Individual scaling of accelerometry to preferred walking speed in the assessment of physical activity in older adults

Background Walking forms a large portion of physical activity (PA) of older adults. We assessed free-living PA using acceleration corresponding to preferred walking speed as a relative cut-point, and studied how it relates to age. We compared the relative cut-point to a common absolute cut-point of moderate-to-vigorous PA (MVPA). Methods 444 community-dwelling adults aged 75, 80 and 85 wore an accelerometer on the thigh during a PA surveillance period and a modified six-minute walking test (6MWT) at preferred speed. Each individual’s mean acceleration (g) during the 6MWT was used as a cut-point for relative PA. Acceleration corresponding to three metabolic equivalents (METs) was used as the cut-point for absolute MVPA. Results When using the acceleration of preferred walking speed as a cut-point, 62 (SD 82) minutes a week of relative PA was detected, compared to 228 (163) minutes of absolute MVPA. For 96 % of the participants, the acceleration generated by their preferred walking speed exceeded the common absolute cut-point for MVPA. Absolute MVPA was lower in the older age groups, and 6MWT speed explained 22 % of its variation (p<0.001), whereas relative PA was independent of walking speed and age. Conclusions Preferred walking speed was a significant contributor to absolute MVPA, and those who walked the slowest accumulated the least MVPA. Assessing relative PA using the intensity of preferred walking speed as a cut-point eliminated the dependency of PA on age and walking speed, and may be a feasible scaling option to evaluate relative PA among older people.