17 December 2001

mailto:l.clare@ucl.ac.uk

Neuropsychological rehabilitation is just as relevant for people with progressive disorders affecting cognitive functioning as it is for people with non-progressive brain injury. If we define the goal of rehabilitation as enabling people to ‘achieve an optimal level of physical, psychological and social functioning’, given any limitations imposed by injury or illness , then it is clear that this is an appropriate goal at any stage of a progressive disorder. Improving well-being implies an enhancement in quality of life, not only for the person with dementia but also for his or her family or caregivers. Indeed, it has been suggested that the broad concept of rehabilitation provides a suitable unifying framework for conceptualising intervention in dementia , and within this broad framework an understanding of neuropsychological functioning is an essential element in addressing the needs of the person and his or her caregivers.

In this chapter, following a brief overview of dementia, I will focus on four key questions:

Why is neuropsychological rehabilitation relevant for people with dementia?

Can people with dementia benefit from neuropsychological rehabilitation?

What is the role of neuropsychological rehabilitation in clinical practice?

The way forward: where do we go from here?

Dementia has been defined as ‘a clinical syndrome characterised by loss of function in multiple cognitive abilities in an individual with previously normal (or at least higher) intellectual abilities and occurring in clear consciousness’ . By implication, this decline in cognitive function will also impact on functioning in various domains of daily living and social interaction. There are numerous possible causes of dementia, but the most frequent dementia diagnosis is that of Alzheimer’s disease . Less frequently diagnosed sub-types of dementia include vascular dementia, dementia with Lewy bodies, and the frontal and temporal (semantic) variants of frontotemporal dementia. These have different neuropsychological profiles, particularly in the earlier stages, and consequently different implications for rehabilitation interventions .

Even within a single sub-type of dementia such as Alzheimer’s disease, there is considerable heterogeneity in both initial presentation and course . Memory is usually the first cognitive function to be affected, although impairments are initially evident only in certain memory systems, particularly episodic memory . In the early stages, attention , executive function and word finding may also be compromised. As the disorder progresses, deficits in these areas become more extensive, psychomotor function is affected, and a decline in global cognitive functioning becomes evident . Visuospatial perception is usually affected only in the later stages, although in atypical cases it may be observed as one of the earliest symptoms . The time course of progression is very variable, with some individuals staying in the mild stages for a number of years and others progressing more rapidly to severe impairment (e.g. . Some individuals develop mild cognitive impairment, with significant difficulties in the domain of episodic memory, but do not show any progression to dementia. Therefore, needs will differ according to both the neuropsychological profile and the extent to which the cognitive impairments have progressed in severity. Individual assessment, formulation and goal planning is always necessary, and a diagnostic or staging label does not suffice as a basis for developing rehabilitation plans.

In recent years a quiet revolution has been taking place in dementia care, and this has brought to the fore the concepts of personhood and person-centred care . The perspective of the person with dementia, hitherto largely neglected , is now being explored and valued alongside that of the family member or caregiver. Psychosocial and social constructionist models of dementia highlight the importance of the unique set of life experiences and coping strategies that each individual brings to the challenge of living with dementia, and the impact of the social environment on the expression and course of neurological impairment. Social and psychological factors are understood to interact with the effects of neurological impairment in a dialectical process, so that a person surrounded by a ‘malignant social psychology’ is likely to show ‘excess disability’ and appear more disabled than the extent of any brain pathology would indicate ought to be the case. Reducing excess disability is therefore an important target for intervention. Rehabilitation has an important role to play in tackling excess disability and enhancing well-being and quality of life for the person with dementia and his or her family. Further, it has been argued that selfhood should not be thought of as lost in dementia , and rehabilitation may help to maintain both sense of self for the person with dementia and the perception of selfhood by others .

The context in which rehabilitation occurs is therefore crucial. In developing a model for neuropsychological rehabilitation in dementia a range of factors need to be taken into account above and beyond the person’s neuropsychological profile. I have argued elsewhere that neuropsychological rehabilitation in dementia requires a psychotherapeutic framework, equivalent to the ‘holistic’ approach taken in brain injury rehabilitation by Prigatano . It is essential to acknowledge the person’s emotional responses and coping strategies, and to work with these . Equally, rehabilitation in dementia requires a systemic perspective, in which the person and the impact of dementia can be viewed in the context of the person’s network of support and care. In many cases, caregivers will be essential allies if the rehabilitation process is to be effective. It is also vital to consider that dementia is typically – although not always – a problem of later life. Rehabilitation for older people with dementia should not be undertaken without an understanding of ageing and its psychological and social implications. Finally, it is necessary to be sensitive to differences in the way dementia is perceived across diverse cultural and religious groups, and to respect values and expectations that may diverge from the way in which dementia is typically viewed by western health professionals.

