- Email: [email protected]
Combining GPS with heart rate monitoring to measure physical activity in children: A feasibility study
Available online at www.sciencedirect.com
Journal of Science and Medicine in Sport 12 (2009) 583–585
Original paper
Combining GPS with heart rate monitoring to measure physical activity in children: A feasibility study J. Scott Duncan ∗ , Hannah M. Badland, Grant Schofield Centre for Physical Activity and Nutrition Research, Faculty of Health and Environmental Sciences, Auckland University of Technology, New Zealand Received 2 May 2008; received in revised form 26 August 2008; accepted 11 September 2008
Abstract The recent development of global positioning system (GPS) receivers with integrated heart rate (HR) monitoring has provided a new method for estimating the energy expenditure associated with children’s movement. The purpose of this feasibility study was to trial a combination of GPS surveillance and HR monitoring in 39 primary-aged children from New Zealand. Spatial location and HR data were recorded during a school lunch break using an integrated GPS/HR receiver (1 Hz). Children averaged a total distance of 1.10 ± 0.56 km at speeds ranging from 0 to 18.6 km h−1 . Activity patterns were characterised by short bursts of moderate to high speeds followed by longer periods of slow speeds. In addition, boys averaged higher speeds than girls (1.77 ± 0.62 km h−1 and 1.36 ± 0.50 km h−1 , respectively; p = 0.003). The percentage of time spent at 0 km h−1 (stationary) ranged from 0.1% to 21.3% with a mean of 6.4 ± 4.6%. These data suggest that while children were relatively active during the lunch period, they spent a substantial portion of time engaged in slow or stationary physical activities. Furthermore, associations between HR, average speed, and stationary time demonstrated that children who moved at faster speeds expended more energy than those who moved at slower speeds. We conclude that the combined approach of GPS and HR monitoring is a promising new method for investigating children’s play-related energy expenditure. There is also scope to integrate GPS data with geographic information systems to examine where children play and accumulate physical activity. © 2008 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Keywords: Activities of daily living; Epidemiology; Global positioning; Heart rate; Methods; Physical fitness
1. Introduction The potential of portable global positioning systems (GPS) to provide researchers with an objective assessment of physical activity location was first posited over 10 years ago.1 It is only recently, however, that two pilot studies have combined GPS with accelerometry and geographical information systems (GIS) to provide an environmental context for free-living physical activity in adults.2,3 While such information could contribute to our understanding of how urban design elements influence overall and travel-related physical activity in people of all ages, GPS monitoring has yet to be tested in a paediatric sample. Recently developed GPS models feature integrated heart rate (HR) receivers for monitoring athlete performance during outdoor sports. We propose that these devices may also provide a means to estimate the ∗
Corresponding author. Tel.: +64 9 921 9999; fax: +64 9 921 9746. E-mail address: [email protected] (J.S. Duncan).
energy expenditure associated with children’s movement patterns during free play. Thus, the purpose of this feasibility study was to trial a combination of GPS surveillance and HR monitoring in primary-aged children.
