Alzheimer's Disease Education and Referral Center

Gap in reading levels may explain apparent differences in cognitive testing

December 15, 2001


The research of Jennifer Manly, PhD, and colleagues at Columbia University suggests that methodological and analytical approaches, not actual racial or ethnic disparities in cognition, may in some cases help explain differences between older African Americans and non-Hispanic whites on cognitive test performance.

The Columbia group's findings, published in the March 2002 Journal of the International Neuropsychological Society, offer a powerful new tool - adjusting test results for quality of education - as one way to make cognitive testing of older people more culturally neutral. In the study, when test scores of participants with similar quality of education, measured by reading level, were compared, differences in test scores between two racial groups mostly disappeared.

The report challenges a standard approach used in research on cognition and dementia. Because differences in education might affect test scores or health status, most studies that have adjusted for the number of years of schooling still found that older African Americans did not score as well on cognitive tests as non-Hispanic whites. Manly's work looks instead at the quality of education. Inequalities in the educational experience of those who grew up in an era of segregated schools, Manly reasons, might be attributed to less funding and resources for such schools as well as fewer days per year that schools in the agrarian South were open for instruction. These discrepancies in early education, Manly suggests, have the potential to persist throughout life, as indicated in differences in reading levels between older African Americans and non-Hispanic whites.

At the NIA, racial and ethnic differences in cognitive function and dementia are an increased focus of research. "Several studies have reported racial and ethnic differences in cognitive performance in older adults and in the prevalence of Alzheimer's disease," notes Molly Wagster, PhD, program director for neuropsychology of aging research and project officer for the Manly study. "Because cultural, educational, or other differences might affect performance on this type of testing, there have been questions about whether measurement techniques are as culturally neutral as they could be."

Participants were drawn from a community-based study of cognitive aging in New York City's ethnically diverse neighborhoods of northern Manhattan. Random sampling was used to select 192 African Americans and 192 non-Hispanic whites age 65 and older. The two groups were matched in age, sex, and number of years of education. Participants were given medical evaluations at the start of the study to confirm that they did not have dementia and that they had no problems with daily activities. A battery of neuropsychological tests was then administered to examine memory, learning, and thinking abilities. Older people with low levels of reading ability could only name letters or read short, common words. Those with higher reading levels were able to pronounce complex, less common words out of context.

The African American group scored lower on the neuropsychological test battery overall as compared to non-Hispanic whites. But when reading scores were factored in, most differences between the groups were no longer statistically significant. The exceptions were category fluency (a test in which people must name objects in a category during a set period of time) and a drawing test, where differences between racial groups were observed even after adjusting for reading level. More information is needed to explain the persistent differences on those two measures, the researchers said.

Manly recommended further research to sort out methodological and analytical approaches, such as accounting for quality of education, from real racial or ethnic disparities in cognitive test performance. Such problems in methodology and analysis make it difficult to establish whether one group or another truly may have greater rates of cognitive decline and dementia in minority groups. "We need to develop measures designed to assess cognitive abilities in a way that is compatible with African American and other cultures," she notes. "With these, we will have greater confidence in any measures which show - or do not show - differences in cognitive health and dementia among various groups. Dr. Manly pursues such research as a neuropsychologist at the NIA-supported Alzheimer's Disease Center at Columbia University. Part of Dr. Manly's focus at the Center concentrates on research to improve the accuracy in diagnosing cognitive impairment and decline over time.

Researchers at Columbia are now adding tests of reading levels to their assessments of cognitive ability and are looking to see how accounting for quality of education might affect previous findings on levels of cognitive performance and rates of dementia among different groups in the community. In addition, Manly and her colleagues are carrying out further research on the effect of quality of education and literacy among Spanish-speaking Caribbean Hispanics through a pilot grant from the Columbia Center for the Active Life of Minority Elders (CALME), a Resource Center for Minority Aging Research (RCMAR) funded by the NIA and the National Institute of Nursing Research.

Share this:
Email Twitter Linkedin Facebook