Predicting metabolic rate during level and uphill outdoor walking using a low-cost GPS receiver

Abstract

The objective of this study was to assess the accuracy of using speed and grade data obtained from a low-cost global positioning system (GPS) receiver to estimate metabolic rate (MR) during level and uphill outdoor walking. Thirty young, healthy adults performed randomized outdoor walking for 6-min periods at 2.0, 3.5, and 5.0 km/h and on three different grades: 1) level walking, 2) uphill walking on a 3.7% mean grade, and 3) uphill walking on a 10.8% mean grade. The reference MR [metabolic equivalents (METs) and oxygen uptake (V̇o2)] values were obtained using a portable metabolic system. The speed and grade were obtained using a low-cost GPS receiver (1-Hz recording). The GPS grade (Δ altitude/distance walked) was calculated using both uncorrected GPS altitude data and GPS altitude data corrected with map projection software. The accuracy of predictions using reference speed and grade (actual[SPEED/GRADE]) data was high [R(2) = 0.85, root-mean-square error (RMSE) = 0.68 MET]. The accuracy decreased when GPS speed and uncorrected grade (GPS[UNCORRECTED]) data were used, although it remained substantial (R(2) = 0.66, RMSE = 1.00 MET). The accuracy was greatly improved when the GPS speed and corrected grade (GPS[CORRECTED]) data were used (R(2) = 0.82, RMSE = 0.79 MET). Published predictive equations for walking MR were also cross-validated using actual or GPS speed and grade data when appropriate. The prediction accuracy was very close when either actual[SPEED/GRADE] values or GPS[CORRECTED] values (for level and uphill combined) or GPS speed values (for level walking only) were used. These results offer promising research and clinical applications related to the assessment of energy expenditure during free-living walking.

Using GPS, accelerometry and heart rate to predict outdoor graded walking energy expenditure

“…The methods of the Acti-GPS project have been extensively described elsewhere. 9 During the inclusion visit, the participants provided their medical histories and underwent a physical examination. Weight, height, and body fat percentage using the skin fold method were measured, as previously reported.…”

“…Weight, height, and body fat percentage using the skin fold method were measured, as previously reported. 9 The resting metabolic rate was measured early in the morning (between 07:00 and 10:30) in the supine position and in a noise-free and thermoneutral (22  1°C) room. Participants were asked to fast for at least 12 h, to refrain from vigorous PA and alcohol for 24 h, and to sleep for 8 h or more prior to testing.…”

“…The V O2peak value was determined using the K4b 2 during an outdoor maximal incremental field-running test performed on a 400m athletic track. 9 On 2 or 3 other and different days, each participant completed six-minute walking periods at speeds of 2.0, 3.5 and 5.0 kmh -1 during each of the three outdoor walking sessions (0.0%, 3.4% and 10.4% mean grades for the three sessions, respectively). During the outdoor walking sessions on the 3.4% and 10.4% mean grades, each walking speed was tested uphill and downhill (Supplemental Material (SM) #1).…”

“…The K4b 2 was calibrated according to the manufacturer's guidelines before each test, and reliability data have been previously published. 9 During all walking sessions, participants were equipped with a low-cost DG100 GPS receiver (GlobalSat® Inc., Taipei, Taiwan; ~AUD80) that recorded at a 1-Hz sampling rate, as previously described. 9 Participants also wore a wGT3X+ set at a 90-Hz sampling rate on the iliac crest of the hip (ActiGraph TM , LLC, Pensacola, FL; ~AUD295; firmware 2.2.1).…”

“…8 Interestingly, we recently showed that the use of speed and grade data obtained using a low-cost global positioning system (GPS) receiver allowed highly accurate estimations of EE during level and uphill outdoor walking compared to the use of actual speed and grade. 9 Although that study supported the feasibility of using GPS alone for estimating EE during level and graded outdoor walking, it is unknown how GPS compares with classical methods, such as accelerometry or HR monitoring, for this purpose.…”