The ABES, with the scoring direction modified, was phone-administered to 937 individuals who self-identified as having one or more arthritis conditions based on a validated, US, national survey assessment tool. Descriptive statistics of demographic variables and factor analysis of scale items were conducted. Scale dimensionality was assessed using principal component analysis (PCA) with oblique rotation. Criteria for assessing factors were eigenvalues > 1, visual assessment of scree plot, and structure and pattern matrices.

The predominantly female (74.2%) and Caucasian (79.9%) sample had a mean age of 61.0 ± 13.1 years, and a mean BMI of 30.2 ± 7.1. Major arthritis conditions reported were rheumatoid arthritis, osteoarthritis and fibromyalgia. A three-factor structure with cronbach alpha values of .84, .85 and .53 was elicited, and accounted for 72% of the variance.

Compared to the two-factor structure evidenced by the original ABES scale in a sample of UK adults, the data from this sample evidenced a three-factor structure with higher variance. The third factor’s cronbach alpha of .53 was low and could be improved by the addition of salient questions derived from further qualitative interviews with patients with arthritis and other rheumatic conditions and from current literature findings.