Demographic Predictors of Older Adults’ Performance on a Novel Virtual Reality–Based Everyday Function Test and Conventional Measures of Cognition and Function

Objective: In contrast to conventional cognitive tests and questionnaires on everyday function, performance-based tests of everyday function typically show little to no relation to race, education, and other demographic variables. This study examined demographic correlates of a novel, virtual reality-based test of everyday function to determine whether its computerized format would yield demographic associations similar to performance-based tests with real objects. Methods: A total of 237 older adults (MAge = 72.7 ± 6.96; MEducation = 15.8 ± 2.75; 66% female; 47.7% White; 45.1% Black) completed conventional cognitive tests and two functional tests requiring them to make a breakfast and a lunch with real-life objects (Real Kitchen) and on a non-immersive virtual reality paradigm (Virtual Kitchen Challenge-Version 2; VKC-2). VKC-2 and Real Kitchen were scored for completion time and errors. Informant reports of everyday functioning were also analyzed. Results: Multiple linear regressions indicated that age significantly predicted nearly all variables. Race did not significantly predict any VKC-2 or Real Kitchen variables, though it significantly predicted nearly all cognitive test scores and informant questionnaires. Education showed similar results to race, but significantly predicted several VKC-2 and Real Kitchen scores. Sex significantly predicted two cognitive tests (Hopkins Verbal Learning Test- Delay Free Recall and Boston Naming Test). Conclusions: Performance-based measures of everyday function, even when administered in VR format, demonstrate weaker associations with race, sex, and education than traditional cognitive tests or informant reporting. The VKC-2 offers a valid and efficient approach to assessing cognitive and functional decline in diverse older adult populations.