Construct Validity of the ALLFTD Mobile App Cognitive Tests: Associations with Gold Standard Clinical Measures and Brain Volume

Abstract

AbstractBackgroundRemote digital cognitive assessments validated for use in frontotemporal dementia (FTD) and other neurodegenerative diseases would benefit clinical trials and clinical care. The ALLFTD Mobile App includes several smartphone‐based measures of cognition, and prior studies have provided preliminary support for their reliability. Here we evaluate their validity by testing associations with gold standard clinical outcomes and expected neural correlates.MethodA diagnostically mixed sample of 92 participants (CDR®+NACC‐FTLD = 0 [n = 33]; CDR®+NACC‐FTLD = 0.5 [n = 15]; CDR®+NACC‐FTLD≥1 [n = 16]; unknown [n = 28]) were recruited from a UCSF healthy aging study (Hillblom Network) and a North American FTD study (ALLFTD). Participants were asked to remotely complete four ALLFTD Mobile App tests of executive functioning (EF) and a memory test on their smartphones. Gold standard clinical measures were also collected; outcomes were the UDSv3 EF composite, CVLT‐SF Total Immediate Recall (memory), and CDR®+NACC‐FTLD Box Score (disease severity). Volumetric brain MRI measurements were available for a subset (n = 37). Expected neural correlates were frontoparietal gray matter volume for EF tests and hippocampal volume for the memory test. Linear models were fitted to evaluate the association of app tests with the gold standard outcomes, controlling for education, sex, and total intracranial volume (imaging only).ResultSmartphone EF measures were strongly and positively associated with the UDS3‐EF composite (ß’s = 0.6–0.8, all p<.001). The memory test was strongly correlated with the CVLT‐SF (ß = 0.7, p<.001). Tests were less strongly associated with divergent cognitive domains (e.g., associations with a visuospatial test were ß’s = 0.1–0.3). Worse performance on all tests was significantly associated with greater disease severity (ß’s = 0.5–0.7, all p<.001). Worse app‐based EF scores were associated with lower frontoparietal volume (ß’s = 0.4–0.6, all p<.015) and worse memory scores with lower hippocampal volume (ß = 0.5, p<.001).ConclusionALLFTD Mobile App scores were strongly associated with gold standard neuropsychological tests of the same cognitive domain (i.e., convergent validity) and less associated with measures of a different domain (i.e., divergent validity). Worse performance on the smartphone tests was associated with increasing disease severity and volume loss in expected brain regions. These results provide preliminary support for the construct validity of the ALLFTD Mobile App cognitive tests.