Comparison of three models of actigraph accelerometers during free living and controlled laboratory conditions

Abstract

The aim of this study was to compare the outputs of three commonly used uniaxial Actigraph models (Actitrainer, 7164 and GT1M) under both free-living and controlled laboratory conditions. Ten adults (mean age = 24.7±1.1 years) wore the three Actigraph models simultaneously during one of day free-living and during a progressive exercise protocol on a treadmill at speeds between 1.5 and 5.5 miles per hour (mph). During free-living the three Actigraph models produced comparable outputs in moderate, vigorous and moderate-to-vigorous physical activity (MVPA) with effect sizes typically <0.2, but lower comparability was seen in sedentary and light categories, as well as in total step counts (effect sizes often >0.30). In controlled conditions, acceptable comparability between the three models was seen at all treadmill speeds, the exception being walking at 1.5 mph (mean effect size = 0.48). It is concluded that care should be taken if different Actigraph models are to be used to measure and compare light physical activity, step counts and walking at very low speeds. However, using any of these three different Actigraph models to measure and compare levels of MVPA in free-living adults seems appropriate.     

Comparison of two Actigraph models for assessing free-living physical activity in Indian adolescents

This aim of this study was to compare the new Actigraph (GT1M) with the widely used Model 7164. Seven days of free-living physical activity were measured simultaneously using both the Model 7164 and GT1M in 30 Indian adolescents (mean age 15.8 years, s = 0.6). The GT1M was on average 9% lower per epoch than model 7164, thus a correction factor of 0.91 is suggested for comparison between the two monitors. The differences between monitors increased in magnitude with intensity of activity (P < 0.001) but remained randomly distributed (r = 0.01, P = 0.96). No significant difference was observed between monitors for time spent in moderate (P = 0.31) and vigorous (P = 0.34) physical activity when using the same epoch length. The Model 7164 classified less time as sedentary (P < 0.001) and more time as light-intensity activity (P < 0.001) than the GT1M. In conclusion, data from the GT1M can be compared with historical data using average counts per minute with a correction factor, and the two models might be comparable for assessing time spent in moderate to vigorous physical activity in children when using the same epoch length.     

Validity of the New Lifestyles NL-1000 Accelerometer for Measuring Time Spent in Moderate-to-Vigorous Physical Activity in School Settings

Current interest in promoting physical activity in the school environment necessitates an inexpensive, accurate method of measuring physical activity in such settings. Additionally, it is recognized that physical activity must be of at least moderate intensity in order to yield substantial health benefits. The purpose of the study, therefore, was to determine the validity of the New Lifestyles NL-1000 (New Lifestyles, Inc., Lee's Summit, Missouri, USA) accelerometer for measuring moderate-to-vigorous physical activity in school settings, using the Actigraph GT1M (ActiGraph, Pensacola, Florida, USA) as the criterion. Data were collected during a cross-country run (n = 12), physical education (n = 18), and classroom-based physical activities (n = 42). Significant and meaningful intraclass correlations between methods were found, and NL-1000 estimates of moderate-to-vigorous physical activity were not meaningfully different from GT1M-estimated moderate-to-vigorous physical activity. The NL-1000 therefore shows promising validity evidence as an inexpensive, convenient method of measuring moderate-to-vigorous physical activity in school settings.     

Cut points of the Actigraph GT9X for moderate and vigorous intensity physical activity at four different wear locations

ABSTRACT

Accelerometer cut points are an important consideration for distinguishing the intensity of activity into categories such as moderate and vigorous. It is well-established in the literature that these cut points depend on a variety of factors, including age group, device, and wear location. The Actigraph GT9X is a newer model accelerometer that is used for physical activity research, but existing cut points for this device are limited since it is a newer device. Furthermore, there is not existing data on cut points for the GT9X at the ankle or foot locations, which offers some potential benefit for activities that do not involve arm and/or core motion. A total of N = 44 adults completed a four-stage treadmill protocol while wearing Actigraph GT9X sensors at four different locations: foot, ankle, wrist, and hip. Metabolic Equivalent of Task (MET) levels assessed by indirect calorimetry along with Receiver Operating Characteristic (ROC) curves were used to establish cut points for moderate and vigorous intensity for each wear location of the GT9X. Area under the ROC curves indicated high discrimination accuracy for each case.     

