Introduction: Accelerometers are commonly used to assess time-use behaviors related to physical activity, sedentary behavior, and sleep; however, as new accelerometer technologies emerge, it is important to ensure consistency with previous devices. This study aimed to evaluate the concurrent validity of the commonly used accelerometer, ActiGraph GT3X+, and the relatively new Axivity AX3 (fastened to the lower back) for detecting physical activity intensity and body postures when using direct observation as the criterion measure. Methods: A total of 41 children (aged 6–16 years) and 33 adults (aged 28–59 years) wore both monitors concurrently while performing 10 prescribed activities under laboratory conditions. The GT3X+ data were categorized into different physical activity intensity and posture categories using intensity-based cut points and ActiGraph proprietary inclinometer algorithms, respectively. The AX3 data were first converted to ActiGraph counts before being categorized into different physical activity intensity categories, while activity recognition models were used to detect the target postures. Sensitivity, specificity, and the balanced accuracy for intensity and posture category classification were calculated for each accelerometer. Differences in balanced accuracy between the devices and between children and adults were also calculated. Results: Both accelerometers obtained 74–96% balanced accuracy, with the AX3 performing slightly better (∼4% higher, p < .01) for detecting postures and physical activity intensity. Error in both devices was greatest when contrasting sitting/standing, sedentary/light intensity, and moderate/light intensity. Conclusion: In comparison with the GT3X+ accelerometer, AX3 was able to detect various postures and activity intensities with slightly higher balanced accuracy in children and adults.
Leila Hedayatrad, Tom Stewart, and Scott Duncan
Geeta Sharma, Tom Stewart, and Scott Duncan
Background: Curriculum-integrated dance programs are a promising but relatively under-researched strategy for increasing children’s physical activity (PA). The aim of this study was to determine the impact of a curriculum-integrated dance program on children’s PA. Methods: A total of 134 primary children aged 7–9 years from 4 New Zealand schools were assigned to either a dance group (n = 78) or a control group (n = 56). The dance group participated in a 6-week curriculum-integrated dance program during school time. Although the dance program focused on curricular learning, fitness and coordination were embedded in the dance sessions. Intensity of PA varied according to the focus of each dance session. PA was measured at baseline and postintervention using a waist-mounted ActiGraph GT3X+ accelerometer for 8 consecutive days. Results: There were no significant intervention effects on PA levels between the dance and control groups postintervention. Conclusion: Dance-embedded learning did not increase overall levels of PA in this study. Future studies may consider assessing longer term effects of a dance-based intervention, or programs that place more focus on PA promotion.
Anantha Narayanan, Farzanah Desai, Tom Stewart, Scott Duncan, and Lisa Mackay
Background: Application of machine learning for classifying human behavior is increasingly common as access to raw accelerometer data improves. The aims of this scoping review are (1) to examine if machine-learning techniques can accurately identify human activity behaviors from raw accelerometer data and (2) to summarize the practical implications of these machine-learning techniques for future work. Methods: Keyword searches were performed in Scopus, Web of Science, and EBSCO databases in 2018. Studies that applied supervised machine-learning techniques to raw accelerometer data and estimated components of physical activity were included. Information on study characteristics, machine-learning techniques, and key study findings were extracted from included studies. Results: Of the 53 studies included in the review, 75% were published in the last 5 years. Most studies predicted postures and activity type, rather than intensity, and were conducted in controlled environments using 1 or 2 devices. The most common models were support vector machine, random forest, and artificial neural network. Overall, classification accuracy ranged from 62% to 99.8%, although nearly 80% of studies achieved an overall accuracy above 85%. Conclusions: Machine-learning algorithms demonstrate good accuracy when predicting physical activity components; however, their application to free-living settings is currently uncertain.
Melody Oliver, Scott Duncan, Celia Kuch, Julia McPhee, and Grant Schofield
Aims were to investigate sex, ethnicity, and age differences in achieving daily step count and television (TV) watching recommendations in schoolchildren.
