Premium
The latent construct of dementia phenotype: Validation and longitudinal examination in the Sydney Memory and Ageing Study
Author(s) -
Lam Ben CP,
Crawford John D,
Lipnicki Darren M,
Numbers Katya T,
Kochan Nicole A,
Draper Brian,
Trollor Julian N,
Brodaty Henry,
Sachdev Perminder S
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.040876
Subject(s) - dementia , clinical dementia rating , structural equation modeling , psychology , neuropsychology , gerontology , cognition , longitudinal study , construct validity , medicine , psychiatry , clinical psychology , psychometrics , cognitive impairment , disease , pathology , statistics , mathematics
Background The latent continuous construct delta (δ) has been proposed as a novel approach to model dementia phenotype using structural equation modelling that reflects the “cognitive correlates of functional status” (Royall & Palmer, 2012. J Neuropsychiatry Clin Neurosci; Royall et al., 2012. J Alzheimers Dis). This δ factor has been demonstrated to be associated with clinically diagnosed dementia status and severity of dementia. However, thus far there are few studies validating the model longitudinally and these are in American samples. To establish the potential research and clinical utility of δ, the current research constructs and validates this latent dementia factor over a 6‐year period in a community sample of Australian older adults. Method A community‐dwelling sample of Australian older adults without dementia (at baseline) from the Sydney Memory and Ageing Study was used ( n = 1037; M age = 78.65 years; 55% females). Biennially, participants completed a battery of neuropsychological tests measuring performance in four major cognitive domains, and informants rated their functional status on instrumental activities of daily living. Dementia status and severity were established through consensus diagnosis by an expert panel of clinicians and the Clinical Dementia Rating Scale Sum of Boxes (CDR‐SOB), respectively. Result A latent growth curve model of δ and Spearman’s general intelligence factor (g) built on four waves of cognitive and function data revealed good fit: CFI = 0.97, RMSEA = 0.04, SRMR = 0.06. A significant increase in δ over time was observed, and this latent change in δ (Δδ) was significantly associated with CDR‐SOB at wave 4 after controlling for demographics, APOE*4 , and baseline CDR‐SOB. Cox regression revealed a significant association between Δδ and incident dementia. Further, Δδ accurately discriminated diagnosed dementia cases at wave 4 (ROC area under the curve = 0.91, 95% CI [0.88, 0.95]). Conclusion This study tests and validates the δ framework in Australian older adults by demonstrating that the change in δ over 6 years is associated with dementia risk and prospective severity of dementia. Future research should further test the model using longitudinal data from geographically and ethnoculturally diverse samples.