25 Biases in Informant Ratings of Functional Abilities in Mild Cognitive Impairment: Predictors of Caregiver-Rated Functional Abilities While Controlling for Patient Objective Cognitive Status

Abstract

Objective:

An understanding of factors that contribute to informant ratings of patients’ functional abilities is crucial, not only because these ratings are used to diagnose individuals with mild cognitive impairment (MCI) versus dementia, but also because these ratings are commonly used as outcome measures in clinical trials. While these ratings are assumed to be largely accurate, research shows they are subject to biases. Caregiver distress, higher caregiver educational attainment, and higher patient age are associated with a higher discrepancy between informant and patient reports of functional abilities. Studies on informant ratings of functional abilities that simultaneously control for patient objective cognitive abilities remain sparse. The current study aims to evaluate caregiver characteristics as predictors of informant-rated functional status while controlling for patient objective cognitive abilities in MCI.

Participants and Methods:

Individuals with a clinical diagnosis of MCI (Albert, 2011 criteria) were referred to the Cognitive Empowerment Program (CEP), a comprehensive lifestyle program addressing modifiable risk factors associated with progression. This study included cross-sectional data from 118 newly enrolled individuals and their caregivers who served as informants. Patient cognitive functioning was assessed with the Montreal Cognitive Assessment (MoCA). Predictors of interest included caregiver-rated functional abilities (Functional Activities Questionnaire; FAQ), caregiver burden (Zarit Burden Interview; ZBD), caregiver depressive symptoms (Center for Epidemiological Studies Depression scale; CES-D), caregiver stress (Perceived Stress Scale; PSS), and caregivers’ self-rated communicative effectiveness (Communicative Effectiveness Index; CETI). Hierarchical linear regression models were run to predict FAQ while controlling for patient MoCA scores. Separate models were run for the caregiver variables of interest including caregiver age, ZBD, CES-D, PSS, and CETI.

Results:

Caregivers were 75.6% spouses, 17.1% adult children, 3.3% unmarried partners/cohabitating partners, and 4.1% friends. The mean age of individuals with MCI was 74.7 years (SD: 6.96, mean education = 16.2±2.60 years; 47% female) and the mean age of caregivers was 66.4 (SD: 12.88, mean education = 16.3±2.34; 66% female). Worse ratings of functional abilities on the informant-rated FAQ were found for patients with lower MoCA scores (ß = .242, p = .008). Importantly, while controlling for MoCA scores, worse ratings of functional abilities on the FAQ were found for informants with lower age (ß = -0.269, p = .003), higher perceived stress (ß = 0.267, p = .003), higher caregiver burden (ß = 0.289, p < 0.001), and lower self-rated communication effectiveness (ß = -0.324, p < .001). Caregiver depression (ß = 0.089, p = .084) and education (ß = -0.137, p = .147) were not significant predictors of functional ability ratings while controlling for MoCA scores.

Conclusions:

Results of the current study highlight the potential for biases in informant ratings regarding functional abilities in MCI. Informant ratings were found to be significantly influenced by caregiver age, stress, burden, and communicative effectiveness. A key finding is that younger caregivers, such as adult children, may report greater functional impairment in individuals with MCI. The current findings have implications for the use of perceived functional ratings, both for diagnostic purposes and as outcome measures in clinical trials.