Harris, TB; Caserotti, P; Chen, K; Glynn, N; Brychta, R; Matthews, C; Maia, BL; Van Domelen, D; Shen, HM; Hung, M; Koster, A
3rd International Conference on Ambulatory Monitoring of Physical Activity and Movement
INTRODUCTION: Wrist-worn accelerometer use is likely to increase in clinically important groups such as older populations. However, there is a dearth of information regarding wrist accelerometry in older persons and a lack of easy-to-apply analytic algorithms. PURPOSE: To evaluate: 1. Can data from wrist, hip, arm, or thigh accelerometers equally reflect tasks of daily living? 2. Which accelerometry position best allows assessment of physical function? And, 3. Are accelerometry measures independent contributors to physical function in older persons? METHODS: We added an accelerometry component to a methods study at the University of Pittsburgh. Subjects wore multiple accelerometers: Actigraph GT3X+ on right and left wrist and right hip (80 Hz), a Sensewear on the left arm, and for 50 subjects, an ActivPal on the right thigh. While wearing a Cosmed oxygen monitor, subjects performed five sets of activities: (1) laying still, standing still; (2) upper body movement while standing: simulated dishwashing, dough kneading, dressing, folding towels; (3) upper body movement while walking: vacuuming, shopping; (4) upper body movement while sitting: writing, dealing cards, chair stands; (5) lower body movement: usual and fast paced walks for 20 meters performed with and without using arms. Subjects were asked to provide 7 days of free-living data while wearing all monitors. Other functional measures were collected at the same session. RESULTS: We have complete data on 71 older men and women (50% male), with a mean age of 79.5. Average body mass index (BMI, weight in kilograms over height squared) is 26.6. Average walking speed at usual pace (UWS) is 1.13 (±.15). After accounting for nonwear time, we examined correlations of selected variables from the free-living data with UWS, controlling for total valid minutes, age, gender and BMI. ActivPal step count (.31, P=.04), ActivPal sedentary time (-.32), P=.04), ActivPal Upright Time (.31, P=.05) all correlated with usual walking speed; ActivPal step time was similar (.30, P=.05) but neither of the position change variables were associated. Evaluation of the hip and wrist worn Actigraph GT3X+ with UWS and other measures of function is ongoing. CONCLUSION: We have created a data resource to understand the contribution of accelerometry to assessment of older populations. We seek collaborations with statistical groups (see abstract by B. He) to help us compare the accelerometry output from wrists, arm, thigh, and hip during specific activities and in free-living data to identify whether these measures add independently to estimation of daily activity in older people. SUPPORT: Intramural Research Program, NIAHarris, TB; Caserotti, P; Chen, K; Glynn, N; Brychta, R; Matthews, C; Maia, BL; Van Domelen, D; Shen, HM; Hung, M; Koster, A Aging Research Evaluating Accelerometry (area): Methodologic Comparison Of Accelerometry Performance In The Very Old 3rd International Conference on Ambulatory Monitoring of Physical Activity and Movement 2013. http://www.umass.edu/sphhs/sites/default/files/ICAMPAM_Slide_Poster_Composite_6.6.pdf.