Youth Oriented Activity Trackers: Comprehensive Laboratory- and Field-Based Validation.

BACKGROUND: Commercial activity trackers are growing in popularity among adults and some are beginning to be marketed to children. There is, however, a paucity of independent research examining the validity of these devices to detect physical activity of different intensity levels. OBJECTIVES: The purpose of this study was to determine the validity of the output from 3 commercial youth-oriented activity trackers in 3 phases: (1) orbital shaker, (2) structured indoor activities, and (3) 4 days of free-living activity. METHODS: Four units of each activity tracker (Movband [MB], Sqord [SQ], and Zamzee [ZZ]) were tested in an orbital shaker for 5-minutes at three frequencies (1.3, 1.9, and 2.5 Hz). Participants for Phase 2 (N=14) and Phase 3 (N=16) were 6-12 year old children (50% male). For Phase 2, participants completed 9 structured activities while wearing each tracker, the ActiGraph GT3X+ (AG) research accelerometer, and a portable indirect calorimetry system to assess energy expenditure (EE). For Phase 3, participants wore all 4 devices for 4 consecutive days. Correlation coefficients, linear models, and non-parametric statistics evaluated the criterion and construct validity of the activity tracker output. RESULTS: Output from all devices was significantly associated with oscillation frequency (r=.92-.99). During Phase 2, MB and ZZ only differentiated sedentary from light intensity (P<.01), whereas the SQ significantly differentiated among all intensity categories (all comparisons P<.01), similar to AG and EE. During Phase 3, AG counts were significantly associated with activity tracker output (r=.76, .86, and .59 for the MB, SQ, and ZZ, respectively). CONCLUSIONS: Across study phases, the SQ demonstrated stronger validity than the MB and ZZ. The validity of youth-oriented activity trackers may directly impact their effectiveness as behavior modification tools, demonstrating a need for more research on such devices.

Advantages and Limitations of Wearable Activity Trackers: Considerations for Patients and Clinicians.

BACKGROUND: Exercise, light physical activity, and decreased sedentary time all have been associated with health benefits following cancer diagnoses. Commercially available wearable activity trackers may help patients monitor and self-manage their behaviors to achieve these benefits. OBJECTIVES: This article highlights some advantages and limitations clinicians should be aware of when discussing the use of activity trackers with cancer survivors. METHODS: Limited research has assessed the accuracy of commercially available activity trackers compared to research-grade devices. Because most devices use confidential, proprietary algorithms to convert accelerometry data to meaningful output like total steps, assessing whether these algorithms account for differences in gait abnormalities, functional limitations, and different body morphologies can be difficult. Quantification of sedentary behaviors and light physical activities present additional challenges. FINDINGS: The global market for activity trackers is growing, which presents clinicians with a tremendous opportunity to incorporate these devices into clinical practice as tools to promote activity. This article highlights important considerations about tracker accuracy and usage by cancer survivors.

Energy Expenditure for 70 Activities in Children and Adolescents.

BACKGROUND: Thirty-five percent of the activities assigned MET values in the Compendium of Energy Expenditures for Youth were obtained from direct measurement of energy expenditure (EE). The aim of this study was to provide directly measured EE for several different activities in youth. METHODS: Resting metabolic rate (RMR) of 178 youths (80 females, 98 males) was first measured. Participants then performed structured activity bouts while wearing a portable metabolic system to directly measure EE. Steady-state oxygen consumption data were used to compute activity METstandard (activity VO2/3.5) and METmeasured (activity VO2/measured RMR) for the different activities. RESULTS: Rates of EE were measured for 70 different activities and ranged from 1.9 to 12.0 METstandard and 1.5 to 10.0 METmeasured. CONCLUSION: This study provides directly measured energy cost values for 70 activities in children and adolescents. It contributes empirical data to support the expansion of the Compendium of Energy Expenditures for Youth.

Utility of Consumer Physical Activity Trackers as an Intervention Tool in Cardiovascular Disease Prevention and Treatment.

