Dataset Information

Estimating physical activity in youth using a wrist accelerometer.

ABSTRACT:

Purpose

The purpose of this study was to develop and validate methods for analyzing wrist accelerometer data in youth.

Methods

A total of 181 youth (mean ± SD; age, 12.0 ± 1.5 yr) completed 30 min of supine rest and 8 min each of 2 to 7 structured activities, selected from a list of 25. Receiver operating characteristic (ROC) curves and regression analyses were used to develop prediction equations for energy expenditure (child-METs; measured activity V˙O2 divided by measured resting V˙O2) and cut points for computing time spent in sedentary behaviors (SB), light (LPA), moderate (MPA), and vigorous (VPA) physical activity. Both vertical axis (VA) and vector magnitude (VM) counts per 5 s were used for this purpose. The validation study included 42 youth (age, 12.6 ± 0.8 yr) who completed approximately 2 h of unstructured PA. During all measurements, activity data were collected using an ActiGraph GT3X or GT3X+, positioned on the dominant wrist. Oxygen consumption was measured using a Cosmed K4b. Repeated-measures ANOVA were used to compare measured versus predicted child-METs (regression only) and time spent in SB, LPA, MPA, and VPA.

Results

All ROC cut points were similar for area under the curve (≥0.825), sensitivity (≥0.756), and specificity (≥0.634), and they significantly underestimated LPA and overestimated VPA (P < 0.05). The VA and VM regression models were within ±0.21 child-METs of mean measured child-METs and ±2.5 min of measured time spent in SB, LPA, MPA, and VPA, respectively (P > 0.05).

Conclusions

Compared to measured values, the VA and VM regression models developed on wrist accelerometer data had insignificant mean bias for child-METs and time spent in SB, LPA, MPA, and VPA; however, they had large individual errors.

SUBMITTER: Crouter SE

PROVIDER: S-EPMC4362848 | biostudies-literature | 2015 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Estimating physical activity in youth using a wrist accelerometer.

Crouter Scott E SE Flynn Jennifer I JI Bassett David R DR

Medicine and science in sports and exercise 20150501 5

<h4>Purpose</h4>The purpose of this study was to develop and validate methods for analyzing wrist accelerometer data in youth.<h4>Methods</h4>A total of 181 youth (mean ± SD; age, 12.0 ± 1.5 yr) completed 30 min of supine rest and 8 min each of 2 to 7 structured activities, selected from a list of 25. Receiver operating characteristic (ROC) curves and regression analyses were used to develop prediction equations for energy expenditure (child-METs; measured activity V˙O2 divided by measured resti ...[more]

PMID: 25207928

Similar Datasets

Project description:PurposeThis study aimed to determine the validity of existing methods to estimate sedentary behavior (SB) under free-living conditions using ActiGraph GT3X+ accelerometers (AG).MethodsForty-eight young (18-25 yr) adults wore an AG on the right hip and nondominant wrist and were video recorded during four 1-h sessions in free-living settings (home, community, school, and exercise). Direct observation videos were coded for postural orientation, activity type (e.g., walking), and METs derived from the Compendium of Physical Activities, which served as the criterion measure of SB (sitting or lying posture, <1.5 METs). Thirteen methods using cut points from vertical counts per minute (CPM), counts per 15-s (CP15s), and vector magnitude (VM) counts (e.g., CPM1853VM), raw acceleration and arm angle (sedentary sphere), Euclidean norm minus one (ENMO) corrected for gravity (mg) thresholds, uni- or triaxial sojourn hybrid machine learning models (Soj1x and Soj3x), random forest (RF), and decision tree (TR) models were used to estimate SB minutes from AG data. Method bias, mean absolute percent error, and their 95% confidence intervals were estimated using repeated-measures linear mixed models.ResultsOn average, participants spent 34.1 min per session in SB. CPM100, CPM150, Soj1x, and Soj3x were the only methods to accurately estimate SB from the hip. Sedentary sphere and ENMO44.8 overestimated SB by 3.9 and 6.1 min, respectively, whereas the remaining wrist methods underestimated SB (range, 9.5-2.5 min). In general, mean absolute percent error was lower using hip methods compared with wrist methods.ConclusionAccurate group-level estimates of SB from a hip-worn AG can be achieved using either simpler count-based approaches (CPM100 and CPM150) or machine learning models (Soj1x and Soj3x). Wrist methods did not provide accurate or precise estimates of SB. The development of large open-source free-living calibration data sets may lead to improvements in SB estimates.

Project description:Over the last decade use of raw acceleration metrics to assess physical activity has increased. Metrics such as Euclidean Norm Minus One (ENMO), and Mean Amplitude Deviation (MAD) can be used to generate metrics which describe physical activity volume (average acceleration), intensity distribution (intensity gradient), and intensity of the most active periods (MX metrics) of the day. Presently, relatively little comparative data for these metrics exists in youth. To address this need, this study presents age- and sex-specific reference percentile values in England youth and compares physical activity volume and intensity profiles by age and sex. Wrist-worn accelerometer data from 10 studies involving youth aged 5 to 15 y were pooled. Weekday and weekend waking hours were first calculated for youth in school Years (Y) 1&2, Y4&5, Y6&7, and Y8&9 to determine waking hours durations by age-groups and day types. A valid waking hours day was defined as accelerometer wear for ≥ 600 min·d-1 and participants with ≥ 3 valid weekdays and ≥ 1 valid weekend day were included. Mean ENMO- and MAD-generated average acceleration, intensity gradient, and MX metrics were calculated and summarised as weighted week averages. Sex-specific smoothed percentile curves were generated for each metric using Generalized Additive Models for Location Scale and Shape. Linear mixed models examined age and sex differences. The analytical sample included 1250 participants. Physical activity peaked between ages 6.5-10.5 y, depending on metric. For all metrics the highest activity levels occurred in less active participants (3rd-50th percentile) and girls, 0.5 to 1.5 y earlier than more active peers, and boys, respectively. Irrespective of metric, boys were more active than girls (p < .001) and physical activity was lowest in the Y8&9 group, particularly when compared to the Y1&2 group (p < .001). Percentile reference values for average acceleration, intensity gradient, and MX metrics have utility in describing age- and sex-specific values for physical activity volume and intensity in youth. There is a need to generate nationally-representative wrist-acceleration population-referenced norms for these metrics to further facilitate health-related physical activity research and promotion.

