INS NYC 2024 Program

Cognitive dispersion refers to the within-person variability across tasks. Relative to the normative approach, cognitive dispersion may be more sensitive to subtle changes in preclinical Alzheimer’s disease (AD). Greater dispersion has been associated with increased risk for amyloid-beta PET positivity, conversion to mild cognitive impairment and faster rates of medial temporal lobe atrophy. This metric can be derived from a regular neuropsychological assessment and complements traditional outcome approaches, particularly in those under-represented in normative data. This study evaluated cognitive dispersion as an independent predictor of functional status across racial and ethnic groups and explore its association with AD biomarkers.

Baseline data from the U.S. POINTER clinical trial on 2111 older adults without significant cognitive impairment was used (Mean Age=68.2±5.2 years; 69% female, 30% with less-than-college education, 31% from underrepresented group). Cognitive dispersion was defined as the standard deviation of individuals’ z-transformed scores of 8 tests from a Neuropsychological Test Battery modified for U.S. POINTER. Clinical status was estimated by Clinical Dementia Global Ratings (CDR 0.5 vs. 0). Other covariates included subjective cognitive decline (SCD; Cognitive Function Instrument and Measurement of Everyday Cognition 12 item), cardiovascular risk (Framingham Heart Study 10-year risk score), level of physical activities, and global cognitive composite z-score. Correlations between cognitive dispersion and other brain health risk covariates were tested. Logistic regression models were fitted to predict CDR to evaluate cognitive dispersion as an independent predictor in the whole sample and in White participants and Black participants, controlled for age, gender, education, cardiovascular risk, physical activity, SCD, and global cognition. In a subset of 795 individuals with neuroimaging data, the relationship between cognitive dispersion, amyloid status (florbetaben PET; cortical summary SUVR>1.08 defines amyloid positive), and entorhinal cortex (ERC) tau (flortaucipir PET) deposition were examined, followed by testing for an amyloid x tau interaction effect.

Greater dispersion was correlated with older age (r=0.08, p<0.001) and greater vascular risk (r=0.11, p<0.001), but not SCD or physical activity. Cognitive dispersion was significantly higher in Black participants, those with less education, and those with higher vascular risk. Cognitive dispersion was correlated with CDR in Black participants (OR=4.4, p=0.01), but this association was not significant in the whole sample (OR=1.4, p=0.20) and in White participants (OR=0.93, p=0.81). In the subset of participants with imaging data, cognitive dispersion did not differ between positive and negative amyloid status groups. An amyloid status x tau interaction was found (t=2.11, p=0.04), such that cognitive dispersion correlated with ERC tau (r=0.14, p=0.04) only among those who were amyloid positive, even when adjusted for age.

Results supported cognitive dispersion as a sensitive objective metric, in addition to norm-based global cognitive performance and self-report SCD, to identify individuals showing the earliest clinical signs of decline, particularly among those under-represented in traditional norms. Cognitive dispersion may also track with medial temporal lobe tau deposition in amyloid positive individuals. Further analysis is needed to understand cognitive dispersion in the context of AD pathology and possible cultural bias in some neuropsychological tests and to assess trajectories of cognitive dispersion over time.