2. Methods A total of 40 children (20 boys, 20 girls) from school Years 1–2 (5–7 years) and 5–6 (9–10 years) were randomly selected from a local primary school. The two age groups were targeted to facilitate comparisons between younger and older children. Only one participant did not provide useable data, resulting in a final sample size of 39 children (20 boys, 19 girls). Each participant and their legal guardian provided written informed consent. Spatial location was assessed to the nearest metre using a 12-channel F500 GPS receiver (FRWD Technologies Ltd., Oulu, Finland), while a coded transmitter belt (Polar Electro, Kempele, Finland) attached to the
1440-2440/$ – see front matter © 2008 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jsams.2008.09.010
584
J.S. Duncan et al. / Journal of Science and Medicine in Sport 12 (2009) 583–585
Table 1 Descriptive data for age, duration, distance, and speed. Boys
Girls
Years 1–2 (n = 10) Age (years) Duration of observation (min) Maximum N–S distance (m) Maximum E–W distance (m) Total distance travelled (km) Average speed (km h−1 ) Maximum speed (km h−1 )
5.89 ± 0.609 (5.14–6.86) 44.7 ± 13.2 (17.0–58.2) 121 ± 59.8 (37.0–260) 88.9 ± 28.7 (33.0–123)
Years 5–6 (n = 10)
Years 1–2 (n = 10)
Years 5–6 (n = 9)
9.81 ± 0.52 (9.08–10.7) 41.7 ± 8.81 (20.9–53.4)
6.07 ± 0.627 (5.30–7.22) 51.7 ± 8.32 (42.2–67.8)
95.4 ± 37.0 (37.0–138)
90.7 ± 38.8 (29.0–157)
101 ± 41.6 (15.0–154)
82.6 ± 25.7 (48.0–130)
121 ± 45.7 (20.0–180)
126 ± 22.1 (73.0–161)
10.0 ± 0.528 (9.34–10.9) 44.4 ± 6.75 (34.5–57.9)
1.38 ± 0.637 (0.235–2.46)
0.972 ± 0.460 (0.207–1.90) 1.15 ± 0.691 (0.380–2.86)
0.895 ± 0.292 (0.427–1.36)
1.78 ± 0.594 (1.07–2.81)
1.77 ± 0.675 (0.260–2.63)
1.33 ± 0.433 (0.640–2.13)
10.4 ± 3.52 (7.20–18.6)
11.0 ± 5.04 (3.40–17.4)
1.39 ± 0.580 (0.590–2.56) 8.40 ± 3.63 (4.00–14.1)
10.4 ± 3.00 (4.00–13.3)
Data are presented as mean ± S.D. (range).
chest of each participant enabled HR monitoring. The F500 model is well suited for child monitoring given the water resistant construction and the lack of external buttons or controls. Location, distance, speed, and HR data were collected during school lunch periods using a 1-s recording interval. Units were attached to the back of the children using a harness designed by the manufacturers specifically for the F500 receiver. Children were instructed to wear the unit during their normal activities and return to the researchers at the end of the lunch period. This process was completed and during the 10 min children were required to sit and eat their lunch. After this initial period, children were free to play for an additional 50 min until the end of the lunch break. The resting heart rate (RHR) of each participant was measured in the early morning after lying supine for 10 min. This enabled the percentage of time spent above 25% (PAHR-25) and 50% (PAHR-50) of RHR to be calculated as a standardised measure of energy expenditure (EE).4 All data were imported directly from the GPS unit (via bluetooth connection) as CSV files before collation in a SPSS spreadsheet (SPSS Inc., Chicago, IL) for analyses. Descriptive statistics for all variables of interest were generated, and associations between selected factors investigated using correlation analysis. Ethical approval for this study was obtained from our institutional ethics committee.
than Year 1–2 children (p < 0.05 within genders and p < 0.01 across the whole sample). Activity patterns were characterised by short bursts of moderate to high speeds followed by longer periods of slow speeds. Fig. 1 shows the distribution of time spent at each average speed category; no differences in the percentage of time in each speed group were detected between genders or age groups and subsequently all data were combined. The percentage of time spent at 0 km h−1 (stationary) ranged from 0.1% to 21.3% with a mean of 6.4 ± 4.6%. The average HR of this sample was 141 ± 17 bpm, with an average PAHR-25 and PAHR-50 of 93.4 ± 12.8% and 68.1 ± 26.3%, respectively. PAHR-25 and PAHR-50 were positively correlated with average speed (r = 0.411 and 0.344; p = 0.009 and 0.032) and PAHR-25 was negatively correlated with time spent at 0 km h−1 (r = −0.414; p = 0.009).