Project description:BackgroundThe rapid advancements in science and technology of wrist-wearable activity devices offer considerable potential for clinical applications. Self-monitoring of physical activity (PA) with activity devices is helpful to improve the PA levels of adolescents. However, knowing the accuracy of activity devices in adolescents is necessary to identify current levels of PA and assess the effectiveness of intervention programs designed to increase PA.ObjectiveThe study aimed to determine the validity of the 11 commercially available wrist-wearable activity devices for monitoring total steps and total 24-hour total energy expenditure (TEE) in healthy adolescents under simulated free-living conditions.MethodsNineteen (10 male and 9 female) participants aged 14 to 18 years performed a 24-hour activity cycle in a metabolic chamber. Each participant simultaneously wore 11 commercial wrist-wearable activity devices (Mi Band 2 [XiaoMi], B2 [Huawei], Bong 2s [Meizu], Amazfit [Huamei], Flex [Fitbit], UP3 [Jawbone], Shine 2 [Misfit], GOLiFE Care-X [GoYourLife], Pulse O2 [Withings], Vivofit [Garmin], and Loop [Polar Electro]) and one research-based triaxial accelerometer (GT3X+ [ActiGraph]). Criterion measures were total EE from the metabolic chamber (mcTEE) and total steps from the GT3X+ (AGsteps).ResultsPearson correlation coefficients r for 24-hour TEE ranged from .78 (Shine 2, Amazfit) to .96 (Loop) and for steps ranged from 0.20 (GOLiFE) to 0.57 (Vivofit). Mean absolute percent error (MAPE) for TEE ranged from 5.7% (Mi Band 2) to 26.4% (Amazfit) and for steps ranged from 14.2% (Bong 2s) to 27.6% (Loop). TEE estimates from the Mi Band 2, UP3, Vivofit, and Bong 2s were equivalent to mcTEE. Total steps from the Bong 2s were equivalent to AGsteps.ConclusionsOverall, the Bong 2s had the best accuracy for estimating TEE and total steps under simulated free-living conditions. Further research is needed to examine the validity of these devices in different types of physical activities under real-world conditions.

Project description:PurposeThis study aimed to examine the validity of wrist acceleration cut points for classifying moderate (MPA), vigorous (VPA), and moderate-to-vigorous (MVPA) physical activity.MethodsFifty-seven children (5-12 yr) completed 15 semistructured activities. Three sets of wrist cut points (>192 mg, >250 mg, and >314 mg), previously developed using Euclidian norm minus one (ENMO192+), GENEActiv software (GENEA250+), and band-pass filter followed by Euclidian norm (BFEN314+), were evaluated against indirect calorimetry. Analyses included classification accuracy, equivalence testing, and Bland-Altman procedures.ResultsAll cut points classified MPA, VPA, and MVPA with substantial accuracy (ENMO192+: κ = 0.72 [95% confidence interval = 0.72-0.73], MVPA: area under the receiver operating characteristic curve (ROC-AUC) = 0.85 [0.85-0.86]; GENEA250+: κ = 0.75 [0.74-0.76], MVPA: ROC-AUC = 0.85 [0.85-0.86]; BFEN314+: κ = 0.73 [0.72-0.74], MVPA: ROC-AUC = 0.86 [0.86-0.87]). BFEN314+ misclassified 19.7% non-MVPA epochs as MPA, whereas ENMO192+ and GENEA250+ misclassified 32.6% and 26.5% of MPA epochs as non-MVPA, respectively. Group estimates of MPA time were equivalent (P < 0.01) to indirect calorimetry for the BFEN314+ MPA cut point (mean bias = -1.5%, limits of agreement [LoA] = -57.5% to 60.6%), whereas estimates of MVPA time were equivalent (P < 0.01) to indirect calorimetry for the ENMO192+ (mean bias = -1.1%, LoA = -53.7% to 55.9%) and GENEA250+ (mean bias = 2.2%, LoA = -56.5% to 52.2%) cut points. Individual variability (LoA) was large for MPA (min: BFEN314+, -60.6% to 57.5%; max: GENEA250+, -42.0% to 104.1%), VPA (min: BFEN314+, -238.9% to 54.6%; max: ENMO192+, -244.5% to 127.4%), and MVPA (min: ENMO192+, -53.7% to 55.0%; max: BFEN314+, -83.9% to 25.3%).ConclusionWrist acceleration cut points misclassified a considerable proportion of non-MVPA and MVPA. Group-level estimates of MVPA were acceptable; however, error for individual-level prediction was larger.

Dataset Information

Estimating physical activity in youth using a wrist accelerometer.

Purpose

Methods

Results

Conclusions

Publications

Estimating physical activity in youth using a wrist accelerometer.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets