Low literacy is highly prevalent in the United States. According to the U.S. Department of Education's National Adult Literacy Survey (NALS) of 1992, 21 percent of adults possess skills in the lowest level of prose, document, and quantitative literacy. Another 27 percent of adults fall under the second lowest level of literacy proficiency and may be hindered in routine daily activities. The prevalence of low literacy increases dramatically with age: 44 percent of U.S. adults age 65 and older had level 1 reading skills on the NALS. Individuals with inadequate literacy have worse health in cross sectional studies and higher hospitalization rates in prospective studies. No previous study has examined whether literacy independently predicts death. In addition, the mechanisms by which literacy affects health outcomes remain unclear. The primary aims of this study are: 1) To determine whether elderly persons with inadequate or marginal functional health literacy have higher mortality than persons with adequate literacy after adjusting for age, gender, race/ethnicity, and ? baseline health and chronic conditions, and 2) To determine the importance of different causal pathways between literacy and mortality and the residual effect of literacy on mortality after adjusting for all pathways examined, including a) adverse health behaviors (smoking, alcohol use, obesity, and seat belt use), b) under use of preventive services (mammography, Papanicolau smear, and pneumococcal and influenza vaccination), c) occupation and income, and d) lower levels of social support. Data from a cohort of 3260 community dwelling elderly who completed a one hour in-home interview that measured health behaviors, health care use, physical and mental health, functional health literacy (Short Test of Functional Health Literacy in Adults) and cognitive functioning (Mini Mental Status Examination) will be linked to the National Death Index to track mortality. Cox proportional hazards models will be used to determine the independent association between literacy and mortality. The pathways described above will be sequentially added to the models to determine the change in the beta coefficients for literacy to determine the degree to which each pathway mediates the relationship between literacy and mortality. ? ? ?