Comparison of two Actigraph models for assessing free-living physical activity in Indian adolescents

Abstract

This aim of this study was to compare the new Actigraph (GT1M) with the widely used Model 7164. Seven days of free-living physical activity were measured simultaneously using both the Model 7164 and GT1M in 30 Indian adolescents (mean age 15.8 years, s = 0.6). The GT1M was on average 9% lower per epoch than model 7164, thus a correction factor of 0.91 is suggested for comparison between the two monitors. The differences between monitors increased in magnitude with intensity of activity (P < 0.001) but remained randomly distributed (r = 0.01, P = 0.96). No significant difference was observed between monitors for time spent in moderate (P = 0.31) and vigorous (P = 0.34) physical activity when using the same epoch length. The Model 7164 classified less time as sedentary (P < 0.001) and more time as light-intensity activity (P < 0.001) than the GT1M. In conclusion, data from the GT1M can be compared with historical data using average counts per minute with a correction factor, and the two models might be comparable for assessing time spent in moderate to vigorous physical activity in children when using the same epoch length.     

Adolescent physical activity levels: discrepancies with accelerometer data analysis

This study investigated the effects of epoch length and cut point selection on adolescent physical activity intensity quantification using vertical axis and vector magnitude (VM) measurement with the ActiGraph GT3X+ accelerometer. Four hundred and nine adolescents (211 males; 198 females) aged 12–16 years of age wore accelerometers during waking hours. The GT3X+ acceleration counts were reintegrated into 1, 5, 15, 30 and 60 s epoch lengths for both vertical axis and VM counts. One cut point was applied to vertical axis counts and three different cut points were applied to VM counts for each epoch length. Significant differences (P < 0.01) in mean total counts per day were observed between vertical axis and VM counts, and between epoch lengths for VM only. Differences in physical activity levels were observed between vertical and VM cut points, and between epoch lengths across all activity intensities. Our findings illustrate the magnitude of differences in physical activity outcomes that occur between axis measurement, cut points and epoch length. The magnitude of difference across epoch length must be considered in the interpretation of accelerometer data and seen as a confounding variable when comparing physical activity levels between studies.     

Comparing the Validity and Output of the GT1M and GT3X Accelerometer in 5- to 9-Year-Old Children

The purpose of this study was to compare the validity and output of the biaxial ActiGraph GT1M and the triaxial GT3X (ActiGraph, LLC, Pensacola, FL, USA) accelerometer in 5- to 9-year-old children. Thirty-two children wore the two monitors while their energy expenditure was measured with indirect calorimetry. They performed four locomotor and four play activities in an exercise laboratory and were further measured during 12 minutes of a sports lesson. Validity evidence in relation to indirect calorimetry was examined with linear regression equations applied to the laboratory data. During the sports lessons predicted energy expenditure according to the regression equations was compared to measured energy expenditure with the Wilcoxon-signed rank test and the Spearman correlation. To compare the output, agreement between counts of the two monitors during the laboratory activities was assessed with Bland-Altman plots. The evidence of validity was similar for both monitors. Agreement between the output of the two monitors was good for vertical counts (mean bias =??14 ± 22 counts) but not for horizontal counts (?17 ± 32 counts). The current results indicate that the two accelerometer models are able to estimate energy expenditure of a range of physical activities equally well in young children. However, they show output differences for movement in the horizontal direction.     

Accelerometer responsiveness to change between structured and unstructured physical activity in children and adolescents

This study examined if accelerometer-based assessments of physical activity were responsive to changes in physical activity level commensurate with performing structured versus unstructured activity in youth. Youth (6–16 years; N = 206) participated in a simulated after-school program that included structured and unstructured games on four occasions over a 3-year period. Recruitment occurred in 2007/2008 and data collection ended in 2011. Participants wore an Actigraph GT1M accelerometer on the hip. The Evenson cut-points were used to determine the time spent in each physical activity intensity, and standardized response means (SRM) were calculated and converted to standard effect sizes to be interpreted according to Cohen’s guidelines. SRMs ranged from trivial (0.16) to high (2.07), with the majority (75%) being classified as moderate or high. Our findings suggest that accelerometry was sensitive to differences in physical activity associated with structured compared to unstructured play, supporting the utility of accelerometry in evaluating activity-promoting interventions.     

