AbstractBackgroundWith a growing racial and ethnic disparity in the aging population, it is important to understand factors that could influence relationships between neurodegeneration and cognitive decline. Differences in early socioeconomic exposure, educational background, and/or lack of culturally and linguistically appropriate testing protocols could impact the ability to see or interpret such relationships. We examined this issue by comparing the association between MRI‐derived measures of cortical thickness and performance on tests of global cognitive function, memory and language in elderly Latino and Non‐Latino White (NLW) individuals before and after adjustment for education.MethodsParticipants were 36 Latino and 126 NLW elderly individuals with cognitive impairment (i.e., Mild Cognitive Impairment, Alzheimer Dementia), and 33 Latino and 95 NLW cognitively normal controls (NC). Participants were evaluated annually at the Shiley‐Marcos Alzheimer's Research Center (ADRC) where they received cognitive assessment that included measures of global cognitive function (i.e., MMSE), memory (Logical Memory Test, Word List Learning) and language (Category Fluency Test). MRI scans were performed within 6 months of testing and measures of hippocampal volume and cortical thickness were determined using Freesurfer procedures.ResultsDemographically, the Latino NC group was younger (71.0+7.9) than the Latino‐Impaired group (73.2+10.8) and both NLW groups (NLW NC=72.8+5.7; NLW‐Impaired: 74.4+7.7; p < 0.01). Latino‐Impaired (9.9+4.8) had fewer years of education than Latino NC (13.8+3.2), and both Latino groups fewer than the NLW groups (NC=16.5+2.2; Impaired: 16.2+2.9). Regression analyses showed that MMSE, Word List Learning, Logical Memory, and Category Fluency were associated with entorhinal cortical thickness or hippocampal volume in the Latino and NLW groups. Furthermore, these associations were stronger in the Latino‐Impaired group than in the NLW Impaired group (all p’s < 0.05). However, these latter differences no longer reached significance when adjusted for education (i.e., in regression analyses or when the overall Latino group was restricted to the education range of the NLW group.ConclusionsThe large education gap between elderly Latinos and NLWs can influence perceived differences in the association between measures of neurodegeneration and cognitive decline. Further investigation of cultural and ethnic influences on the manifestation of dementia and its underlying causes is needed.