Two perspectives, one disease: discrepancy scores for measuring anosognosia in Alzheimer's disease

Anosognosia (i.e., the inability to recognize the presence of impairments characteristics of a neurological disease) represents a frequent symptom of Alzheimer’s Disease (AD). Its presence exerts a profound impact on both patients and their caregivers, influencing disease progression and complicating patient management. Currently, several methods are employed to evaluate anosognosia, and no consensus has yet been reached regarding which is the most accurate. Among them, the patient-informant discrepancy method quantifies anosognosia as the signed difference between ratings given by an informant (usually the caregiver) and those given by the patient on a questionnaire regarding the patient’s abilities, with positive values indicating greater impairment reported by the caregiver. However, a major limitation of this method is represented by the possible cancellation of opposing discrepancies across different items of the same questionnaire, leading to the misclassification of patients who are unaware of only some but not all of their abilities (domain-specific anosognosia). The first aim of this study was to compare different methods for measuring anosognosia in a cohort of patients with AD and to examine their relationships with clinical variables concerning the patient as well as with variables concerning the caregiver. The second aim was to address the cancellation problem inherent to the patient-informant discrepancy method, in order to improve the characterization of patients’ anosognosia. Thirty-eight patients diagnosed with AD were recruited for our study. All participants underwent clinical and neuropsychological evaluation, and their level of anosognosia was assessed using the Clinical Insight Rating Scale (CIRS), the Anosognosia Questionnaire for Dementia (AQ-D) and the Everyday Cognition (ECog) questionnaires. A novel score, termed the Internal Inconsistency (II) index, was developed for use alongside traditional discrepancy measures. The II index increases when substantial item-level cancellation occurs, while also accounting for overall disagreement between the patient and the informant. Correlation analyses were performed between anosognosia measures, neuropsychological variables and the II index. Caregiver’s depression and burden were strongly related with each anosognosia score and subscore. Clinical assessment of anosognosia (CIRS) was also associated with the degree of cognitive impairment. Anosognosia measures positively correlated with one another, and each subscore of the discrepancy-based measures (AQ-D and ECog) correlated both with their total score and with the other subscores of the same questionnaire. Application of the II index to our sample revealed the presence of domain-specific anosognosia in patients who would have been otherwise classified as fully aware. No relationship was observed between the II index and any clinical or neuropsychological variable, suggesting that this index conveys an independent information. Furthermore, a significant correlation was found between the II index applied to the AQ-D and the II index applied to the ECog. Our results demonstrate the significant impact of anosognosia on the development of caregiver’s depression and burden, underscoring the importance of accurately assessing unawareness in order to implement adequate supportive interventions for caregivers. Our findings also highlight how the discrepancy-based measures employed for anosognosia assessment capture distinct aspects of unawareness, even when mutually correlated, thus confirming the absence of a clearly superior method. Finally, the present research suggests that the II index is a valuable tool to be used alongside traditional discrepancy measures to facilitate the identification of domain-specific anosognosia.