Non-wear or sleep? Evaluation of five non-wear detection algorithms for raw accelerometer data

ABSTRACT

Detection of non-wear periods is an important step in accelerometer data processing. This study evaluated five non-wear detection algorithms for wrist accelerometer data and two rules for non-wear detection when non-wear and sleep algorithms are implemented in parallel. Non-wear algorithms were based on the standard deviation (SD), the high-pass filtered acceleration, or tilt angle. Rules for differentiating sleep from non-wear consisted of an override rule in which any overlap between non-wear and sleep was deemed non-wear; and a 75% rule in which non-wear periods were deemed sleep if the duration was < 75% of the sleep period. Non-wear algorithms were evaluated in 47 children who wore an ActiGraph GT3X+ accelerometer during school hours for 5 days. Rules for differentiating sleep from non-wear were evaluated in 15 adults who wore a GeneActiv Original accelerometer continuously for 24 hours. Classification accuracy for the non-wear algorithms ranged between 0.86–0.95, with the SD of the vector magnitude providing the best performance. The override rule misclassified 37.1 minutes of sleep as non-wear, while the 75% rule resulted in no misclassification. Non-wear algorithms based on the SD of the acceleration signal can effectively detect non-wear periods, while application of the 75% rule can effectively differentiate sleep from non-wear when examined concurrently.     

In-school and out-of-school physical activity in primary and secondary school children

The aim of this study was to compare in-school and out-of-school physical activity within a representative sample. Socio-demographic, physical activity, and anthropometric data were collected from a random sample of children (250 boys, 253 girls) aged 3-16 years attending nine primary and two secondary schools. Actigraph GT1M accelerometers, worn for seven days, were used to estimate physical activity levels for in-school (typically 09.00-15.00 h), out-of-school (weekday), and weekend periods. Physical activity as accelerometer counts per minute were lower in school versus out of school overall (in school: 437.2 +/- 172.9; out of school: 575.5 +/- 202.8; P < 0.001), especially in secondary school pupils (secondary: 321.6 +/- 127.5; primary: 579.2 +/- 216.3; P < 0.001). Minutes of moderate-to-vigorous physical activity accumulated in school accounted for 29.4 +/- 9.8% of total weekly moderate-to-vigorous physical activity overall but varied by sector (preschool: 37.4 +/- 6.2%; primary: 33.6 +/- 8.1%; secondary: 23.0 +/- 9.3%; F = 114.3, P < 0.001). Approximately half of the children with the lowest in-school activity compensated out of school during the week (47.4%) and about one-third at the weekend (30.0%). Overall, physical activity during the school day appears to be lower than that out of school, especially in secondary school children, who accumulate a lower proportion of their total weekly moderate-to-vigorous physical activity at school than younger children. As low in-school activity was compensated for beyond the school setting by less than half of children, promoting physical activity within the school day is important, especially in secondary schools.     

Metabolic Thresholds and Validated Accelerometer Cutoff Points for the Actigraph GT1M in Young Children Based on Measurements of Locomotion and Play Activities

The purpose of the current study was to determine metabolic thresholds and subsequent activity intensity cutoff points for the ActiGraph GT1M with various epochs spanning from 5 to 60 sec in young children. Twenty-two children, aged 4 to 9 years, performed 10 different activities including locomotion and play activities. Energy expenditure was measured with indirect calorimetry. Thresholds and cutoff points were determined through receiver operating characteristic curves. The lower metabolic threshold was 6.19 kcal·kg^?1·h^?1 for moderate and 9.28 kcal·kg^?1·h^?1 for vigorous intensity. The cutoff points for the GT1M accelerometer appear to be lower than those for the previous model (7164). For 5-sec epochs, a cutoff point of 143 counts resulted for moderate intensity and of 208 counts for vigorous intensity activity. Whether short or long epochs were chosen when collecting data to determine cutoff points, does not appear to have an influence on the resulting cutoff values. Similarly, comparable results are seen when analyses are based on locomotion only as opposed to a wide range of activities including children's play.     

