To establish how participant and informant reports of cognitive decline may differ between groups or remain consistent based on race and level of education in a large, national sample.

Participants were selected using the National Alzheimer’s Coordinating Center (NACC) database. Participants who were cognitively healthy at baseline and at least 65 years of age were selected. All informants either lived with the participant or visited the participant weekly (N = 9300). Participant racial groups included White American (n = 7534), Black American (n = 1453), Native American/Alaskan (n = 68), or Asian American (n = 239). Native Hawaiians were not included in this study, given the small sample size (n = 6). Participant education-levels included less than high school degree (n = 395), high school degree or GED (n = 1326), some post-secondary education (n = 1727), bachelor’s degree (n = 2184), and graduate studies (n = 3668). Pairwise comparisons examined each racial and educational attainment group by subjective reports of cognitive decline using Bayesian contingency tables to find reliable evidence to support the null or alternative hypothesis. Participant and informant reports of decline were coded to create a single variable to express no reported decline, participant reported decline, informant reported decline, or agreed decline.

Pairwise race comparisons found moderate evidence that Native Americans reliably reported cognitive decline differently than Black (BF10 = 6.973) and White Americans (BF10 = 3.634). In both cases, the Native American group reported more cases of decline than expected in all groups and reported no decline less than expected. Further analysis found very strong evidence for the null hypothesis when comparing White Americans with Black (BF01 = 60.506) and Asian Americans (BF01 = 65.72). A comparison of Black and Asian Americans found extremely strong evidence for the null hypothesis (BF01 = 199.464). No conclusive evidence was found when comparing reports of Native and Asian Americans (BF01 = 2.401). Pairwise comparisons of educational attainment with subjective reporting of cognitive decline found no evidence of reliable differences between groups. No conclusive evidence was found when comparing the reporting pattern of individuals with some post-secondary education and individuals who did not complete high school (BF01 = 1.257). Moderate evidence for the null hypothesis was found when comparing individuals with a bachelor’s degree with those who did not complete high school (BF01 = 8.57). Strong evidence for the null hypothesis was found when comparing individuals who did not complete high school with those who have studied at the graduate level (BF01 = 17.141) and those who completed high school (BF01 = 16.306). When making all other pairwise comparisons (BF01 > 100), there was extremely strong evidence for the null hypothesis.

These findings suggest that how participants and their informants report cognitive decline does not differ based upon educational attainment in almost all cases, and no evidence was found supporting differences based upon educational attainment. There is evidence that Native Americans/Native Alaskans and their informants report more cognitive decline compared to White and Black Americans. However, the findings suggest that White, Black, and Asian Americans do not differ in how participants and their informants report cognitive decline.