The Relationships among race/ethnicity, physical activity, education and cognitive function in late middle age

OBJECTIVES: This study describes cross-sectional and longitudinal racial/ethnic differences in cognitive function scores (memory and mental status), and the mediating roles of education and physical activity in late middle aged adults. The purpose of the study is to explain racial/ethnic disparities in cognitive function scores. METHODS: The Health and Retirement Study collected data on men and women aged 51-61 in the United States from 1992-2002 (n = 9204) of the following racial/ethnic groups: black, Hispanic, and white. Multivariable and mixed modeling techniques were used to describe racial/ethnic disparities on tests of memory and mental status (orientation, attention, naming), and to determine if physical activity or education mediated the relationship between race/ethnicity and cognitive function scores. Models were adjusted for demographic, social, and health-related covariates. RESULTS: Black and Hispanic adults had lower memory and mental status scores than whites (p<0.0001). However, in a subsample of those working for pay, Hispanic adults had no differences in mental status scores from their white counterparts. Leisure-time physical activity and education mediated some relationships between race/ethnicity and cognitive function scores, and were associated with cognitive function scores at each wave. Nevertheless, physical activity, education, and race/ethnicity were not associated with differential rates of change in cognitive function scores. CONCLUSIONS: Racial/ethnic differences in cognitive function scores exist in late middle age. Leisure-time physical activity and education partially mediate the relationships; however the covariates used in the analyses were unable to explain all of the differences. Although being white or Hispanic, increased education, increased vigorous activity, and less work-related physical activity were associated with higher memory scores, none differentially impacted rates of change in memory score. Being white and increased education were positively related to mental status scores at each wave. However, no variables of interest were associated with differential rates of mental status score change.