RESPONSE

Dear Editor-in-Chief: We thank Jehn et al. (3) for their interest in our article. The primary objective of our study (2) was to examine the validity of step counts measured with the Omron HJ-112 pedometer and to assess the effect of pedometer placement during treadmill walking in healthy, normal/overweight, and obese adults. We examined pedometer accuracy during a variable-speed condition and at slow (1.12 m·s−1 or 2.5 mph), moderate (1.34 m·s−1 or 3.0 mph), and fast (1.56 m·s−1 or 3.5 mph) speeds. The primary findings of this study were that the Omron HJ-112 pedometer validly assesses steps in both normal/overweight and obese individuals during constant and variable-speed walking when positioned around the neck, in a shirt pocket, or at the hip. Although there are limited data regarding preferred walking speeds for the general population, recently published data from Kang and Dingwell (5) and Finnis and Walton (1) suggest that the speeds used in our study were appropriate for our healthy adult population. In healthy young adults (n = 17, age = 23.3 ± 2.6 yr, body mass index [BMI] = 23.5 ± 1.7 kg·m−2), Kang and Dingwell (5) reported a preferred walking speed of 1.30 ± 0.10 m·s−1 (range = 1.16-1.56 m·s−1). In healthy older adults (n = 18, age = 72.1 ± 6.0 yr, BMI = 25.4 ± 3.2 kg·m−2), the preferred walking speed was 1.29 ± 0.15 m·s−1 (range = 0.93-1.52 m·s−1). Furthermore, researchers who have measured pedestrian walking speeds for the purpose of planning pedestrian walking facilities have reported an average walking speed of 1.33 m·s−1 for the general population (1). Thus, a walking speed of 1.12 m·s−1 may exceed that typically considered as slow and should have been designated as slow relative to the other constant speed conditions included in our study. Nevertheless, we are confident that the speeds chosen for this particular study accurately represent the range of preferred walking speeds in a free-living environment for a healthy, ambulatory, adult population. Jehn et al. (3) correctly note that previous studies have found inaccuracies in pedometer step counting at speeds below 0.83 m·s−1. Assessing pedometer accuracy at extremely slow speeds would most likely be uncomfortable for a healthy, ambulatory participant and would likely result in changes in walking gait such as stride length and frequency (4). Each of these walking parameters could potentially change the accuracy of the pedometer. However, healthy, ambulatory, adult individuals rarely walk at such slow speeds for a prolonged period. We agree with Jehn et al. (3) that the reliability of the piezoelectric pedometer should be validated at walking speeds of 0.83 m·s−1 or slower, which are typically reported in special populations with irregular walking gaits; however, that was not the purpose of our study. Our goal was to validate the Omron HJ-112 pedometer for different BMI groups because the advances in pedometer design were likely to make the pedometer less susceptible to problems associated with positioning in obese participants; and we did so at speeds that are preferred in a healthy, ambulatory, adult population. Rebecca E. Hasson Jeannie Haller David M. Pober Department of Kinesiology University of Massachusetts Amherst, MA John S. Staudenmayer Department of Mathematics and Statistics University of Massachusetts Amherst, MA Patty S. Freedson Department of Kinesiology University of Massachusetts Amherst, MA