The Contribution of Health Literacy to Disparities in Self-Rated Health Status and Preventive Health Behaviors in Older Adults

Thank you for publishing this article that provides more evidence of the possible factors influencing levels of health literacy and health disparities among the U.S. elderly population. Your concluding remarks about needing interventions that address low health literacy encouraged me to highlight the work of the Medical Library Association (MLA) in this area. MLA is creating curricula to educate health care providers of the value of health information literacy. Soon, there will be a free self- guided Web-based tutorial on the MLA web site that health care providers can take to learn more about health literacy and the information resources and services available to them and their patients from the National Library of Medicine, as well as local health sciences librarians. Many medical librarians are willing to conduct training classes as well on the topic. The curricula and other useful related information will be available at http://mlanet.org/resources/healthlit/

Competing interests:   None declared

It's great that more research is being done to discover how literacy relates to health outcomes, especially with nationally representative data.

I'm also glad that causality has been addressed in these TRACK comments. The article left me wondering whether more could have been done to show support of health literacy as a true mediator of these disparities. In particular, I would have been interested to know whether educational status could be considered a mediator of the relationship between literacy and health outcomes- similar to reverse causality. I think this would have only involved measuring b’ (using the framework in figure 1) and comparing it to b, since the bivariate analysis for relationship a could be taken as bidirectional.

Educational status could mediate health literacy’s effects (or be caused by them) for the following reasons: First, it’s within the realm of possibility that individuals terminate their formal education (or have it terminated) as a result of a literacy problem that was already present. Second, the article seems to indicate that education and literacy are more closely linked than race (a more decidedly “independent” variable) and literacy.

All that said, I don’t purport to be a professional statistician when it comes to analyzing mediation. So really any information you have on what this study does offer about educational status not being a mediator of health literacy, would be helpful in addressing causality. Thank you again for this helpful article.

Competing interests:   None declared

I appreciate the comments of Dr. Weiss who brought the concept that literacy might influence health in the US in the early 1990's and continues to carry out important research in this field. Our primary goal in carrying out the analyses described in this paper, as pointed out by Dr. Weiss, was to extend and confirm the associations between health and literacy to a nationally representative sample using a more robust measure of literacy than has generally been used.

This being said, I urge caution in presuming that the relationship between literacy and health are directly causal as observational studies are not capable of fully supporting these presumptions. Clearly there is a need to assess whether interventions can reduce the literacy-linked disparities shown in health outcomes. Experimental designs will do much to move this important area of clinical work forward.

In addition, because the influence of literacy on health is likely to be lifelong, and potentially cummulative, life course studies may be needed to see significant changes in endpoints such as mortality.

Competing interests:   None declared