Psychometric properties of the activities-specific balance confidence scale and the survey of activities and fear of falling in older women

OBJECTIVES: To compare the psychometric properties of the Activities-specific Balance Confidence Scale (ABC) and the Survey of Activities and Fear of Falling in the Elderly (SAFE). DESIGN: Secondary analysis using baseline and 12-week data from a randomized, controlled trial on fall prevention. SETTING: Upper Midwest metropolitan area with assessments conducted in participants' homes. PARTICIPANTS: Population-based sample of 272 noninstitutionalized female Medicare beneficiaries aged 70 and older at risk of falling. MEASUREMENTS: Participants self-administered the ABC, SAFE, Geriatric Depression Scale, and Medical Outcomes Study 36-item Short Form Survey. During a home visit, a nurse practitioner administered the Berg Balance Test and Timed Up and Go, measured gait speed, and asked about falls and chronic illnesses. RESULTS: Baseline internal consistency measured using Cronbach alpha was 0.95 for the ABC and 0.82 for the SAFE. Baseline concurrent validity between the ABC and SAFE measured using a correlation coefficient was -0.65 (P<.001). ABC and SAFE scores were significantly correlated at baseline with physical performance tests and self-reported health status. The ABC had stronger baseline correlations than the SAFE with most measures. Neither instrument demonstrated responsiveness to change at 12 weeks. CONCLUSION: The ABC and SAFE demonstrated strong internal-consistency reliability and validity when self-administered. The ABC had stronger associations with physical functioning and may be more appropriate for studies focused on improving physical function. Both instruments demonstrated ceiling effects, which may explain the lack of responsiveness to change in relatively nonfrail older women. Instruments sensitive to measuring lower levels of fear of falling are needed to capture the full range of this phenomenon in this population.