by the 43 participants with moderate-stage dementia before and after the intervention; these showed a reduction in depression, psychopathological symptoms, and disturbances in social behavior. No significant changes were found in memory functioning, activities of daily living or self-care. As no measures specifically targeting aspects of self or identity were included in the study, it is difficult to assess whether improvements in these variables occurred.

Cohen-Mansfield, Parpura-Gill, and Golander (2006) tested an individualized treatment based on role identities, whereby each participant was assessed to determine their most prominent identity role using the Self-Identity in Dementia Questionnaire (SID; Cohen-Mansfield et al., 2006). An activity related to this specific identity role was then developed and was delivered to participants for 30 minutes each day for 5 days. The researchers defined identity as the roles that people take on throughout their lives, which are demonstrated through both speech and behavior, and four main roles were defined: occupational roles, family roles, leisure activities, and attributes. A control group was also used in this study, where participants received the usual activities and care. Measures were completed before and after the intervention by research assistants and care staff. A greater awareness of identity was recorded in the treatment group post-intervention by the research assistants but not by participants themselves, and greater pleasure and interest in activities were observed in the treatment group relative to the control group. No changes in anxiety or depression were seen post-intervention.

Haight, Gibson, and Michel (2006) created a life review/life storybook intervention in which 15 people in care homes participated, as well as 15 people in the control group who did not receive the intervention. The researchers believed that it would be possible to “preserve personhood” by reinforcing personal identity both through conversation and also through retaining tangible reminders of experiences from a person’s lifetime. Those in the intervention group participated in weekly life review sessions for eight weeks and also created a life storybook with help from care staff, using photographs and captions. A variety of measures were used, pre- and post-intervention, and significant improvements were recorded on cognition, depression, positive mood, and communication in the intervention group. No differences were seen in independence, memory, or behavioral problems between the two groups, and again, no specific measures of aspects of self or identity were used, making it difficult to ascertain whether actual changes in self or identity occurred.

The final study was conducted with a single participant in the moderate stages of Alzheimer’s disease and involved developing and installing an in-home display of the participant’s life history on a screen called a “biography theatre” (Massimi et al., 2008). The participant spent eight sessions with the researcher, selecting photographs and arranging them into themes, which were then displayed on a touch-screen computer in the participant’s own home which was permanently left on and was accessible at all times. Measures were completed at baseline, after collection of the biographical material, and four weeks after installation of the biography theater. Researchers noted an improvement in identity on the interim and final assessments, which was recorded on the Self Image Profile–Adult (Butler & Gasson, 2004).

Thus, these latter studies have been clearer about their aim of supporting self or identity, but as with some of the studies described earlier, measures of self/identity have not always been used. Often studies have focused on various aspects of well-being, and it appears that a sense of identity is automatically being equated with well-being in some studies. While it is of interest to see whether the intervention is responsible for changes in the well-being in participants, without a specific measure of identity it is not possible to determine whether there have been changes in self or identity, or whether such changes did in fact mediate any changes in well-being. This makes it challenging to evaluate the intervention in terms of its stated aims.