Multidimensional Assessment of Awareness in Early-Stage Dementia: A Cluster Analytic Approach

Background/Aims: Research on awareness in dementia has yielded variable and inconsistent associations between awareness and other factors. This study examined awareness using a multidimensional approach and applied cluster analytic techniques to identify associations between the level of awareness and other variables. Methods: Participants were 101 individuals with early-stage dementia (PwD) and their carers. Explicit awareness was assessed at 3 levels: performance monitoring in relation to memory, evaluative judgement in relation to memory, everyday activities and socio-emotional functioning, and metacognitive reflection in relation to the experience and impact of the condition. Implicit awareness was assessed with an emotional Stroop task. Results: Different measures of explicit awareness scores were related only to a limited extent. Cluster analysis yielded 3 groups with differing degrees of explicit awareness. These groups showed no differences in implicit awareness. Lower explicit awareness was associated with greater age, lower MMSE scores, poorer recall and naming scores, lower anxiety and greater carer stress. Conclusion: MultidiAccepted: March 7, 2011 Published online: April 21, 2011 Prof. Linda Clare School of Psychology, Bangor University Bangor, Gwynedd LL57 2AS (UK) E-Mail l.clare @ bangor.ac.uk © 2011 S. Karger AG, Basel 1420–8008/11/0315–0317$38.00/0 Accessible online at: www.karger.com/dem D ow nl oa de d by : 54 .7 0. 40 .1 1 11 /1 9/ 20 17 2 :5 0: 50 P M Clare /Whitaker /Nelis /Martyr /Markova / Roth /Woods /Morris Dement Geriatr Cogn Disord 2011;31:317–327 318 tency may in part reflect variations in the conceptualisation of awareness, differences in the object of awareness selected, and differences in the type of assessment measure used. The range of terminology used to describe states of reduced awareness, including ‘lack of insight’, ‘anosognosia’ and ‘denial’ [1] , is indicative of different underlying assumptions. Since these terms are often used interchangeably and without explicit clarification of the theoretical connotations, this creates a potentially confusing picture. The lack of conceptual clarity evident in this field has hindered the development of comprehensive explanatory models [8] , and many studies fail to outline a clear conceptual framework or to provide a precise operational definition of awareness. Attempts at explanation have typically focused on specific elements such as cognitive processes [9] or defensive denial [10] ; however, there is a need to develop a comprehensive framework to account for a range of influences on the expression of awareness, both biological and psychosocial or contextual [11] . Hitherto, the most frequently studied objects of awareness in dementia are memory and activities of daily living (ADLs) [12] , either discretely or in combination [13, 14] . Some attention has also been given to social functioning [5] . Findings suggest that the extent of awareness differs in relation to different objects [15] and that awareness phenomena may be domain specific. The main methods of assessing awareness are general ratings of awareness based on interviews conducted by a clinician or researcher [16] , calculation of discrepancies between patient and informant ratings on parallel questionnaires [17] , and calculation of discrepancies between patient estimates of performance and actual task performance [18] . All have significant limitations [12, 19, 20] . In view of this, some studies have compared findings from two different types of measure, typically participant-informant discrepancies and participant-test discrepancies [21–24] . Others have combined one or both of these with clinician ratings of awareness and in some cases qualitative interview data [25–29] . Awareness scores appear to differ according to the assessment method used, and awareness scores obtained from different methods are typically not highly correlated [14, 30] . As a result of these conceptual and methodological difficulties, current empirical evidence provides only a limited basis for determining how the degree of awareness shown by a PwD may be evaluated in a meaningful and potentially useful way. This situation would be improved by use of relevant theoretical models, provision of clear operational definitions and precise selection of objects. For the purposes of the present study, awareness will be broadly defined as ‘a reasonable or realistic perception or appraisal of a given aspect of one’s situation, functioning or performance, or of the resulting implications’ [31] . Awareness will be examined in the context of a theoretical framework which specifies that in earlystage dementia this broad definition encompasses cognitive-affective operations at 3 main levels: performance monitoring in relation to selected tasks, evaluative judgement about aspects of functioning such as memory or everyday activities, and meta-cognitive reflection, for example in relation to the impact and implications of the condition. For research purposes, it is necessary to identify precisely the specific clinical phenomena that will be elicited empirically [1] . In this study, awareness at each level will be assessed in relation to a clearly identified object or objects. A multiplicity of levels and objects calls in turn for an appropriate range of measurement approaches. In addition to the standard discrepancy and rating methods, based on directly eliciting explicit responses, there is scope for further methodological development. For example, the use of more indirect methods of accessing participants’ subjective experience of, and knowledge about, the condition might be experienced as less threatening than direct questioning, and hence elicit evidence of greater awareness. Similarly, where awareness is not expressed explicitly, signs of awareness may still be present at an implicit level [9, 32] . This might be evident at the level of behavioural adaptation, or at the conceptual level, with tacit knowledge about the condition expressed through reference. Therefore, if there is evidence that condition-salient information affects cognitive processing and consequently impacts on behavioural responses in the absence of explicit expressions of awareness, this might indicate that awareness is retained at an implicit level. Here a dissociation between implicit and explicit awareness scores might be anticipated. This study incorporated novel assessment methods to measure these aspects of awareness alongside more established techniques. Given that individual scores are likely to differ across different levels and objects of awareness and when assessed using multiple measurement approaches, the question then arises as to whether it is possible to identify any meaningful groupings among participants on the basis of a range of measures tapping different awareness phenomena [33] . This study will attempt to answer this question using a cluster analytic approach. The following specific research questions will be addressed: D ow nl oa de d by : 54 .7 0. 40 .1 1 11 /1 9/ 20 17 2 :5 0: 50 P M Multidimensional Assessment of Awareness in Early-Stage Dementia Dement Geriatr Cogn Disord 2011;31:317–327 319 (1) How do results from different assessment methods applied at different levels of awareness to different objects of awareness relate to each other? (2) Is there a dissociation between the scores achieved on explicit and implicit measures of awareness? (3) Is it possible to group participants according to degree of awareness by means of cluster analysis when using a range of explicit awareness measures targeting different levels and objects of awareness? (4) If so, does the resulting grouping differentiate participants on other characteristics sometimes found to be related to awareness: memory, language and executive function, mood, quality of life and degree of carer stress?