Abstract
Direct observation and time-study techniques were used with a sample of 25 ambulatory, cognitively impaired subjects drawn from two long-term care settings to evaluate wandering behavior. The purposes of this study were (a) to describe the 24-hour distribution of wandering and direct ambulating cycles, (b) to examine the stability of wandering behavior over a 3-day interval, (c) to evaluate whether wandering during a 2-hour epoch is representative of that of a 24-hour day, and (d) to evaluate whether large-scale integrated (LSI) activity meters can substitute as an index or proxy for direct observation in the study of wandering behavior. Subjects displayed a daily average of 20.1 cycles encompassing 43.9 minutes of wandering ambulation and 28.8cycles encompassing 40.4 minutes of direct ambulation. Wandering behavior was present in all subjects. However, wandering was highly variable from subject to subject. For a given subject, wandering was only moderately stable over a 3-day interval, but more so than direct ambulation. Similarly, a standard 2-hour epoch was moderately representative of daily wandering ambulation, but more so than for direct ambulation. Finally, LSI meters, when applied at the ankle and worn over longer (24-hr) rather than shorter (2-hr) intervals, are a promising means to index wandering behavior.