Aims: To compare the prevalence of anosognosia in patients with Alzheimer's disease (AD) using two different evaluation methodologies, to analyse the discrepancy between them and to determine the variables associated with this discrepancy.
Patients and methods: The methodology involved a cross-sectional, observational study conducted in patients with AD. The Anosognosia Questionnaire-Dementia (AQ-D) and the Experimenter Rating Scale (ERS) were administered. Clinical and sociodemographic information was recorded about both the patient and his/her caregiver by means of standardised instruments. Prevalence was determined with each instrument and the degree of agreement between them was then established. A variable called 'agreement/non-agreement' was created and a logistic regression model was adjusted to determine the variables associated to the discrepancy.
Results: The prevalence of anosognosia in the mild cases was 13.6% (95% confidence interval, 95% CI = 5.5-21.7) and 17.3% (95% CI = 8.4-26.1), and 44.2% (95% CI = 28.2-60.2) and 55.8% (95% CI = 38.8-71.8) in the moderate cases, according to the ERS and the AQ-D, respectively. The degree of agreement between the two measurements yielded a kappa-value of 0.7. The regression model with the discrepancy between the ERS and the AQ-D as the dependent variable was associated to a lower functional capacity (odds ratio = 0.080; 95% CI = 0.855-0.997; p = 0.042).
Conclusions: The data from the study showed an increase in the prevalence of anosognosia linked to the severity and a higher disagreement in its evaluation with two different methods as the degree of the patient's functional disability increased.