- Email: [email protected]
Validity of the ActiGraph GT3X+ and SenseWear Armband to predict energy expenditure during physical activity and sport
Abstracts / Journal of Science and Medicine in Sport 20S (2017) e67–e105
238
239
Predictors of different response trajectories to non-surgical management in individuals with patellofemoral OA
Validity of the ActiGraph GT3X+ and SenseWear Armband to predict energy expenditure during physical activity and sport
K. Mills 1,∗ , J. Eyles 2,3 , M. Martin 1 , M. Hancock 4 , D. Hunter 2,3
P. Gastin 1,∗ , C. Cayzer 1 , S. Robertson 1,2 , D. Dwyer 1
1 Centre for Physcial Health, Macquarie University, Australia 2 Department of Rheumatology, Royal North Shore Hospital, Australia 3 Kolling Institute of Medical Research, Institute of Bone and Joint Research, Australia 4 Actuarial Studies and Applied Statistics, The Australian National University, Australia
1 Centre for Sport Research, Exercise and Nutrition Sciences, Deakin University, Australia 2 ISEAL, Victoria University, Australia
Introduction: Individualised non-surgical management is widely recommended for people with knee osteoarthritis (OA). However, not all patients will experience the same trajectory of response or experience improvement of the same magnitude. The aim of this study was to describe the different trajectories of response over 26-weeks as individuals with patellofemoral (PF) OA participated in a chronic care OA management program. Further, to describe baseline characteristic that predicted these trajectories. Methods: 88 participants (55 females, BMI 29.97 (SD 5.49) Age 66 (SD 6)) with a radiographic PF OA were consecutively recruited from a multidisciplinary non-surgical care program that primarily draws from the elective joint replacement list. At baseline, anthropometric, structural, pain, performance, lower limb strength and patient reported outcomes measures were recorded. At 6-, 12-, 18- and 26-weeks, participants completed a global transition scale asking “compared with when you commenced this program are you. . .” with responses ranging from “completely recovered” to “much worse”. Latent class growth analysis determined if there were unique trajectories of response to the program. Baseline characteristics of individuals who followed different trajectories were compared using general linear models and chi square tests. Significant baseline characteristics were entered into a step-wise regression to determine the best predictors of different trajectories. Results: Fifty-eight participants completed the program. Two distinct trajectories were found. Group 1 consisted of individuals who reported improvement (58%) and no change (27%) over 26-weeks; Group 2 consisted of individuals who reported marked improvement (76%) or complete recovery (14.3%) after 26-weeks. Group 1 was found to exhibit significantly higher baseline BMI (3.06 kg/m2 (95% CI 0.2, 6.1), weaker hip abductors (−0.22 Nm/kg (−0.34, −0.1), hip external rotators (−0.1 Nm/kg (−0.17, −0.03) and quadriceps (−0.35 Nm/kg (−0.55, −0.15), higher rates of anxiety (2 = 5.245, p = 0.022) and lower proximal pain thresholds (−8.75 N (−17.13, −0.37). Of these, hip abductor strength <0.63 Nm/kg and anxiety >2/5 on the DASS-21 increased the odds of being in Group 1 by 4.6 (1.3–16.37) and 5.14 (1.4–18.89) respectively. Discussion: While the majority of patients report improvement when participating in a multidisciplinary care program, there appears to be an identifiable group of patients with PF OA who respond relatively less than others. These individuals appear to exhibit weakness in the muscles that support the knee, potential central sensitization of pain and higher rates of anxiety. Importantly, all of these characteristics are modifiable, and suggest that if these characteristics are identified non-surgical OA management should commence far earlier in the disease process in order to minimise their influence on late-stage disease management. http://dx.doi.org/10.1016/j.jsams.2017.01.086
e103
