206717 Evidence of functional bias in the SF-36

Wednesday, November 11, 2009: 12:30 PM

Willi Horner-Johnson, PhD , Center on Community Accessibility, Oregon Health & Science University, Portland, OR
Gloria L. Krahn, PhD, MPH , National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, GA
Rie Suzuki, PhD , Department of Health Sciences and Administration, University of Michigan-Flint, Flint, MI
Jana J. Peterson, MPH, PhD , UMKC Institute for Human Development (UCEDD), University of Missouri - Kansas City, Kansas City, MO
Gale Roid, PhD , Southern Methodist University, Dallas, TX
Trevor Hall, PhD , RRTC: Health and Wellness, Oregon Health & Science University, Portland, OR
RRTC Expert Panel on Health Measurement , Oregon Health & Science University, Portland, OR

Measures of self-perceived health are important tools for assessing health, identifying disparities, and evaluating health promotion interventions. The SF-36 is the most widely used measure of self-perceived health in the United States. However, it includes items that appear biased toward an assumption that respondents are unhealthy if they are unable to perform certain functions such as walking or climbing stairs.


Healthy adults with and without disabilities were recruited and screened to ensure absence of significant chronic conditions or health problems. Screened participants completed a series of measures, including the SF-36. Differential item functioning analysis was used to identify items that performed differently for a particular disability group in comparison to the nondisabled group, even when overall scores were similar.


When controlling for each respondent's scores in physical or mental health domains of the SF-36, individuals with spinal cord injuries scored significantly worse than the reference group on eight out of ten items on the Physical Functioning subscale and two items on the Role-Physical subscale. People with sensory disabilities scored significantly worse on one Physical Functioning item and one Role-Emotional item. People with mental health disabilities scored significantly worse on Self-Rated Health and two Role-Emotional items.


Results indicate that the SF-36 penalizes healthy adults who have functional limitations. For disability and health researchers, it is critically important to disentangle confounding between health and function. Without this distinction, researchers cannot use health measures to accurately assess health disparities or effectiveness of health interventions intended to reduce disparities.

Learning Objectives:
1. Discuss the problems created by including measures of function within measures of health. 2. Explain how differential item functioning analysis can help identify bias in a measure. 3. Describe specific evidence of functional bias in the SF-36.

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I have training and experience in disability and health research.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.