The essentially collaborative and individually-targeted nature of rehabilitation means that it sits easily within this contextual framework. In addition, there is a strong rationale for the specific relevance of neuropsychological rehabilitation for people with at least some forms of dementia. This derives from evidence regarding the neuropsychology and neuroanatomy of cognitive impairments, and the capacity of the person with dementia for new learning. The memory impairments of early-stage Alzheimer’s disease provide a useful illustration of this rationale.

Memory can be considered in terms of dissociable systems , distinct processes, and neuroanatomical structures. Within long-term memory systems, episodic memory is typically severely impaired in early-stage Alzheimer’s . Semantic memory is less likely to be significantly affected, although impairments may be observed in some individuals . Procedural memory is preserved , as are some aspects of priming . Within working memory systems, the central executive is typically impaired in early-stage Alzheimer’s , as is the visuospatial sketch-pad, but the phonological loop appears unaffected . The observation that some aspects of memory are severely impaired while others are relatively unaffected suggests that there is scope for interventions aimed at improving memory functioning that can build on the intact aspects of memory, as well as for interventions aimed at developing ways of compensating for impaired aspects of memory.

Memory can also be considered in terms of the processes of encoding, storage and retrieval. The memory problems of Alzheimer’s are not primarily problems of storage; that is to say, rates of forgetting over the longer term are no greater than those found in healthy individuals. The major impairment is thought to lie in encoding, so that the acquisition of new memories is affected , although this of course does not account for retrograde memory loss , and therefore cannot be a complete explanation. However, if the main problems are in encoding, this suggests that if assistance is given with getting information into the memory store, there is a reasonable likelihood of retention.

In Alzheimer’s, pathology is found in the medial temporal lobe, basal forebrain, thalamus and neocortex . Within the medial temporal lobes, pathology in the medial temporal region results in a functional disconnection of the hippocampus from related areas at an early stage . The hippocampal structures are vital for establishing and consolidating new memories, and linking these with existing knowledge, but are not required for long-term storage. This may have important implications. suggests, for example, that other brain areas may be able to perform the function of integrating new semantic information with existing memories, provided that strategies are offered at encoding to facilitate the linking function. Similarly, where the frontal lobes are affected, people may be expected to have particular difficulty implementing strategies to help themselves remember, and interventions can attempt to compensate for this by providing additional support for remembering.

Understanding the neuropsychological profile of the person with dementia, therefore, allows us to derive some specific pointers that can assist in devising appropriate interventions. Experimental studies of learning confirm that learning is possible in people with dementia. Both classical and operant conditioning of responses has been demonstrated , as has retention of verbal information . However, explicit learning will be seen only where the conditions are favourable; argues that appropriate support for memory must be provided both at encoding and at retrieval (termed ‘dual cognitive support’), and notes that the level of support required will increase as the severity of dementia increases. A number of experimental studies demonstrate beneficial effects of different types of cognitive support. For example, memory performance is facilitated when multiple sensory modalities are involved at encoding or where participants physically enact the target task at encoding . Similarly, in accordance with the encoding-specificity principle, provision of retrieval cues that are compatible with conditions at encoding assists recall ; an example would be where a semantic orienting task is used at encoding (e.g. categorising ‘apple’ as ‘fruit) followed by provision of category cues at retrieval (e.g. ‘it’s a kind of fruit’) . Results presented by Lipinska and colleagues indicate that participants perform better with self-generated than with experimenter-provided cues. emphasise that personalising a task by allowing the participant to make choices about it increases perceived control and motivation, and consequently is likely to benefit performance.

A number of studies, then, demonstrate that elaboration and effortful processing can improve memory performance. At the same time, this needs to be balanced with the goal of reducing or eliminating errors during the learning process , as the principle of errorless learning has also been shown to be useful in improving memory performance in early-stage Alzheimer’s . Current knowledge about memory, and experimental evidence about learning and remembering in dementia, therefore, indicate that neuropsychological rehabilitation is relevant to, and potentially beneficial for, people with dementia. Consequently, there is a good rationale for developing clinical interventions using this approach.

