Photo Credit: cifotart

The following is a summary of “Evaluating Social Determinants of Health–Based Alternatives to Race-Based Cognitive Normative Models,” published in the November 2024 issue of Neurology by Schneider et al.

Race and ethnicity are proxies for sociocultural factors influencing cognitive test performance, which may vary across different groups.  

Researchers conducted a retrospective study to compare regression-based cognitive normative models adjusting for demographics and social determinants of health (SDoH).  

They analyzed data from the Atherosclerosis Risk in Communities Study (ARIC) for Black and White participants without mild cognitive impairment (MCI)/dementia who attended visit 5 (2011–2013). Participants completed 11 cognitive tests across 3 domains, memory, executive function, and language, 6 separate normative models were fitted for each cognitive test, adjusting for age and education, with different combinations of race, the Wide Range Achievement Test (WRAT; education quality proxy), and the Area Deprivation Index (ADI; neighborhood deprivation). Model fits, and concordance/discordance was compared using z-scores from each model, with a z-score threshold of <−1.5 indicating impairment.  

The results showed 2,392 participants with a mean age of 74.4 years, 60.4% were female, and 17.1% were of self-reported Black race. The “Full” model, which included race and SDoH factors, consistently outperformed other nested submodels (likelihood ratios ≥ 100) for all cognitive domains and tests, except for Delayed Word Recall. Models using education quality alone (WRAT) generally outperformed those incorporating neighborhood deprivation (ADI) or race (Race) alone for memory and language tests. In contrast, race-based models performed better for executive function tests. Adding neighborhood deprivation to education quality (WRAT + ADI) did not improve model performance compared to using WRAT alone. Across all domains and tests, concordance with the “Full” model was lower for education and ADI models than for other nested models. Discordance was greater in Black participants (range = 8.2%–23.2%) compared with White participants (range = 2.2%–3.4%), particularly for the Boston Naming Test and executive function tests.  

They concluded that education quality outperforms neighborhood deprivation in cognitive normative models, suggesting that education-related measures were valuable in cognitive aging research.  

Source: neurology.org/doi/10.1212/WNL.0000000000210030