4. Discussion This study used a portable GPS unit to provide the first data describing the spatial movement of children when engaged
3. Results During the observation period (approximately 50 min), children averaged a distance of 1.10 ± 0.56 km at speeds ranging from 0 to 18.6 km h−1 . Table 1 shows the GPS variables grouped by gender and year level. While no significant gender differences were detected within each year level, boys of all ages averaged a higher speed (1.77 ± 0.62 km h−1 ) than girls (1.36 ± 0.50 km h−1 ; p = 0.003). The only significant difference evident between year levels was the maximum East–West distance, with Year 5–6 showing higher values
Fig. 1. Percentage of total time (mean ± S.D.) spent at selected speed categories.
J.S. Duncan et al. / Journal of Science and Medicine in Sport 12 (2009) 583–585
in physical activity. Our results suggest that most children are active during the lunch break, spending the majority of time in motion. Interestingly, the most common movement speeds (between 0.1 and 2.5 km h−1 ) were slower than previous estimates of children’s self-selected slow walking pace (5.0 ± 0.5 km h−1 ).5 This finding is of particular relevance to researchers who use monitoring devices that underestimate activity at slow walking speeds.6,7 However, it was apparent from our observations that children moving at low speeds often engage in non-locomotor activities (e.g. bouncing a ball, jumping) rather than slow walking. Nevertheless, our results show that children who move at faster speeds expend more energy than those who exhibit slower movement patterns. It was interesting to note the gender differences in average and maximum speeds; this was likely due to differences in the type of sports preferred by boys and girls, or the intensity at which the sports are played. In addition to providing a profile of children’s movement patterns, GPS data can be imported into a GIS database to determine the areas most frequently used for active play. Locating activity is especially useful for studies of the built environment and its effect on behaviour. In the present study, variation in the E–W distance between younger and older children may reflect differences in the school areas preferred for physical activity. Enhancing this information with GIS integration would enable the facilities that support physical activity engagement to be developed and promoted across schools. While GPS is only effective in outdoor areas not covered by trees or other structures, the integration of an HR monitor within the GPS unit enables EE to be recorded both indoors and outdoors, providing an ongoing record of activity frequency, intensity, and duration. Overall, our experience using GPS in children was relatively straightforward. The specially designed harness held the receiver in place adequately, and the lack of display or obvious buttons on the device precluded any interference from the children or their friends. However, data collection took place in a restricted outdoor area over a relatively short timeframe. The application of GPS technology in a freeliving setting over multiple days will require further testing. For example, extended periods spent indoors may hinder GPS signal acquisition and result in excessive data loss. The posi-
585
tion of the GPS unit on the back of the children may also need to be reviewed if they are unable to sit comfortably. Expansion of this pilot study across several days would provide valuable information regarding the feasibility of GPS/HR methodology in free-living children.
5. Conclusion Combining GPS with HR monitoring is a promising new method for investigating both the spatial location and EE associated with children’s physical activity.
Acknowledgments This research was supported by a grant from the School of Sport and Recreation, Auckland University of Technology. HB is supported by a New Zealand National Heart Foundation Research Fellowship.
References 1. Schutz Y, Chambaz A. Could a satellite-based navigation system (GPS) be used to assess the physical activity of individuals on earth? Eur J Clin Nutr 1997;51(5):338–9. 2. Rodriguez DA, Brown AL, Troped PJ. Portable global positioning units to complement accelerometry-based physical activity monitors. Med Sci Sports Exerc 2005;37(Suppl. 5):S572–81. 3. Troped PJ, Oliveira MS, Matthews CE, et al. Prediction of activity mode with global positioning system and accelerometer data. Med Sci Sports Exerc 2008;40(5):972–8. 4. Trost SG. Objective measurement of physical activity in youth: current issues, future directions. Exerc Sport Sci Rev 2001;29(1):32–6. 5. Trost SG, Way R, Okely AD. Predictive validity of three ActiGraph energy expenditure equations for children. Med Sci Sports Exerc 2006;38(2):380–7. 6. Le Masurier GC, Tudor-Locke C. Comparison of pedometer and accelerometer accuracy under controlled conditions. Med Sci Sports Exerc 2003;35(5):867–71. 7. Duncan JS, Schofield G, Duncan EK, Hinckson EA. Effects of age, walking speed, and body composition on pedometer accuracy in children. Res Q Exerc Sport 2007;78(5):420–8.
