We present the process of further validation of the 16-item Activities-specific Balance Confidence scale (ABC-16) and a short version (ABC-6) derived by us, to assess balance confidence and fear of falling (FOF). The ABC-16 was administrated to three groups who were anticipated to have a range of balance confidence: 70 patients with higher level gait disorders (HLGDs), 68 healthy controls, and 19 patients with Parkinson's disease (PD). Item reduction was based on identifying items with the lowest scores (high FOF) among the patients. Internal consistency and discriminative validity were assessed using Cronbach's alpha and logistic regression, respectively. The intraclass correlation (ICC) between the short and long versions was assessed using a mixed model approach, accounting for the difference between the scores of the two versions. Six items were found to reflect the most frightening conditions, especially in the patient groups, and to form the short version (ABC-6). Internal consistency of the ABC-16 and ABC-6 were high in the three groups: Cronbach's alpha was between 0.83 and 0.91 and 0.81 and 0.90, respectively. Compared to the control group, the sensitivity of the ABC-16 was 96% for identification of patients with HLGDs (greatest FOF) and 58% for identification of PDs (moderate FOF), based only on the ABC scores. Similar values were obtained for the short version, i.e., 91% for HLGDs and 53% for PDs. ICCs between the short and the long versions was 0.88 (HLGDs), 0.83 (PDs), and 0.78 (Controls). To conclude, the short version of the ABC has properties analogous to the parent questionnaire and is apparently useful in assessing FOF.
Copyright 2006 Movement Disorder Society.