In search of ‘low health literacy’: Threshold vs. gradient effect of literacy on health status and mortality Michael S. Wolf * , Joseph Feinglass, Jason Thompson, David W. Baker Health Literacy and Learning Program, Division of General Internal Medicine, Feinberg School of Medicine, Northwestern University, United States article info Article history: Available online xxx Keywords: Health literacy Reading ability Measurement Health status Mortality Physical function Mental health USA Elderly abstract Studies have demonstrated significant associations between limited literacy and health outcomes. Yet differences in literacy measurement and the cutoffs used for analysis have made it difficult to fully understand the relationship between literacy and health across the entire spectrum of literacy (i.e., whether the relationship is continuous and graded or whether a threshold exists below which literacy is independently associated with health). To analyze this question, we re-examined the relationship between literacy, baseline physical functioning and mental health, and all-cause mortality for a cohort of 3260 US community-dwelling elderly who were interviewed in 1997 to determine demographics, socioeconomic status, chronic conditions, self-reported physical and mental health (SF-36 subscales), health behaviors, and literacy based upon the Short Test of Functional Health Literacy in Adults (S-TOFHLA). All-cause mortality was determined using data from the US National Death Index through 2003. Seven categories of S-TOFHLA literacy scores were created and used in this analysis instead of the existing three categories identified with the measure. In multivariate analyses, a continuous, graded relationship between literacy and baseline physical functioning was identified. However, participants scoring below the third literacy category had significantly worse mental health compared to the highest literacy category, displaying a notable threshold. Finally, all six literacy categories were significantly associated with greater all-cause mortality risk compared to the highest literacy category, but again there was a marked threshold below the third category at which the adjusted mortality rate significantly increased compared to all other categories. We conclude that the nature of the relationship between literacy and health may vary depending upon the outcome under examination. Ó 2010 Published by Elsevier Ltd. Introduction Health literacy, as defined by the Institute of Medicine and National Library of Medicine is ‘‘the degree to which individuals have the capacity to obtain, process, and understand basic health infor- mation and services needed to make appropriate health decisions’’ (Institute of Medicine, 2004). It is a multifaceted concept, of which reading ability is a fundamental component (Baker, 2006; Rudd, Kirsch, & Yamamoto, 2004). An individual’s ability to read, comprehend, and take action based on health-related material is closely related to the more general ability to read, comprehend, and take action based on non-health materials. Regardless of the terms and measures used, the field of health literacy is based on numerous studies that have found associations between measures of literacy skills and a wide variety of health outcomes (Baker, 2006; Dewalt, Berkman, Sheridan, Lohr, & Pignone, 2004). Studies have linked lower adult literacy with more infrequent use of preventive services, delays in diagnoses, poor understanding of one’s medical condition and treatment, non-adherence to medical instructions, inadequate self-management skills, worse physical and mental health, and increased mortality (Baker, Wolf, Feinglass, Gazmararian, & Thompson, 2007; Dewalt et al., 2004; Sudore et al., 2006; Wolf, Gazmararian, & Baker, 2005; Wolf, Knight et al., 2006). However, it remains unclear whether the relationship between literacy and health outcomes is continuous and graded, or whether there is a threshold below which an individual’s reading ability is so limited that he or she is unable to do the essential things needed to prevent disease and injury, seek prompt medical attention, and perform essential self-management tasks (e.g., take medications correctly). What is the requisite level of literacy skills necessary to adequately process, understand and act on health information? If such a threshold effect for inadequate literacy exists, screening tests could identify those patients who are at greatest risk, and it would be possible to develop more explicit guidelines and stan- dards for the design of appropriate health materials and instruc- tional strategies (Baker, 2006; Rudd, 2007; Weiss et al., 2005; Wolf, Davis et al., 2006). * Corresponding author. Tel.: þ1 312 503 5592. E-mail address: mswolf@northwestern.edu (M.S. Wolf). Contents lists available at ScienceDirect Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed ARTICLE IN PRESS 0277-9536/$ – see front matter Ó 2010 Published by Elsevier Ltd. doi:10.1016/j.socscimed.2009.12.013 Social Science & Medicine xxx (2010) 1–7 Please cite this article inpress as: Wolf, M. S., et al., In search of ‘low health literacy’: Threshold vs. gradient effect of literacy..., Social Science & Medicine (2010), doi:10.1016/j.socscimed.2009.12.013