Before considering how neuropsychological rehabilitation can best be put into practice in dementia care, it is important to understand something of the history of psychological interventions in dementia, and to consider how neuropsychological rehabilitation relates to other approaches. One of the earliest approaches to psychological intervention was the adaptation of reality orientation (RO) for use with people who have dementia . This represented a major breakthrough as it demonstrated the possibility that psychological approaches had something to offer. The concept of reality orientation attained widespread acceptance in long-term care settings, but its implementation was not always of a high standard, with interventions sometimes applied in a rather insensitive manner. Subsequently the approach was heavily criticised for overlooking the emotional needs of the person with dementia, and alternative methods emerged in the form of practices such as validation therapy . More recent work has demonstrated that interventions based on the principles of reality orientation can have positive effects on cognition and behaviour in people with dementia , and has attempted to ‘rehabilitate’ reality orientation through the development of group interventions for people with severe dementia in residential settings . Similarly, evaluation of ‘cognitive stimulation’ group interventions for people with dementia shows positive results . Because interventions based on RO or cognitive stimulation typically incorporate a number of elements, it remains difficult to determine the relative contribution of different components to the positive outcomes observed, or to derive a theoretical understanding of the mechanisms by which the interventions exert their positive effects.

In the meantime, the reality orientation tradition was succeeded by a related body of work focused on ‘memory training’ or ‘memory retraining’ for people with dementia. Memory training is similar to cognitive stimulation approaches, in that the goals set and tasks used are general rather than individually-designed, but differs in that there is a more specific focus on one or more aspects of memory functioning. Memory training has been criticised in turn, firstly on the basis that gains are limited and maintenance is poor, and secondly because it is said to have negative effects on mood and well-being for people with dementia or their caregivers . It should be noted that these criticisms were made in the context of a strong emphasis on the value of pharmacological treatments, although the effectiveness of the pharmacological treatments currently available for people with dementia is modest . Other reviews, in contrast, have argued that this kind of approach may be beneficial; for example, Gatz and colleagues classify ‘memory retraining’ as a ‘probably efficacious’ form of intervention which warrants more research. One reason for this divergence of views may be that increased understanding of the conditions required in order for people with dementia to learn effectively has resulted in more appropriate intervention protocols, as indicated by . There is a continuing interest in ‘memory training’ for people with dementia, with a variety of methods in evidence. For example, a recent randomised controlled trial (Davis, Massman, & Doody, 2001) reported gains in performance on targeted areas, but no generalised improvements. There is also a growing interest in computerised cognitive training for people with dementia (e.g. .

In the last few years, the focus has shifted once again and researchers have begun to apply the concepts of cognitive or neuropsychological rehabilitation to dementia care . This strand of research draws on the concept and practice of neuropsychological rehabilitation with brain-injured people. Interventions are devised on the basis of theoretical principles derived from neuropsychology, cognitive psychology and learning theory, and are targeted specifically to the individual on the basis of the person’s neuropsychological profile. In addition, following Prigatano’s holistic model of neuropsychological rehabilitation , they should endeavour to take into account the person’s emotional and practical needs and social context, drawing on perspectives from psychotherapy and systems theory.

It is necessary to distinguish the work carried out in the tradition of reality orientation, encompassing memory training and cognitive stimulation, from attempts at implementing neuropsychological or cognitive rehabilitation. At the same time, it is important to acknowledge that there are some shared elements and that knowledge or evidence derived from one can sometimes be applicable to the other. The review that follows will focus primarily on neuropsychological rehabilitation rather than on memory training and cognitive stimulation.

Neuropsychological rehabilitation can provide both a general framework for intervention and a means of tackling specific issues. As a general framework, it allows for a biopsychosocial formulation within which an understanding and acknowledgement of the person’s cognitive impairments is central. This means, for example, that explanations and advice can be provided to the person and his or her carers, helping them to make sense of some of their difficult and distressing experiences. Specific difficulties can be addressed using methods devised for people with dementia or adapted from those reported to be useful for people with brain injury. The way in which these two aspects of neuropsychological rehabilitation are implemented in practice varies according to the needs of the individual, so that the emphasis is likely to be quite different in early- and later-stage dementia.