Introduction: Field sports and active free-play are well suited to helping individuals meet national guidelines that focus on time spent in moderate to vigorous activity. Locomotor activities in field sports range from walking through to sprinting, and include acceleration, deceleration and change of direction. These movements are also characteristic of those observed in free-play in children. Assessment of physical activity using accelerometers is now widespread, however recent evidence questions the validity of accelerometers to predict energy expenditure (EE) at higher exercise intensities. The purpose of this study was to assess the validity of two common wearable sensors, the ActiGraph GT3X+ (GT3X + ) and the SenseWear Armband (SWA), to predict EE during exercise and field sport locomotor movements. Methods: Twenty-six active adults completed a single 90 min session involving alternating intervals of exercise (5 min) and recovery (10 min). Exercise involved walking (4 km/h), jogging (8 km/h), running (12 km/h) or a sport-simulated circuit (three intervals). Participants wore two triaxial accelerometers (GT3X+ and SWA) and a portable gas analyser (MetaMax 3B), used as the criterion measure (VO2). The overall energy cost of each exercise bout was taken as the EE (minus resting VO2) during the 5 min of exercise and the 10 min of recovery. Results: Total EE was significantly underestimated (p < 0.01) by the GT3X+ (percent difference: 29.3%; mean bias ± SD: −374.5 ± 132.84 kJ) and SWA (18.2%; −244.3 ± 148.0 kJ). Overestimations (p < 0.008) were made by both accelerometers during the walk (GT3X+: 27.4 ± 30.8 kJ; SWA: 32.1 ± 15.4 kJ) and jog (38.0 ± 30.0 kJ; 34.5 ± 31.6 kJ). Underestimations (p < 0.008) were evident during the run (GTX+: −41.2 ± 25.1 kJ; SWA: −43.8 ± 33.5 kJ) and circuit (e.g., C1: GTX+: −127.2 ± 41.6 kJ; SWA: −86.1 ± 40.2 kJ). As the intensity of exercise increased, the error of prediction increased in magnitude (GT3X+: 40.8–143.0 kJ; SWA: 35.5–102.0 kJ) and went from overestimation to underestimation (GT3X+: 25.3% to −59.3%; SWA: 36.7% to −37.3%). Effect size differences ranged from moderate to very large. Discussion: The GT3X+ and SWA do not accurately predict EE during physical activity and field sport movements. Poor accuracy and large precision errors, particularly during high intensity and intermittent movement patterns, indicated inadequate validity for use in the field. http://dx.doi.org/10.1016/j.jsams.2017.01.087
238
239
Predictors of different response trajectories to non-surgical management in individuals with patellofemoral OA
Validity of the ActiGraph GT3X+ and SenseWear Armband to predict energy expenditure during physical activity and sport
K. Mills 1,∗ , J. Eyles 2,3 , M. Martin 1 , M. Hancock 4 , D. Hunter 2,3
P. Gastin 1,∗ , C. Cayzer 1 , S. Robertson 1,2 , D. Dwyer 1
1 Centre for Physcial Health, Macquarie University, Australia 2 Department of Rheumatology, Royal North Shore Hospital, Australia 3 Kolling Institute of Medical Research, Institute of Bone and Joint Research, Australia 4 Actuarial Studies and Applied Statistics, The Australian National University, Australia
1 Centre for Sport Research, Exercise and Nutrition Sciences, Deakin University, Australia 2 ISEAL, Victoria University, Australia
Introduction: Individualised non-surgical management is widely recommended for people with knee osteoarthritis (OA). However, not all patients will experience the same trajectory of response or experience improvement of the same magnitude. The aim of this study was to describe the different trajectories of response over 26-weeks as individuals with patellofemoral (PF) OA participated in a chronic care OA management program. Further, to describe baseline characteristic that predicted these trajectories. Methods: 88 participants (55 females, BMI 29.97 (SD 5.49) Age 66 (SD 6)) with a radiographic PF OA were consecutively recruited from a multidisciplinary non-surgical care program that primarily draws from the elective joint replacement list. At baseline, anthropometric, structural, pain, performance, lower limb strength and patient reported outcomes measures were recorded. At 6-, 12-, 18- and 26-weeks, participants completed a global transition scale asking “compared with when you commenced this program are you. . .” with responses ranging from “completely recovered” to “much worse”. Latent class growth analysis determined if there were unique trajectories of response to the program. Baseline