Participants were 615 children (n = 325) and adolescents (n = 290) aged 5 to 16 years. Activity was assessed over 5 days using pedometers; TV time was collected via parental proxy-report and self-report. Ethnic, sex, and age differences in step counts, TV time, and odds of meeting TV and step count recommendations were examined for weekdays, weekend days, and overall using generalized estimation equation modeling.
Overall, boys were more active than girls (P < .001). Adolescents were more active than children (P = .044), watched more TV (P = .005), and were less likely to meet TV watching recommendations (P = .004). Non-European children watched significantly more TV (P = .008), and were significantly less likely to meet TV recommendations than non-European children (P = .001). Participants watched more TV and accumulated less steps on weekend days than weekdays.
Multifaceted interventions focusing on both increasing activity and decreasing TV time are needed, especially on weekends. Children and girls may benefit more from activity interventions, while ethnic-specific interventions focusing on TV habits may be most efficacious for adolescents.
Carolyn A. Duncan, Scott N. MacKinnon, and Wayne J. Albert
The purpose of this study was to examine how wave-induced platform motion effects postural stability when handling loads. Twelve participants (9 male, 3 female) performed a sagittal lifting/lowering task with a 10 kg load in different sea conditions off the coast of Halifax, Nova Scotia, Canada. Trunk kinematics and foot center of force were measured using the Lumbar Motion Monitor and F-Scan foot pressure system respectively. During motion conditions, significant decreases in trunk velocities were accompanied by significant increases in individual foot center of pressure velocities. These results suggest that during lifting and lowering loads in moving environments, the reaction to the wave-induced postural disturbance is accompanied by a decrease in performance speed so that the task can be performed more cautiously to optimize stability.
Scott Duncan, Kate White, Losi Sa’ulilo, and Grant Schofield
The aim of this study was to assess the convergent validity of a new piezoelectric pedometer and an omnidirectional accelerometer for assessing children’s time spent in moderate to vigorous physical activity (MVPA). A total of 114 children (51 boys, 63 girls) aged 5–11 years wore a sealed NL-1000 piezoelectric pedometer (New Lifestyles Inc, Lee’s Summit, MO) and an Actical accelerometer (Mini Mitter, Bend, OR) over one school day. The NL-1000 pedometers were randomized to one of two manual intensity thresholds used to define MVPA (1): Level 3 = 2.9 metabolic equivalent test (MET) and (2) Level 4 = 3.6 MET. Compared with the Actical, the NL-1000 underestimated the time spent in MVPA by 37% and 45% at intensity levels 3 and 4, respectively. In addition, the 95% limits of agreement were wide at both intensity levels (level 3 = -144%, 70%; level 4 = -135%, 45%), indicating a low level of precision.
Melody Oliver, Hannah Badland, Suzanne Mavoa, Mitch J. Duncan, and Scott Duncan
Global positioning systems (GPS), geographic information systems (GIS), and accelerometers are powerful tools to explain activity within a built environment, yet little integration of these tools has taken place. This study aimed to assess the feasibility of combining GPS, GIS, and accelerometry to understand transport-related physical activity (TPA) in adults.
Forty adults wore an accelerometer and portable GPS unit over 7 consecutive days and completed a demographics questionnaire and 7-day travel log. Accelerometer and GPS data were extracted for commutes to/from workplace and integrated into a GIS database. GIS maps were generated to visually explore physical activity intensity, GPS speeds and routes traveled.
GPS, accelerometer, and survey data were collected for 37 participants. Loss of GPS data was substantial due to a range of methodological issues, such as low battery life, signal drop out, and participant noncompliance. Nonetheless, greater travel distances and significantly higher speeds were observed for motorized trips when compared with TPA.