Journal of Science and Medicine in Sport 12 (2009) 583–585
Original paper
Combining GPS with heart rate monitoring to measure physical activity in children: A feasibility study J. Scott Duncan ∗ , Hannah M. Badland, Grant Schofield Centre for Physical Activity and Nutrition Research, Faculty of Health and Environmental Sciences, Auckland University of Technology, New Zealand Received 2 May 2008; received in revised form 26 August 2008; accepted 11 September 2008
Abstract The recent development of global positioning system (GPS) receivers with integrated heart rate (HR) monitoring has provided a new method for estimating the energy expenditure associated with children’s movement. The purpose of this feasibility study was to trial a combination of GPS surveillance and HR monitoring in 39 primary-aged children from New Zealand. Spatial location and HR data were recorded during a school lunch break using an integrated GPS/HR receiver (1 Hz). Children averaged a total distance of 1.10 ± 0.56 km at speeds ranging from 0 to 18.6 km h−1 . Activity patterns were characterised by short bursts of moderate to high speeds followed by longer periods of slow speeds. In addition, boys averaged higher speeds than girls (1.77 ± 0.62 km h−1 and 1.36 ± 0.50 km h−1 , respectively; p = 0.003). The percentage of time spent at 0 km h−1 (stationary) ranged from 0.1% to 21.3% with a mean of 6.4 ± 4.6%. These data suggest that while children were relatively active during the lunch period, they spent a substantial portion of time engaged in slow or stationary physical activities. Furthermore, associations between HR, average speed, and stationary time demonstrated that children who moved at faster speeds expended more energy than those who moved at slower speeds. We conclude that the combined approach of GPS and HR monitoring is a promising new method for investigating children’s play-related energy expenditure. There is also scope to integrate GPS data with geographic information systems to examine where children play and accumulate physical activity. © 2008 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Keywords: Activities of daily living; Epidemiology; Global positioning; Heart rate; Methods; Physical fitness
1. Introduction The potential of portable global positioning systems (GPS) to provide researchers with an objective assessment of physical activity location was first posited over 10 years ago.1 It is only recently, however, that two pilot studies have combined GPS with accelerometry and geographical information systems (GIS) to provide an environmental context for free-living physical activity in adults.2,3 While such information could contribute to our understanding of how urban design elements influence overall and travel-related physical activity in people of all ages, GPS monitoring has yet to be tested in a paediatric sample. Recently developed GPS models feature integrated heart rate (HR) receivers for monitoring athlete performance during outdoor sports. We propose that these devices may also provide a means to estimate the ∗
Corresponding author. Tel.: +64 9 921 9999; fax: +64 9 921 9746. E-mail address: [email protected] (J.S. Duncan).
energy expenditure associated with children’s movement patterns during free play. Thus, the purpose of this feasibility study was to trial a combination of GPS surveillance and HR monitoring in primary-aged children.