Consumer activity trackers have grown in popularity over the last few years. These devices are typically worn on the hip or wrist and provide the user with information about physical activity measures such as steps taken, energy expenditure, and time spent in moderate to vigorous physical activity. The consumer may also use the computer interface (e.g. device websites, smartphone applications) to monitor and track achievement of PA goals and compete with other users. This review will describe some of the most popular consumer devices and discuss the user feedback tools. We will also present the limited evidence available about the accuracy of these devices and highlight how they have been used in cardiovascular disease management. We conclude with some recommendations for future research, focusing on how consumer devices might be used to assess effectiveness of various cardiovascular treatments.

Performance of Activity Classification Algorithms in Free-Living Older Adults.

PURPOSE: The objective of this study is to compare activity type classification rates of machine learning algorithms trained on laboratory versus free-living accelerometer data in older adults. METHODS: Thirty-five older adults (21 females and 14 males, 70.8 ± 4.9 yr) performed selected activities in the laboratory while wearing three ActiGraph GT3X+ activity monitors (in the dominant hip, wrist, and ankle; ActiGraph, LLC, Pensacola, FL). Monitors were initialized to collect raw acceleration data at a sampling rate of 80 Hz. Fifteen of the participants also wore GT3X+ in free-living settings and were directly observed for 2-3 h. Time- and frequency-domain features from acceleration signals of each monitor were used to train random forest (RF) and support vector machine (SVM) models to classify five activity types: sedentary, standing, household, locomotion, and recreational activities. All algorithms were trained on laboratory data (RFLab and SVMLab) and free-living data (RFFL and SVMFL) using 20-s signal sampling windows. Classification accuracy rates of both types of algorithms were tested on free-living data using a leave-one-out technique. RESULTS: Overall classification accuracy rates for the algorithms developed from laboratory data were between 49% (wrist) and 55% (ankle) for the SVMLab algorithms and 49% (wrist) to 54% (ankle) for the RFLab algorithms. The classification accuracy rates for SVMFL and RFFL algorithms ranged from 58% (wrist) to 69% (ankle) and from 61% (wrist) to 67% (ankle), respectively. CONCLUSIONS: Our algorithms developed on free-living accelerometer data were more accurate in classifying the activity type in free-living older adults than those on our algorithms developed on laboratory accelerometer data. Future studies should consider using free-living accelerometer data to train machine learning algorithms in older adults.

Validity of Activity Monitor Step Detection Is Related to Movement Patterns.

BACKGROUND: There is a need to examine step-counting accuracy of activity monitors during different types of movements. The purpose of this study was to compare activity monitor and manually counted steps during treadmill and simulated free-living activities and to compare the activity monitor steps to the StepWatch (SW) in a natural setting. METHODS: Fifteen participants performed laboratory-based treadmill (2.4, 4.8, 7.2 and 9.7 km/h) and simulated free-living activities (eg, cleaning room) while wearing an activPAL, Omron HJ720-ITC, Yamax Digi- Walker SW-200, 2 ActiGraph GT3Xs (1 in "low-frequency extension" [AGLFE] and 1 in "normal-frequency" mode), an ActiGraph 7164, and a SW. Participants also wore monitors for 1-day in their free-living environment. Linear mixed models identified differences between activity monitor steps and the criterion in the laboratory/free-living settings. RESULTS: Most monitors performed poorly during treadmill walking at 2.4 km/h. Cleaning a room had the largest errors of all simulated free-living activities. The accuracy was highest for forward/rhythmic movements for all monitors. In the free-living environment, the AGLFE had the largest discrepancy with the SW. CONCLUSION: This study highlights the need to verify step-counting accuracy of activity monitors with activities that include different movement types/directions. This is important to understand the origin of errors in step-counting during free-living conditions.

Methods to estimate aspects of physical activity and sedentary behavior from high-frequency wrist accelerometer measurements.