Pragmatic issues of using GPS monitoring to understand TPA behaviors and methodological recommendations for future research were identified. Although methodologically challenging, the combination of GPS monitoring, accelerometry and GIS technologies holds promise for understanding TPA within the built environment.
Scott Duncan, Kate White, Suzanne Mavoa, Tom Stewart, Erica Hinckson, and Grant Schofield
The distance between home and school is the most consistent predictor of active transport in youth: the closer an individual lives to school, the more likely they are to use active transport. While this suggests that it is preferable to live as close to school as possible, the limited physical activity accumulated during short trips may not offer substantial benefits to active transporters.
The current study investigated the predicted physical activity benefits associated with a range of home-school distances in 595 young people aged 5 to 16 years (Years 1 to 11). Physical activity was measured using sealed pedometers over 7 days. Participants’ home addresses and usual transport mode to and from school were collected via a questionnaire completed by parents (Years 1 to 6) and participants (Years 7 to 11).
A nonlinear relationship between predicted weekday activity and distance was detected, such that the high probability of active transport at short distances was offset by the low physical activity associated with walking short distances.
A distance of approximately 2 km was associated with the best physical activity outcomes related to active transport (9% to 15% increase on weekdays). These findings have potential implications for future interventions and for planning residential developments or facilities.
Alessandra Madia Mantovani, Scott Duncan, Jamile Sanches Codogno, Manoel Carlos Spiguel Lima, and Rômulo Araújo Fernandes
Physical activity level is an important tool to identify individuals predisposed to developing chronic diseases, which represent a major concern worldwide.
To identify correlates of daily step counts measured using pedometers, as well as analyze the associations between health outcomes and 3 different amounts of daily physical activity.
The sample comprised 278 participants (126 men and 153 women) with a mean age of 46.51 ± 9.02 years. Physical activity was assessed using pedometers for 7 consecutive days, and 3 amounts of daily physical activity were considered: ≥10,000 steps/day, ≥7500 steps/day, and <5000 steps/day. Sleep quality was assessed through a questionnaire, and dual-energy x-ray absorptiometry was used to measure body fat. Sociodemographic and anthropometric data were also collected.
The percentages of adults achieving at least 10,000 and 7500 steps/day on a minimum of 5 days of the evaluated week were 12.9% and 30.9%, respectively. Adults who reached ≥7500 steps/day had a lower likelihood of being obese (odds ratio [OR] = 0.38, 95% confidence interval [CI], 0.17–0.85) and reporting worse sleep quality (OR = 0.58, 95% CI, 0.34–0.99). Adults who reached <5000 steps/day had a higher likelihood of reporting worse sleep quality (OR = 2.11, 95% CI, 1.17–3.82).
Physical activity in adulthood, as measured by pedometer, constituted a behavior related to lower adiposity and better sleep quality.
Ralph Maddison, Samantha Marsh, Erica Hinckson, Scott Duncan, Sandra Mandic, Rachael Taylor, and Melody Smith
In this article, we report the grades for the second New Zealand Report Card on Physical Activity for Children and Youth, which represents a synthesis of available New Zealand evidence across 9 core indicators.
An expert panel of physical activity (PA) researchers collated and reviewed available nationally representative survey data between March and May 2016. In the absence of new data, (2014–2016) regional level data were used to inform the direction of existing grades. Grades were assigned based on the percentage of children and youth meeting each indicator: A is 81% to 100%; B is 61% to 80%; C is 41% to 60%, D is 21% to 40%; F is 0% to 20%; INC is Incomplete data.
Overall PA, Active Play, and Government Initiatives were graded B-; Community Environments was graded B; Sport Participation and School Environment received a C+; Sedentary Behaviors and Family/Peer Support were graded C; and Active Travel was graded C-.
Overall PA participation was satisfactory for young children but not for youth. The grade for PA decreased slightly from the 2014 report card; however, there was an improvement in grades for built and school environments, which may support regional and national-level initiatives for promoting PA.