In the early stages of dementia, the main focus for intervention is likely to be everyday problems arising from difficulties with long-term episodic memory or executive function. The discussion here will concentrate on memory functioning.

The most appropriate approach will be determined through a careful assessment of the neuropsychological profile and the person’s everyday functioning. This assessment should be made in the context of a broader evaluation encompassing the person’s past experience and preferred ways of coping, psychological well-being, awareness of difficulties and readiness to address them, and any other possible blocks to successful outcome that may need to be overcome before the intervention begins. The way in which the person perceives his or her difficulties is likely to be particularly important here, as it has been demonstrated that expressed awareness of memory difficulties and their impact is associated with better outcome in cognitive rehabilitation interventions in the early stages of dementia . The person’s support systems also need to be considered along with the willingness of family members or friends to be involved. The assessment should lead to a collaborative exercise in setting goals for intervention, in which the person with dementia plays a full part. Interventions are most likely to be effective when they address issues that are important to the person and family, and relevant to everyday life. Where the goals of the person with dementia and the family are markedly discrepant, careful and sensitive negotiation is required in order to try to reach a consensus that is acceptable to both parties, acknowledging the different emotional and practical needs of all involved.

Some people with early-stage dementia may already be engaging in self-help activities, for example those provided in books about improving memory. This can be facilitated through provision of appropriate material or suggestions. Information about memory problems and how these may be tackled (e.g. can be helpful for the individual and for family members, empowering them to identify their own solutions to specific issues or problems.

Consistent with memory rehabilitation in brain injury, specific interventions for memory difficulties in early-stage dementia take two main forms . Firstly, assistance can be given with learning or relearning information and skills, in order to enhance residual episodic or procedural memory performance. Secondly, strategies can be developed that enable the person to compensate for aspects of memory that are impaired and functioning poorly.

Interventions aimed at facilitating residual memory performance need to incorporate the twin guiding principles of effortfulness and errorlessness in the learning process. This can be achieved in practice by using one or more of a number of methods.

Expanding rehearsal, or spaced retrieval, has been used extensively with people who have dementia. The act of retrieving an item of information is a powerful aid to subsequent retention under any conditions. In addition, the temporal sequencing of retrieval attempts affects the extent to which benefits are observed as a result of retrieval practice, with maximum benefit occurring when test trials are spaced at gradually expanding intervals . Experimental studies have demonstrated that expanding rehearsal can aid new learning in people with memory disorders following brain injury . The method has been adapted for use in Alzheimer’s , with very short retrieval intervals - typically the first interval is 15 or 30 seconds long, and the length is repeatedly doubled. A series of studies have demonstrated clear benefits in teaching face-name associations, object naming , memory for object location, and prospective memory assignments . A further advantage of the expanding rehearsal method is that it can easily be used by caregivers, with back-up support from professionals as required . Expanding rehearsal does not rule out the possibility of errors occurring, but in practice, because the initial recall intervals are so short, errors are rare. It therefore approximates well to an errorless learning procedure, while also requiring the effort of retrieving the information.

Another method that can be applied is the use of cueing . This can take various forms. In one version, termed ‘vanishing cues’ or ‘decreasing assistance’ , the number of cues is gradually reduced. When learning a name, for example, this would mean that at each presentation an additional letter was removed from the end of the name. Another version, termed ‘forward cueing’ or ‘increasing assistance’ begins by offering just the initial letter and adds a letter on each subsequent presentation until the word or name is correctly given, after which the cues may be faded again as in the vanishing cues method. Cueing methods were used by Clare and colleagues , and were directly compared in one single case study, where forward cueing was found to be more effective than vanishing cues.

Strategies such as visual imagery mnemonics, chunking of information, the method of loci, the story method and initial letter cueing have been described in relation to the cognitive rehabilitation of memory disorders following brain injury, although some of these strategies may prove too difficult or demanding for many brain injured patients . There is limited evidence for the success of strategies of this kind in Alzheimer’s. People with Alzheimer’s are likely to have difficulty both in learning an explicit mnemonic strategy and in remembering to use it appropriately , although in some cases simple strategies may be remembered and implemented. It is, however, important to distinguish between the use of mnemonic strategies as a way of facilitating learning in specific tasks and the aim of developing spontaneous and independent use of the strategy in a wider sphere. The former is often a more appropriate goal in memory rehabilitation, and clinicians can draw on a number of strategies to facilitate learning for the person with dementia.