This investigation developed models to estimate aspects of physical activity and sedentary behavior from three-axis high-frequency wrist-worn accelerometer data. The models were developed and tested on 20 participants (n = 10 males, n = 10 females, mean age = 24.1, mean body mass index = 23.9), who wore an ActiGraph GT3X+ accelerometer on their dominant wrist and an ActiGraph GT3X on the hip while performing a variety of scripted activities. Energy expenditure was concurrently measured by a portable indirect calorimetry system. Those calibration data were then used to develop and assess both machine-learning and simpler models with fewer unknown parameters (linear regression and decision trees) to estimate metabolic equivalent scores (METs) and to classify activity intensity, sedentary time, and locomotion time. The wrist models, applied to 15-s windows, estimated METs [random forest: root mean squared error (rSME) = 1.21 METs, hip: rMSE = 1.67 METs] and activity intensity (random forest: 75% correct, hip: 60% correct) better than a previously developed model that used counts per minute measured at the hip. In a separate set of comparisons, the simpler decision trees classified activity intensity (random forest: 75% correct, tree: 74% correct), sedentary time (random forest: 96% correct, decision tree: 97% correct), and locomotion time (random forest: 99% correct, decision tree: 96% correct) nearly as well or better than the machine-learning approaches. Preliminary investigation of the models' performance on two free-living people suggests that they may work well outside of controlled conditions.

Validation of the Fitbit wireless activity tracker for prediction of energy expenditure.

OBJECTIVE: The purpose of this study was to examine the accuracy of the Fitbit wireless activity tracker in assessing energy expenditure (EE) for different activities. METHODS: Twenty participants (10 males, 10 females) wore the Fitbit Classic wireless activity tracker on the hip and the Oxycon Mobile portable metabolic system (criterion). Participants performed walking and running trials on a treadmill and a simulated free-living activity routine. Paired t tests were used to test for differences between estimated (Fitbit) and criterion (Oxycon) kcals for each of the activities. RESULTS: Mean bias for estimated energy expenditure for all activities was -4.5 ± 1.0 kcals/6 min (95% limits of agreement: -25.2 to 15.8 kcals/6 min). The Fitbit significantly underestimated EE for cycling, laundry, raking, treadmill (TM) 3 mph at 5% grade, ascent/descent stairs, and TM 4 mph at 5% grade, and significantly overestimated EE for carrying groceries. Energy expenditure estimated by the Fitbit was not significantly different than EE calculated from the Oxycon Mobile for 9 activities. CONCLUSION: The Fitbit worn on the hip significantly underestimates EE of activities. The variability in underestimation of EE for the different activities may be problematic for weight loss management applications since accurate EE estimates are important for tracking/monitoring energy deficit.

The independent and combined effects of exercise training and reducing sedentary behavior on cardiometabolic risk factors.

This pilot study examined if the combination of exercise training and reducing sedentary time (ST) results in greater changes to health markers than either intervention alone. Fifty-seven overweight/obese participants (19 males/39 females) (mean ± SD; age, 43.6 ± 9.9 years; body mass index (BMI), 35.1 ± 4.6 kg·m(-2)) completed the 12-week study and were randomly assigned to (i) EX: exercise 5 days·week(-1) for 40 min·session(-1) at moderate intensity; (ii) rST: reduce ST and increase nonexercise physical activity; (iii) EX-rST: combination of EX and rST; and (iv) CON: maintain behavior. Fasting lipids, blood pressure (BP), peak oxygen uptake, BMI, and 2-h oral glucose tolerance tests were completed pre- and post-intervention. EX and EX-rST increased peak oxygen uptake by ∼10% and decreased systolic BP (both p < 0.001). BMI decreased by -3.3% (95% confidence interval: -4.6% to -1.9%) for EX-rST and -2.2% (-3.5% to 0.0%) for EX. EX-rST significantly increased composite insulin-sensitivity index by 17.8% (2.8% to 32.8%) and decreased insulin area under the curve by 19.4% (-31.4% to -7.3%). No other groups improved in insulin action variables. rST group decreased ST by 7% (∼50 min·day(-1)); however, BP was the only health-related outcome that improved. EX and EX-rST improved peak oxygen uptake and BMI, providing further evidence that moderate-intensity exercise is beneficial. The within-group analysis provides preliminary evidence that exercising and reducing ST may result in improvements in metabolic biomarkers that are not seen with exercise alone, though between-group differences did not reach statistical significance. Future studies, with larger samples, should examine health-related outcomes resulting from greater reductions in ST over longer intervention periods.

