The metabolic cost of changing speeds is significant and may affect preferred walking speeds

In daily life, people seldom walk at exactly constant speeds. Most walking over a day is performed in short bouts [1]. Thus, it is of interest to measure the cost of changing speed. Minetti and co-workers [3] measured the cost of changing speeds but with non-inertial treadmill speed changes or stepfrequency control. Here, we measure the metabolic cost of changing speeds to better quantify the relative contribution of changing-speed versus constant-speed walking costs. Human preferring walking speeds are found to be close to energetically optimal speeds [4]. The cost of changing speeds may cause a change in preferred walking behavior, in some situations. We test this by measuring the walking speed for a range of short-distance bouts, for which acceleration-deceleration constitutes a significant fraction of the total bout.