Key points are not available for this paper at this time.
Abstract Objective Informant-report measures often do not correlate with neuropsychological tests measuring the same cognitive construct, yet they may offer insight into everyday difficulties not captured by objective tests. Performance-based measures of functional ability, developed to capture real-world skills, may better align with informant-report measures. Clarifying the relationship between subjective and objective assessment tools in older adults could enhance precision in selecting the most appropriate measures. Method As part of a comprehensive neuropsychological battery, community-dwelling older adults who were cognitively unimpaired or who presented with subjective cognitive decline or mild cognitive impairment completed tests of executive functioning (EF), (D-KEFS Trail Making Test 4, Letter Fluency, WCST, WAIS-Comprehension) and financial capacity, Financial Capacity Instrument (FCI-SF). Informants responded to measures of functional ability, Amsterdam Instrumental Activities of Daily Living Questionnaire Short Version (A-IADL-Q-SV) and Assessment of Functional Capacity Interview (AFCI) and items capturing EF on the Brief Informant Form of Neurobehavioral Symptomatology (BINS). Results Participants’ (N = 97, Mage = 74.64 ± 7.42 years; Meducation = 15.76 ± 2.42 years; 74.2% female) FCI-SF performance significantly correlated with higher A-IADL-Q-SV (r = 0.403, p 0.01) and AFCI (r = −0.294, p 0.01) scores, whereas there was no significant correlation between a composite Z-score for traditional neuropsychological tests and select EF items from the BINS, (r = −0.119, p = 0.252). Conclusion We found greater alignment between objective and subjective measures of functional ability than EF, possibly due to the observable nature of functional tasks. Objective and informant measures of EF do not align, potentially targeting different aspects of cognition in non-demented older adults. This should be considered when selecting instruments measuring specific constructs of interest.
Bodek et al. (2024) studied this question.