Journal Logo

A-29 Free Communication/Poster - Physical Activity Assessment: MAY 28, 2008 7: 30 AM - 12: 30 PM: ROOM: Hall B

Validity of Downloadable Pedometers during Laboratory and Overground Walking in Young and Older Adults


Board #114 May 28 11:00 AM - 12:30 PM

Miller, Nora E.; Strath, Scott J.; Swartz, Ann M.; Loy, Michael S.; Ewalt, Lauren A.

Author Information
Medicine & Science in Sports & Exercise: May 2008 - Volume 40 - Issue 5 - p S197
doi: 10.1249/
  • Free

PURPOSE: To examine the validity of downloadable pedometers to accurately count steps during treadmill and overground walking in young and older adults.

METHODS: Eighty-four participants (n=45 aged 20-49 yrs; n=39 aged 50-80 yrs) completed a treadmill (TM) protocol (2.0-4.0 mph, in 0.5 mph increments) and an overground (OG) walking protocol (< normal, normal, and > normal walking speeds) while wearing three waist-mounted pedometers (Sportbrain iStep X1 (SB), Omron HJ-700IT-T54 (OM), and Kenz Lifecorder EX (LC)). Actual steps were manually tallied by a researcher. Paired samples t-tests were performed to compare total steps from each pedometer with manually assessed steps for each stage. Mean error scores and 95% CI were calculated.

RESULTS: There were no significant differences between either SB or OM and the manual tallied steps for any of the TM stages. Mean error scores (95% CI) for SB ranged from −28.6 to −1.8 (−69.5, 47.9) for the younger group and −50.1 to −26.0 (−100.8, 25.4) for the older group across all TM speeds. Error scores for OM ranged from −9.5 to 8.4 (−19.3, 21.5) for the young group and −14.2 to 0.40 (−33.4, 19.0) for the older group across all TM stages. Alternately, the LC was significantly different from the manual tally at 2.0 mph for the young group, mean error −27.1 (−45.0, −9.1), and older group, mean error of −72.1 (−95.2, −49.0) (p<0.01 for both groups). There were also significant differences between the LC and manual tallies for the older group at 2.5 mph, mean error of −18.2 (−29.3, −7.1) and 3.0 mph, mean error of −4.3 (−7.5, −1.1) (p<0.05 for both speeds). Error scores for LC ranged from −2.8 to 1.6 (−6.8, 6.6) for the young group from 2.5 to 4.0 mph and −1.0 to −0.13 (−4.6, 2.5) for the older group at 3.5 and 4.0 mph. OG walking showed a significant difference between the LC and tallied steps (< normal walking speed) for the younger adults, mean error −13.1 (−26.0, −0.25) (p=0.046).

CONCLUSIONS: Downloadable pedometer accuracy increased with increasing walking speeds, and findings were irrespective of age. Although the SB and OM pedometer appear to accurately count steps at slower walking speeds, caution should be exercised with these interpretations as these two devices had very large 95% CI of error for all speeds during both TM and OG walking. The pedometer with the smallest error range for both conditions was the LC.

©2008The American College of Sports Medicine