One report of successful use of a mnemonic strategy is provided by Hill and colleagues . They describe a single case experiment in which a 66 year old man with Alzheimer’s was taught to use visual imagery to extend his retention interval for names associated with photographs of faces. In an attempt to replicate the findings in a case series of eight participants, which included seven people with Alzheimer’s , only one of the participants with Alzheimer’s showed training gains similar to those demonstrated by . The remaining patients failed to benefit from training. The authors conclude that the generalisability of the approach appears limited, but comment that there might be a subgroup of Alzheimer’s patients who respond well to this form of memory training.

This finding was supported by the results of a single case study which demonstrated effective relearning of names that the participant wished to know using a combination of mnemonic, vanishing cues and expanding rehearsal. The gains were fully maintained at follow up nine months after the intervention. A further study evaluated forgetting over the subsequent two years and found that performance remained well above initial baseline levels three years after the end of the intervention . A similar approach again targeted goals identified by the participants, but explored the feasibility of different techniques in isolation . Good results were obtained for mnemonic, expanding rehearsal and forward cueing strategies, with vanishing cues proving less effective, as noted above. This is in accordance with the findings of , which showed that visual imagery was a more effective strategy than vanishing cues for the acquisition of face-name associations in memory-impaired participants (the sample included one man with dementia). Thoene & Glisky argued that this was because the mnemonic strategy was able to optimise the use of residual explicit memory, in addition to encouraging deep levels of processing and the development of associations with existing knowledge.

The ability to perform everyday skills is particularly important in maintaining independence. Zanetti and colleagues used a training method based on preserved procedural memory for rehabilitation of ADL skills in people with mild-to-moderate Alzheimer’s. Training involved comprehensive prompting, with subsequent fading out of prompts. Preliminary results suggested this approach could be effective and produced some generalisation of improvements to untrained tasks. Another study used individualised training programmes for activities of daily living and showed improvements in three out of four participants, although only one maintained the gains two months later. An important feature of this study was the selection of tasks which were part of the patient's usual routine and which the patient was motivated to carry out; the value of considering motivational factors in designing interventions was emphasised.

Providing external support for remembering in the form of compensatory memory aids can help to reduce the demands on memory. The selection and introduction of external memory aids requires careful consideration, and aids should be targeted as specifically as possible, rather than simply providing a generalised reminder, the reason for which may be unclear to the person with Alzheimer’s . Many people who develop dementia will already be used to relying on external memory aids such as diaries and lists, and it is helpful to build on this and try to ensure that these aids are used to maximum effectiveness; for example, a diary with unstructured pages may be replaced by one that has times of day listed. People with memory impairments are unlikely to start to use new memory aids spontaneously and usually need training in their use, for example by means of prompting and fading of cues, or expanding rehearsal. Effective use of an errorless prompting and fading method to help a woman with dementia use a calendar to find out what day it was instead of repeatedly questioning her husband was reported by .

Developing technology offers increasing opportunities for identification of ingenious aids to remembering. In an early example , a digital watch was set to beep every hour as a cue to prompt engagement in a predetermined activity. Use of technology is now being extended beyond the realm of specific memory aids by developing computer and video equipment to monitor and control the environment of the person with dementia in order to support independent functioning .

In clinical practice with people who have early-stage dementia, the methods and techniques of neuropsychological rehabilitation have been implemented in a variety of ways. As well as individual interventions such as those described above, a number of centres have developed group programmes aimed at helping people with early-stage dementia to cope with memory difficulties (e.g. . Some programmes offer parallel sessions for participants with dementia and caregivers, while in others couples attend sessions together (e.g. . Group programmes typically incorporate information and education about memory and cognitive problems as well as identification of individual goals, introduction of suitable strategies or aids, and practice in their use. When they work well, groups provide an opportunity for members to support and encourage one another, and perhaps develop friendships and social contacts. However, some people with early-stage dementia may be reluctant to attend a group, preferring one-to-one sessions, and individual preferences should be respected. Elements of cognitive rehabilitation have been incorporated in broad-based community rehabilitation programmes which also incorporate aspects such as partnered volunteering , and in psychosocial early intervention programmes (e.g. .