characteristics of individuals who followed different trajectories were compared using general linear models and chi square tests. Significant baseline characteristics were entered into a step-wise regression to determine the best predictors of different trajectories. Results: Fifty-eight participants completed the program. Two distinct trajectories were found. Group 1 consisted of individuals who reported improvement (58%) and no change (27%) over 26-weeks; Group 2 consisted of individuals who reported marked improvement (76%) or complete recovery (14.3%) after 26-weeks. Group 1 was found to exhibit significantly higher baseline BMI (3.06 kg/m2 (95% CI 0.2, 6.1), weaker hip abductors (−0.22 Nm/kg (−0.34, −0.1), hip external rotators (−0.1 Nm/kg (−0.17, −0.03) and quadriceps (−0.35 Nm/kg (−0.55, −0.15), higher rates of anxiety (2 = 5.245, p = 0.022) and lower proximal pain thresholds (−8.75 N (−17.13, −0.37). Of these, hip abductor strength <0.63 Nm/kg and anxiety >2/5 on the DASS-21 increased the odds of being in Group 1 by 4.6 (1.3–16.37) and 5.14 (1.4–18.89) respectively. Discussion: While the majority of patients report improvement when participating in a multidisciplinary care program, there appears to be an identifiable group of patients with PF OA who respond relatively less than others. These individuals appear to exhibit weakness in the muscles that support the knee, potential central sensitization of pain and higher rates of anxiety. Importantly, all of these characteristics are modifiable, and suggest that if these characteristics are identified non-surgical OA management should commence far earlier in the disease process in order to minimise their influence on late-stage disease management. http://dx.doi.org/10.1016/j.jsams.2017.01.086
e103
Introduction: Field sports and active free-play are well suited to helping individuals meet national guidelines that focus on time spent in moderate to vigorous activity. Locomotor activities in field sports range from walking through to sprinting, and include acceleration, deceleration and change of direction. These movements are also characteristic of those observed in free-play in children. Assessment of physical activity using accelerometers is now widespread, however recent evidence questions the validity of accelerometers to predict energy expenditure (EE) at higher exercise intensities. The purpose of this study was to assess the validity of two common wearable sensors, the ActiGraph GT3X+ (GT3X + ) and the SenseWear Armband (SWA), to predict EE during exercise and field sport locomotor movements. Methods: Twenty-six active adults completed a single 90 min session involving alternating intervals of exercise (5 min) and recovery (10 min). Exercise involved walking (4 km/h), jogging (8 km/h), running (12 km/h) or a sport-simulated circuit (three intervals). Participants wore two triaxial accelerometers (GT3X+ and SWA) and a portable gas analyser (MetaMax 3B), used as the criterion measure (VO2). The overall energy cost of each exercise bout was taken as the EE (minus resting VO2) during the 5 min of exercise and the 10 min of recovery. Results: Total EE was significantly underestimated (p < 0.01) by the GT3X+ (percent difference: 29.3%; mean bias ± SD: −374.5 ± 132.84 kJ) and SWA (18.2%; −244.3 ± 148.0 kJ). Overestimations (p < 0.008) were made by both accelerometers during the walk (GT3X+: 27.4 ± 30.8 kJ; SWA: 32.1 ± 15.4 kJ) and jog (38.0 ± 30.0 kJ; 34.5 ± 31.6 kJ). Underestimations (p < 0.008) were evident during the run (GTX+: −41.2 ± 25.1 kJ; SWA: −43.8 ± 33.5 kJ) and circuit (e.g., C1: GTX+: −127.2 ± 41.6 kJ; SWA: −86.1 ± 40.2 kJ). As the intensity of exercise increased, the error of prediction increased in magnitude (GT3X+: 40.8–143.0 kJ; SWA: 35.5–102.0 kJ) and went from overestimation to underestimation (GT3X+: 25.3% to −59.3%; SWA: 36.7% to −37.3%). Effect size differences ranged from moderate to very large. Discussion: The GT3X+ and SWA do not accurately predict EE during physical activity and field sport movements. Poor accuracy and large precision errors, particularly during high intensity and intermittent movement patterns, indicated inadequate validity for use in the field. http://dx.doi.org/10.1016/j.jsams.2017.01.087