The translation of preschoolers’ physical activity guidelines into a daily step count target

This study aimed at translating the physical activity (PA) guideline (180 min of total PA per day) into a step count target in preschoolers. 535 Flemish preschoolers (mean age: 4.41 ± 0.58) wore an ActiGraph accelerometer (GT1M, GT3X and GT3X+) – with activated step count function – for four consecutive days. The step count target was calculated from the accelerometer output using a regression equation, applying four different cut-points for light-to-vigorous PA: Pate, Evenson, Reilly, and Van Cauwenberghe. The present analysis showed that 180 min of total PA per day is equivalent to the following step count targets: 5,274 steps/day using the Pate cut-point, 4,653 steps/day using the Evenson cut-point, 11,379 steps/day using the Reilly cut-point and 13,326 steps/day using the Van Cauwenberghe cut-point. Future studies should focus on achieving consensus on which cut-points to use in preschoolers before a definite step count target in preschoolers can be proposed. Until then, we propose to use a provisional step count target of 11,500 steps/day as this step count target is attainable, realistic and helpful in promoting preschoolers’ PA.     

Calibration of context-specific survey items to assess youth physical activity behaviour

This study tests calibration models to re-scale context-specific physical activity (PA) items to accelerometer-derived PA. A total of 195 4th–12th grades children wore an Actigraph monitor and completed the Physical Activity Questionnaire (PAQ) one week later. The relative time spent in moderate-to-vigorous PA (MVPA_%) obtained from the Actigraph at recess, PE, lunch, after-school, evening and weekend was matched with a respective item score obtained from the PAQ’s. Item scores from 145 participants were calibrated against objective MVPA_% using multiple linear regression with age, and sex as additional predictors. Predicted minutes of MVPA for school, out-of-school and total week were tested in the remaining sample (n = 50) using equivalence testing. The results showed that PAQ β-weights ranged from 0.06 (lunch) to 4.94 (PE) MVPA_% (P < 0.05) and models root mean square error ranged from 4.2% (evening) to 20.2% (recess). When applied to an independent sample, differences between PAQ and accelerometer MVPA at school and out-of-school ranged from ?15.6 to +3.8 min and the PAQ was within 10–15% of accelerometer measured activity. This study demonstrated that context-specific items can be calibrated to predict minutes of MVPA in groups of youth during in- and out-of-school periods.     

Comparison of three generations of ActiGraph™ activity monitors in children and adolescents

Abstract

In this study, we evaluated agreement among three generations of ActiGraph? accelerometers in children and adolescents. Twenty-nine participants (mean age = 14.2 ± 3.0 years) completed two laboratory-based activity sessions, each lasting 60 min. During each session, participants concurrently wore three different models of the ActiGraph? accelerometers (GT1M, GT3X, GT3X+). Agreement among the three models for vertical axis counts, vector magnitude counts, and time spent in moderate-to-vigorous physical exercise (MVPA) was evaluated by calculating intraclass correlation coefficients and Bland-Altman plots. The intraclass correlation coefficient for total vertical axis counts, total vector magnitude counts, and estimated MVPA was 0.994 (95% CI = 0.989–0.996), 0.981 (95% CI = 0.969–0.989), and 0.996 (95% CI = 0.989–0.998), respectively. Inter-monitor differences for total vertical axis and vector magnitude counts ranged from 0.3% to 1.5%, while inter-monitor differences for estimated MVPA were equal to or close to zero. On the basis of these findings, we conclude that there is strong agreement between the GT1M, GT3X, and GT3X+ activity monitors, thus making it acceptable for researchers and practitioners to use different ActiGraph? models within a given study.     

Validity of the ActiGraph GT1M during walking and cycling

Abstract

The ActiGraph activity monitors have developed and newer versions of the ActiGraph accelerometers (GT1M, GT3X and GT3X +) are now available, including changes in hardware and software compared to the old version (AM7164). This is problematic as most of the validation and calibration work includes the AM7164. The aims of the study were to validate the ActiGraph GT1M during level and graded walking and to assess the potential underestimation of physical activity during cycling. Data were obtained from 20 participants during treadmill walking and ergometer cycling. Energy expenditure was measured via indirect calorimetry and used as the criterion method. Activity counts were highly correlated with energy expenditure during level walking (R² = 0.82) and graded walking at 5% and 8% (R² = 0.82 and R² = 0.67, respectively). There was no linear relationship between activity counts and energy expenditure during cycling. The average activity counts for all data points during cycling was 1,157 counts per minute (CPM) (SD = 974), and mean energy expenditure was 5.0 metabolic equivalents. The GT1M is a valid tool for assessing walking across a wide range of speeds and gradients. However, there is no relationship between activity counts and energy expenditure during cycling and physical activity is underestimated by ≈73% during cycling compared to walking.     

