Energy expenditure is calculated based on oxygen spent for any physical activity. As a rule of thumb, the oxygen uptake is multiplied approximately by 4.9 to determine the energy expenditure value. It is now-a-days very easy to measure oxygen uptake with the arrival of portable cardiopulmonary exercise testing equipment. K4B2 is one such device which is battery operated and extremely portable. The individual needs to wear the face mask.  The device measures the oxygen cost of any activity after analyzing a small sample of the expired air. Request to look into a text book titled 'Exercise Physiology: Nutrition, Energy, and Human Performance' by William D. McArdle, Frank I. Katch, Victor L. Katch for further details. In this text book you will find the energy expenditure of various types of activities like gardening, stair climbing, walking at different speeds, running with different speeds, basketball playing etc.  For your specific query, no matter the walking speed or duration of walking, you first need to estimate oxygen uptake for the activity. The obtained oxygen uptake multiplied by 4.9 would give you the energy expenditure. Oxygen consumption may also be determined from heart rate by an indirect means through a nomogram, though this method is not accurate and an approximate one. Good luck for your research endeavor. 

Thanks for your reply. I am aware of the relationship between oxygen consumption and energy expenditure so I was probably a bit unclear. My problem is exactly: how can I experimentally assess the oxygen uptake for walking at different speeds? My understanding is that during walking at a constant speed I need to reach a steady state condition to be able to accurately determine oxygen uptake and, according to the literature, this happen after 3-6 mins of this activity. But what if I continuously change speed, accelerating and decelerating for example every few seconds?  Thanks

Hai Fausto, HR based measurement might do the work. POLAR heart rate monitor is the most reliable equipment for it. it measures as little as every 5-15 sec of activity HR. KCal will be calculated based on in build formula using gender, age, height, weight and of course heart rate. other options are accelerometer, combination of accelerometer and HR (Actical, Actiheart etc). Choice is based on sample size and financial support rather than gold standard as such because prediction equation used in indirect methods are have there own limitation. as for as literature concerned, right now I found 5620 articles by putting "energy expenditure using polar heart rate monitor. pdf". all the best. 

Hi Fausto, perhaps my opinion about this problem can help and/or stimulate further discussion :-)

I think our problem can be decomposed in two sub-problems.

1) Comparing metabolic cost measurement of people who did not walk at the same average speed over the whole trajectory.

2) Measuring the instantaneous metabolic cost when subjects are continuously altering their speed.

For both solutions it is necessary to have some measurement of the walking speed (e.g. the average walking speed with a stopwatch and distance measurement or the instantaneous walking speed with a GPS or something similar)

For problem (1) I think there are two solutions.

A first solution is expressing metabolic cost per unit distance instead of per unit time. The following paper in our field also did not impose a fixed speed and used this approach to allow comparison between conditions that occurred at different speeds.

Norris JA, Granata KP, Mitros MR, Byrne EM, Marsh AP (2007) Effect of augmented plantarflexion power on preferred walking speed and economy in young and older adults. Gait Posture 25: 620–627.

The problem is that this method does not take into account the fact that the cost of transport for walking depends on walking speed. So if you measure a different cost of transport you do not know if it is due to the different walking speed or to the parameter or group you are comparing. However I think it is ok to use this when comparing measurements that are taken at close to optimal walking speed because the relationship between walking speed and cost of transport is quite “flat” there.

A second solution for problem (1) would be to do a graded walking test for each subject (and/or each condition) to determine a regression equation for the effect of walking speed on cost of transport.

Once you know this equation you can use this to derive from outdoor measurements that occurred at different speeds what would have been the effect if they would have occurred at the same speed. So you can isolate the condition or group effect from the effect of speed. But this seems a very time-intensive solution so not very practical. 

For problem (2) I think the following paper presents the solution to estimate instantaneous metabolic cost during non-steady state walking.

Selinger JC, Donelan JM (2014) Estimating instantaneous energetic cost during non-steady-state gait. J Appl Physiol 117: 1406–1415.

You need to know when the speed changes occur (e.g. from a GPS). You can then (post-hoc) fit an exponential regression with a known time-constant. The people from Sable are also suggesting this approach for improving our desktop based system.

I agree this is all not very straightforward. Probably it is best walking speed in a test where walking speed is not fixed but to measure the effect of the device we are studying in a separate test with imposed speed.

Cheers,

Philippe

Article Effect of augmented plantarflexion power on preferred walkin...

Article Estimating instantaneous energetic cost during non-steady-state gait

Hi Fausto,

Here you are a couple of papers you might be interested in:

-) Minetti, A. E., Gaudino, P., Seminati, E. and Cazzola, D., 2013. The cost of transport of human running is not affected, as in walking, by wide acceleration/deceleration cycles. Journal of applied physiology (Bethesda, Md. : 1985), 114 (4), pp. 498-503.

-) Minetti AE, Ardigò LP, Capodaglio EM, Saibene S. Energetics and mechanics of human walking at oscillating speeds. Am Zool 41: 205–210, 2001

Honestly, oxygen consumption measurement is the best way. Though, you might need to measure the area of the oxygen debt curve in order to consider the extra energy needed to accelerate the body during oscillating walking (or running) speeds.

Another approach might be to run longer trials including a certain number of acceleration and deceleration and then work out the average value.

Happy to have a chat!

Cheers

Dario

Thanks a lot Dario!

We need to find a time to chat. I'll write you soon.