A - 127 Semantic Fluency Performance as a Function of Race and Clinical Status

Abstract Objective Determine whether performance on animal naming (AN) differs between African American (AA) and Caucasian patients with no cognitive disorder (NCD) and suspected Alzheimer’s Disease (ad). Method AN raw scores were obtained from Caucasian (N = 409) and AA (N = 83) patients evaluated at a large Southern medical center. Two-way factorial analysis of covariance was conducted focusing on the interaction among and differences between racial (AA, Caucasian) and clinical (ad, NCD) groups on AN. Covariates included age, education, and premorbid reading level (PRL) measured via Test of Premorbid Functioning (TOPF) raw score. Results Winsorization was used for outliers. The ad group (9.97[5.53]) produced significantly fewer words on average than the NCD group (18.99[5.15]), F(1, 485) = 46.01, p < 0.001. There was no significant main effect of race after controlling for age, education, and PRL, F(1, 485) = 0.42, p = 0.517. However, there was a significant interaction between race and clinical group, with the difference between clinical groups greater among Caucasian patients, F(1, 485) = 13.89, p < 0.001. Specifically, among NCD patients, Caucasians averaged higher scores (19.41[5.15]) than AAs (16.61[4.5]), despite controlling for age, education, and PRL, T(1,353) = 3.72, p < 0.001. Among patients with ad, AAs (10.9[7.01]) and Caucasians (9.72[5.06]) scored similarly (p > 0.05). Conclusions Results suggest differences in AN between ad and NCD are less pronounced for AA patients due to lower NCD AN scores among AAs. Evaluation of socioeconomic conditions (e.g., Area Deprivation Index), geographic linguistic patterns, and item-level error analysis could provide insight into differential AN performance.