Validation of MET estimates and step measurement using the ActivPAL physical activity logger

The purpose of this study was to determine the validity of the metabolic equivalent (MET) equation and step rate function of the ActivPAL? physical activity logger in a group of females. Using a standard treadmill protocol, 62 females aged 15-25 years walked on a treadmill at speeds between 3.2 and 7.0 km · h(-1) while wearing an ActivPAL. Oxygen consumption was measured using expired gas analysis at each speed and METs for each speed were estimated based on each participant's own resting metabolic rate. A sub-set of 18 participants also wore an Actigraph. Results showed that the in-built equation in the ActivPAL significantly underestimated (P < 0.001) METs under treadmill conditions at higher intensities. The ActivPAL equation is based on step rate yet the relationship between counts and measured METs (r = 0.76; P < 0.001) is stronger than that between steps and measured METs (r = 0.59; P < 0.001). Both the ActivPAL and Actigraph step functions showed no significant difference (P > 0.05) to video recorded step rate except at the slowest walking speed where the Actigraph significantly underestimated steps (P < 0.05). The development of a new equation based on the counts-METs relationship that includes a variety of speeds and activities would be useful. The ActivPAL step function performs better than the Actigraph at the slowest walking speed under treadmill conditions.     

Concurrent validation of the Actigraph gt3x+, Polar Active accelerometer,Omron HJ-720 and Yamax Digiwalker SW-701 pedometer step counts in lab-based and free-living settings

Activity monitors are frequently used to assess activity in many settings. But as technology advances, so do the mechanisms used to estimate activity causing a continuous need to validate newly developed monitors. The purpose of this study was to examine the step count validity of the Yamax Digiwalker SW-701 pedometer (YX), Omron HJ-720 T pedometer (OP), Polar Active accelerometer (PAC) and Actigraph gt3x+ accelerometer (AG) under controlled and free-living conditions. Participants completed five stages of treadmill walking (n = 43) and a subset of these completed a 3-day free-living wear period (n = 37). Manually counted (MC) steps provided a criterion measure for treadmill walking, whereas the comparative measure during free-living was the YX. During treadmill walking, the OP was the most accurate monitor across all speeds (±1.1% of MC steps), while the PAC underestimated steps by 6.7–16.0% per stage. During free-living, the OP and AG counted 97.5% and 98.5% of YX steps, respectively. The PAC overestimated steps by 44.0%, or 5,265 steps per day. The Omron pedometer seems to provide the most reliable and valid estimate of steps taken, as it was the best performer under lab-based conditions and provided comparable results to the YX in free-living. Future studies should consider these monitors in additional populations and settings.     

GT3X+ accelerometer placement affects the reliability of step-counts measured during running and pedal-revolution counts measured during bicycling

Accelerometers provide a measure of step-count. Reliability and validity of step-count and pedal-revolution count measurements by the GT3X+ accelerometer, placed at different anatomical locations, is absent in the literature. The purpose of this study was to investigate the reliability and validity of step and pedal-revolution counts produced by the GT3X+ placed at different anatomical locations during running and bicycling.

Twenty-two healthy adults (14 men and 8 women) completed running and bicycling activity bouts (5 minutes each) while wearing 6 accelerometers: 2 each at the waist, thigh and shank. Accelerometer and video data were collected during activity.

Excellent reliability and validity were found for measurements taken from accelerometers mounted at the waist and shank during running (Reliability: intraclass correlation (ICC) ≥ 0.99; standard error of measurement (SEM) ≤1.0 steps; Validity: Pearson ≥ 0.99) and at the thigh and shank during bicycling (Reliability: ICC ≥ 0.99; SEM ≤1.0 revolutions; Validity: Pearson ≥ 0.99). Excellent reliability was found between measurements taken at the waist and shank during running (ICC ≥ 0.98; SEM ≤1.6 steps) and between measurements taken at the thigh and shank during bicycling (ICC ≥ 0.99; SEM ≤1.0 revolutions). These data suggest that the GT3X+ can be used for measuring step-count during running and pedal-revolution count during bicycling. Only shank placement is recommended for both activities.