Previously in aging in action, we discussed a new in-home sensor monitoring system designed to detect patterns in the physical activity of community-dwelling older adults. A forthcoming article presents some initial findings based on 76 users of the system, whose in-home activity was recorded over a four-week period.
The monitoring system used in the study collects data via passive infrared sensors, which detect motion and relay information to a computer. The sensors are set up in a line on the ceiling of a user’s home, and are used to detect when the user is walking under them as well as the user’s walking speed.
Such continuous monitoring of gait speed is useful for several reasons. Gait speed is a useful health marker in research and clinical assessment, and is associated with several other important factors such as executive cognitive function. Clinical and laboratory assessment of gait speed may not be indicative of the individual’s real-world walking abilities, and can only be performed infrequently. Continuous, in-home monitoring enables researchers and clinicians to detect sudden changes in individual activity patterns, and places minimal demand on the walker. The use of motion detection sensors allows for significantly more privacy than video monitoring or other active monitoring alternatives.
For this study, participants were given a clinical assessment that included cognitive testing, health status testing, standard clinical gait measurements, and a Functional Activities Questionnaire (an assessment of daily activities.) The in-home sensor monitoring took place for two weeks before and after the clinical assessment. The in-home walking data was then analyzed and compared to the clinical assessment data.
Over a one-month period, participants averaged 22 walks per day, although there was a great deal of individual variation. The mean in-home walking speed was 61 centimeters per second, or a little under two miles per hour. There were no significant differences by gender on walking speed or frequency, although as only 11 of the participants were male there were not enough participants to make it likely that any such difference would be detected. Age, which ranged from 72 to 97, was not associated with walking speed. Twenty-six participants walked with assistance of a cane or walker, and as a result, walked less frequently and at lower speeds than the other participants. Fifty-three percent of total walking occurred in the morning or early afternoon, and walking was faster at these times compared to later in the day.
The researchers’ hypothesis that continuous, in-home monitoring would record lower average walking speed than clinical assessment was supported by their findings. However, these measurements were significantly associated with clinical measures of walking speed and other measures of motor function—in other words, all participants tended to walk slower at home than they would have during clinical assessments, but those who were faster or slower than average at home were similarly faster or slower than average in clinical assessments. As in clinical and laboratory studies, in-home walking speed was also positively associated with the overall cognitive assessment scores, and in the specific cognitive functions of processing speed and visuospatial ability. This study suggests that continuous, in-home monitoring is a valid and valuable measure of walking function.