Validation of a previous day recall for measuring the location and purpose of active and sedentary behaviors compared to direct observation.

PURPOSE: Gathering contextual information (i.e., location and purpose) about active and sedentary behaviors is an advantage of self-report tools such as previous day recalls (PDR). However, the validity of PDR's for measuring context has not been empirically tested. The purpose of this paper was to compare PDR estimates of location and purpose to direct observation (DO). METHODS: Fifteen adult (18-75 y) and 15 adolescent (12-17 y) participants were directly observed during at least one segment of the day (i.e., morning, afternoon or evening). Participants completed their normal daily routine while trained observers recorded the location (i.e., home, community, work/school), purpose (e.g., leisure, transportation) and whether the behavior was sedentary or active. The day following the observation, participants completed an unannounced PDR. Estimates of time in each context were compared between PDR and DO. Intra-class correlations (ICC), percent agreement and Kappa statistics were calculated. RESULTS: For adults, percent agreement was 85% or greater for each location and ICC values ranged from 0.71 to 0.96. The PDR-reported purpose of adults' behaviors were highly correlated with DO for household activities and work (ICCs of 0.84 and 0.88, respectively). Transportation was not significantly correlated with DO (ICC = -0.08). For adolescents, reported classification of activity location was 80.8% or greater. The ICCs for purpose of adolescents' behaviors ranged from 0.46 to 0.78. Participants were most accurate in classifying the location and purpose of the behaviors in which they spent the most time. CONCLUSIONS: This study suggests that adults and adolescents can accurately report where and why they spend time in behaviors using a PDR. This information on behavioral context is essential for translating the evidence for specific behavior-disease associations to health interventions and public policy.

ActiGraph™ activity monitors: "the firmware effect".

PURPOSE: The objective of this study is to examine the effect of different firmware versions on ActiGraph™ counts from the laboratory, field, and mechanical shaker testing. METHODS: Counts from 5 GT3X and 7 GT1M firmware versions were compared in this study. Monitors uploaded with these firmware versions were worn on the hip by 15 participants (age = 24.9 ± 5.0 yr, BMI= 23.9 ± 2.4 kg · m(-2)) who performed laboratory-based treadmill (walking: 1.5, 3.0 and 4.5 mph; running: 6 mph) and simulated free-living activities (sitting, self-paced walking, filing papers, dusting, vacuuming, and cleaning the room). Testing was also conducted during 1 d of free living and using orbital mechanical shaker testing at 0.7, 1.3, 2.0, and 3.0 Hz. Intermonitor comparisons for vertical, anteroposterior, mediolateral, and triaxial vector magnitude counts were conducted using one-way ANOVA and post hoc pairwise comparisons (P < 0.05). RESULTS: Vertical counts during treadmill walking at 1.5 mph from the GT1M monitor with firmware version 1.1.0. were significantly greater (P < 0.05; 75%) than output from the monitor with firmware version 1.3.0. Shaker testing revealed statistically significant differences in vertical and lateral counts. Although there were no significant differences among activity counts in the free-living comparisons, firmware version 1.1.0. produced the highest vertical counts during this protocol. CONCLUSION: Greater sensitivity of firmware version 1.1.0. to low-frequency sedentary activities resulted in greater counts than other firmware versions. It is recommended that before releasing new firmware, ActiGraph™ perform both human and mechanical shaker testing to verify comparability in outputs between new and previous firmware versions.