Psychometric properties of the Berg Balance Scale in a community-dwelling elderly resident population in Taiwan

Background/Purpose: To investigate the psychometric properties (acceptability, internal consistency reliability, interrater reliability, construct validity) and identify the most challenging items of the Berg Balance Scale (BBS) for elderly people living in the community. Methods: A total of 268 community-dwelling adults 65 years of age or older volunteered to participate in this study. Each subject's performance was assessed with the BBS, timed up and go (TUG) test, and usual gait speed. For testing interrater reliability, the other 68 community-dwelling older adults who met the criteria were also recruited. Results: The BBS demonstrated good internal consistency reliability (Cronbach's alpha=0.77), good interrater reliability (ICC2,1=0.87), and moderate correlation with the TUG and usual gait speed (Spearman's rho=-0.53 and 0.46, respectively). The BBS score of the mobility/IADL (instrumented activities of daily living) able group was also significantly higher than that of the disabled group. Among all items on the BBS, tandem stance (item 13) and one-legged stance (item 14) were found to be the most challenging items for the subjects in the sample. Conclusion: The results of this study suggest that the internal consistency reliability, interrater reliability, and construct validity of the BBS are adequate for measuring balance in community-dwelling older adults. Among all items in the BBS, the tandem stance and one-legged stance are the most challenging items. Further study of their applicability for screening use in the community is warranted.