As dementia progresses, the focus of neuropsychological rehabilitation is likely to change to some extent. There is likely to be more emphasis on addressing behavioural issues and on enhancing well-being through maintaining interaction and engagement.

An understanding of the neuropsychological profile and the possibilities for new learning can be coupled with a behavioural approach that views behaviour as having a meaning or function rather than as a ‘symptom’. This provides a framework for generating creative but highly practical solutions where cognitive impairments appear to play a part in producing ‘problem’ behaviour. This framework has been used, for example, to teach patients to associate a cue with an adaptive behaviour as a means of reducing behaviours that are regarded as problematic .

In some situations, rehabilitation of basic skills is an important focus. Camp and colleagues describe the application of Montessori activities, designed to build skills in a developmental sequence in young children, to dementia care. An example here might be reinstating the ability to feed oneself with a spoon through a sequence of tasks starting with scooping beads with a large scoop, and progressing through scooping rice, sand and eventually liquids with gradually smaller scoops, and so on until a spoon can be used to spoon up soup.

A number of studies have demonstrated improvements resulting from the use of various external memory aids or equivalent environmental support for people with later stage dementia. In some cases these improvements have been maintained after the support has been withdrawn, while in other cases ongoing support has been required. trained in-patients with moderately advanced Alzheimer’s to use a diary, reality orientation board or personal notebook to find out important information, although it is unclear to what extent the improvement was maintained. evaluated the effectiveness of memory wallets in enhancing conversational ability in a small sample of people with moderately advanced Alzheimer’s, and reported significant improvements, with evidence of generalisation to novel utterances. Benefits were maintained at six-week follow up, and for three individuals benefits were retained after 30 months . This finding has recently been replicated with people who have severe dementia . As well as helping the memory-impaired person, memory wallets or memory books offer care staff a means of learning about, and engaging with, the person , and can be especially helpful at times of transition, such as the move into residential care. emphasise the importance of helping the person with dementia to maintain a sense of self, using materials such as memory books in a structured way to facilitate engagement with, and processing of, those aspects of self that are currently most salient for the individual. In this approach, too, memory books can become the focus for constructive interaction between the person with dementia and family members or carers.

Implementing cognitive rehabilitation in later-stage dementia requires particular attention to, and skill in, working with systems. Camp has taken an important lead in addressing practical issues regarding the implementation of rehabilitation interventions in long-term care settings , including issues of cultural and linguistic difference.

This chapter has shown that neuropsychological rehabilitation can be applied in progressive disorders such as dementia with beneficial results. The approach is relevant for both earlier and later stages of dementia, but the focus differs according to the needs of the individual and his or her carers at any given point. Comprehensive recent reviews of the literature support the relevance of this kind of approach in dementia care , suggesting that further research is warranted, and a Cochrane systematic review is in preparation .

The application of neuropsychological rehabilitation for people with progressive disorders such as dementia is a relatively recent development, and there are a number of issues that future research will need to address. At a conceptual level it will be important to ensure that cognitive rehabilitation is clearly distinguished from related, but different, approaches such as reality orientation or memory training. At a practical level, it will be necessary to continue refining our knowledge of methods and techniques that may assist in achieving specific goals. Equally, it will be important to further develop the ‘holistic’ framework for cognitive rehabilitation with people who have progressive disorders, ensuring that emotional needs and responses are attended to, and that the person is considered in the context of his or her social system. It will be vital to identify more clearly the factors that indicate whether or not this kind of approach is likely to be suitable for a given individual at a given time, or whether some other form of support or intervention would be more appropriate. Finally, it will be necessary to situate neuropsychological rehabilitation within a coherent approach to supporting people with progressive disorders that reflects a genuinely biopsychosocial model and espouses the aims and values of person-centred care

References

                    Abrahams, J. P., & Camp, C. J. (1993). Maintenance and generalisation of object

                    Zanetti, O., Zanieri, G., Giovanni, G. d., Vreese, L. P. d., Pezzini, A., Metitieri, T., &       Trabucchi,             M. (2001). Effectiveness of procedural memory stimulation in mild Alzheimer's disease patients: a controlled study. Neuropsychological Rehabilitation. Special Issue: Cognitive Rehabilitation in Dementia, 11, 263-272.