2. Methods A total of 40 children (20 boys, 20 girls) from school Years 1–2 (5–7 years) and 5–6 (9–10 years) were randomly selected from a local primary school. The two age groups were targeted to facilitate comparisons between younger and older children. Only one participant did not provide useable data, resulting in a final sample size of 39 children (20 boys, 19 girls). Each participant and their legal guardian provided written informed consent. Spatial location was assessed to the nearest metre using a 12-channel F500 GPS receiver (FRWD Technologies Ltd., Oulu, Finland), while a coded transmitter belt (Polar Electro, Kempele, Finland) attached to the
1440-2440/$ – see front matter © 2008 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jsams.2008.09.010
584
J.S. Duncan et al. / Journal of Science and Medicine in Sport 12 (2009) 583–585
Table 1 Descriptive data for age, duration, distance, and speed. Boys
Girls
Years 1–2 (n = 10) Age (years) Duration of observation (min) Maximum N–S distance (m) Maximum E–W distance (m) Total distance travelled (km) Average speed (km h−1 ) Maximum speed (km h−1 )
5.89 ± 0.609 (5.14–6.86) 44.7 ± 13.2 (17.0–58.2) 121 ± 59.8 (37.0–260) 88.9 ± 28.7 (33.0–123)
Years 5–6 (n = 10)
Years 1–2 (n = 10)
Years 5–6 (n = 9)
9.81 ± 0.52 (9.08–10.7) 41.7 ± 8.81 (20.9–53.4)
6.07 ± 0.627 (5.30–7.22) 51.7 ± 8.32 (42.2–67.8)
95.4 ± 37.0 (37.0–138)
90.7 ± 38.8 (29.0–157)
101 ± 41.6 (15.0–154)
82.6 ± 25.7 (48.0–130)
121 ± 45.7 (20.0–180)
126 ± 22.1 (73.0–161)
10.0 ± 0.528 (9.34–10.9) 44.4 ± 6.75 (34.5–57.9)
1.38 ± 0.637 (0.235–2.46)
0.972 ± 0.460 (0.207–1.90) 1.15 ± 0.691 (0.380–2.86)
0.895 ± 0.292 (0.427–1.36)
1.78 ± 0.594 (1.07–2.81)
1.77 ± 0.675 (0.260–2.63)
1.33 ± 0.433 (0.640–2.13)
10.4 ± 3.52 (7.20–18.6)
11.0 ± 5.04 (3.40–17.4)
1.39 ± 0.580 (0.590–2.56) 8.40 ± 3.63 (4.00–14.1)
10.4 ± 3.00 (4.00–13.3)
Data are presented as mean ± S.D. (range).
chest of each participant enabled HR monitoring. The F500 model is well suited for child monitoring given the water resistant construction and the lack of external buttons or controls. Location, distance, speed, and HR data were collected during school lunch periods using a 1-s recording interval. Units were attached to the back of the children using a harness designed by the manufacturers specifically for the F500 receiver. Children were instructed to wear the unit during their normal activities and return to the researchers at the end of the lunch period. This process was completed and during the 10 min children were required to sit and eat their lunch. After this initial period, children were free to play for an additional 50 min until the end of the lunch break. The resting heart rate (RHR) of each participant was measured in the early morning after lying supine for 10 min. This enabled the percentage of time spent above 25% (PAHR-25) and 50% (PAHR-50) of RHR to be calculated as a standardised measure of energy expenditure (EE).4 All data were imported directly from the GPS unit (via bluetooth connection) as CSV files before collation in a SPSS spreadsheet (SPSS Inc., Chicago, IL) for analyses. Descriptive statistics for all variables of interest were generated, and associations between selected factors investigated using correlation analysis. Ethical approval for this study was obtained from our institutional ethics committee.
than Year 1–2 children (p < 0.05 within genders and p < 0.01 across the whole sample). Activity patterns were characterised by short bursts of moderate to high speeds followed by longer periods of slow speeds. Fig. 1 shows the distribution of time spent at each average speed category; no differences in the percentage of time in each speed group were detected between genders or age groups and subsequently all data were combined. The percentage of time spent at 0 km h−1 (stationary) ranged from 0.1% to 21.3% with a mean of 6.4 ± 4.6%. The average HR of this sample was 141 ± 17 bpm, with an average PAHR-25 and PAHR-50 of 93.4 ± 12.8% and 68.1 ± 26.3%, respectively. PAHR-25 and PAHR-50 were positively correlated with average speed (r = 0.411 and 0.344; p = 0.009 and 0.032) and PAHR-25 was negatively correlated with time spent at 0 km h−1 (r = −0.414; p = 0.009).
4. Discussion This study used a portable GPS unit to provide the first data describing the spatial movement of children when engaged
3. Results During the observation period (approximately 50 min), children averaged a distance of 1.10 ± 0.56 km at speeds ranging from 0 to 18.6 km h−1 . Table 1 shows the GPS variables grouped by gender and year level. While no significant gender differences were detected within each year level, boys of all ages averaged a higher speed (1.77 ± 0.62 km h−1 ) than girls (1.36 ± 0.50 km h−1 ; p = 0.003). The only significant difference evident between year levels was the maximum East–West distance, with Year 5–6 showing higher values
Fig. 1. Percentage of total time (mean ± S.D.) spent at selected speed categories.
