Evidence for potential bias in the Health and Activity Limitation Index as a health preference measure for persons with disabilities


Sarah E. Boslaugh, Ph.D., M.P.H.†, Elena M. Andresen, Ph.D.‡, Angela Recktenwald, M.P.H., Kathleen Gillespie, Ph.D.

†At the time of the research, Health Communication Research Laboratory, Saint Louis University School of Public Health.

‡At the time of the research, Epidemiology Division, Department of Community Health, Saint Louis University School of Public Health.

‡Conflicts of interest: The author and co-authors have no conflicts of interest relevant to this manuscript. Financial support for this study was provided, in part, by grants from the Centers for Disease Control and Prevention (ATPM/CDC grant R13/CCR717041-01) and a Saint Louis University 2000 Center of Excellence Award.

Disability and Health Journal,  January 2009 Volume 2, Issue 1, Pages 20–26
DOI: http://dx.doi.org/10.1016/j.dhjo.2008.07.004



The American public health plan Healthy People 2010 sets overall goals based on utility-linked information from the Health and Activity Limitation Index (HALex). However, little is known about how utilities measured by the HALex compare with those from established gold standard preference-based utility measures. In addition, distribution of HALex values from the general population underrepresents the experience of persons with disability, leading to uncertain interpretation of their utility values. Therefore, we sought to report the validity and interpretation of HALex scores compared to scores from a preference-based health-related quality of life measure, the Quality of Well-Being (QWB) scale.


A telephone survey with component measures was administered in random order. Participants consisted of 401 adults from a large Midwest metropolitan area: 302 were selected by random-digit-dial methodology and 99 were volunteers with mobility impairment extending the disability values of the sample. Multiple regression analysis predicted HALex scores from QWB scores and from demographic and self-reported health characteristics.


QWB scores accounted for 41% of the variance in HALex scores. The addition of five demographic and health factors increased the variance explained to 63%. Scores on the QWB and HALex were similar on mid-range values and discrepant at the extremes; that is, persons with extreme HALex scores tended to have more moderate QWB scores. HALex scores were higher for white adults than predicted by their QWB scores and lower for people with chronic diseases and disabilities.


Conclusions: Utilities as measured by the HALex and QWB can differ markedly, particularly if the person is classified at either end of the spectrum of function. Of similar concern is that fact that HALex scores show a systematic bias in relationship to QWB scores depending on a person’s demographic and health-related characteristics.