Evaluating Walking Intensity with Hip-Worn Accelerometers in Elders

Abstract

Purpose: Accelerometer activity counts are used to objectively categorize physical activity levels. We examined the association between accelerometer activity counts and metabolic effort in older adults.

Methods: Forty-five older adults (76.3 ± 5.1 yr) completed a 400-m walk at both a usual and a rapid pace. A portable metabolic unit measured pulmonary gas exchange, whereas a hip-worn accelerometer measured activity counts. Participants were categorized as either a "slow walker" or a "fast walker" based on the threshold of 1.0 m·s during usual-pace walking.

Results: Activity counts during rapid (r = 0.62, P < 0.01) but not usual-pace (r = 0.24, P = 0.11) walking were significantly associated with METs. Slow walkers attained only half the activity counts of fast walkers during each walk condition (P < 0.01), while at the same time achieving between 82% and 90% of their MET level.

Conclusion: Accelerometers may misclassify the activity level of functionally impaired older adults with slow walking speed.

FIGURE 1

Scatter plot and lines of best fit predicting metabolic rate with vertical axis activity counts recorded with a hip-worn accelerometer.

FIGURE 2

Oxygen consumption and average activity counts•min^-1 during 400-m walk at A. usual and B. rapid pace. The data are categorized by individuals who walked greater than or less than 1.0 m•s^-1 at a usual pace – a pace that designates older individuals as “slow walkers” with potential functional impairments. Dotted lines represent cut points for metabolic rate (3.0 METs) or accelerometer counts (2020 counts•min^-1 and 1041 counts•min^-1) that designate the walking intensity equal to or greater than moderate. Significant differences occurred in accelerometer counts between slow and fast walkers (P < 0.05).