J.S. Duncan et al. / Journal of Science and Medicine in Sport 12 (2009) 583–585
in physical activity. Our results suggest that most children are active during the lunch break, spending the majority of time in motion. Interestingly, the most common movement speeds (between 0.1 and 2.5 km h−1 ) were slower than previous estimates of children’s self-selected slow walking pace (5.0 ± 0.5 km h−1 ).5 This finding is of particular relevance to researchers who use monitoring devices that underestimate activity at slow walking speeds.6,7 However, it was apparent from our observations that children moving at low speeds often engage in non-locomotor activities (e.g. bouncing a ball, jumping) rather than slow walking. Nevertheless, our results show that children who move at faster speeds expend more energy than those who exhibit slower movement patterns. It was interesting to note the gender differences in average and maximum speeds; this was likely due to differences in the type of sports preferred by boys and girls, or the intensity at which the sports are played. In addition to providing a profile of children’s movement patterns, GPS data can be imported into a GIS database to determine the areas most frequently used for active play. Locating activity is especially useful for studies of the built environment and its effect on behaviour. In the present study, variation in the E–W distance between younger and older children may reflect differences in the school areas preferred for physical activity. Enhancing this information with GIS integration would enable the facilities that support physical activity engagement to be developed and promoted across schools. While GPS is only effective in outdoor areas not covered by trees or other structures, the integration of an HR monitor within the GPS unit enables EE to be recorded both indoors and outdoors, providing an ongoing record of activity frequency, intensity, and duration. Overall, our experience using GPS in children was relatively straightforward. The specially designed harness held the receiver in place adequately, and the lack of display or obvious buttons on the device precluded any interference from the children or their friends. However, data collection took place in a restricted outdoor area over a relatively short timeframe. The application of GPS technology in a freeliving setting over multiple days will require further testing. For example, extended periods spent indoors may hinder GPS signal acquisition and result in excessive data loss. The posi-
585
tion of the GPS unit on the back of the children may also need to be reviewed if they are unable to sit comfortably. Expansion of this pilot study across several days would provide valuable information regarding the feasibility of GPS/HR methodology in free-living children.
5. Conclusion Combining GPS with HR monitoring is a promising new method for investigating both the spatial location and EE associated with children’s physical activity.
Acknowledgments This research was supported by a grant from the School of Sport and Recreation, Auckland University of Technology. HB is supported by a New Zealand National Heart Foundation Research Fellowship.
References 1. Schutz Y, Chambaz A. Could a satellite-based navigation system (GPS) be used to assess the physical activity of individuals on earth? Eur J Clin Nutr 1997;51(5):338–9. 2. Rodriguez DA, Brown AL, Troped PJ. Portable global positioning units to complement accelerometry-based physical activity monitors. Med Sci Sports Exerc 2005;37(Suppl. 5):S572–81. 3. Troped PJ, Oliveira MS, Matthews CE, et al. Prediction of activity mode with global positioning system and accelerometer data. Med Sci Sports Exerc 2008;40(5):972–8. 4. Trost SG. Objective measurement of physical activity in youth: current issues, future directions. Exerc Sport Sci Rev 2001;29(1):32–6. 5. Trost SG, Way R, Okely AD. Predictive validity of three ActiGraph energy expenditure equations for children. Med Sci Sports Exerc 2006;38(2):380–7. 6. Le Masurier GC, Tudor-Locke C. Comparison of pedometer and accelerometer accuracy under controlled conditions. Med Sci Sports Exerc 2003;35(5):867–71. 7. Duncan JS, Schofield G, Duncan EK, Hinckson EA. Effects of age, walking speed, and body composition on pedometer accuracy in children. Res Q Exerc Sport 2007;78(5):420–8.