Impact of ActiGraph Sampling Rate and Intermonitor Comparability on Measures of Physical Activity in Adults

Clevenger, Kimberly A.; Brønd, Jan Christian; Arvidsson, Daniel; Montoye, Alexander H.K.; Mackintosh, Kelly A.; McNarry, Melitta A.; Pfeiffer, Karin A.

doi:10.1123/jmpb.2021-0016

Cited by 6 publications

(11 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Only the prior study in free-living adults explored the impact of using a 100 versus 30 Hz sampling rate on the MAD metric, finding a mean absolute difference of 2.2 mg (5.8%), which was larger than the present study (0.7 mg; 1.8%) (Clevenger et al 2021). Smaller differences in the present study compared to prior research may be due to our focus on a pediatric sample instead of adults.…”

Section: Confidence Interval For Equivalence Testcontrasting

confidence: 89%

“…counts•min −1 . These smaller differences are likely due to our focus on children instead of adults (Clevenger et al 2021) or free-living behavior instead of laboratory-based activities, which often consist of a greater proportion of higher-intensity activities (Clevenger et al 2019). As the primary issue with using a sampling rate other than 30 Hz (or a multiple thereof) is that high frequency signals bypass ActiGraph's filter, this seems to have occurred to a lesser degree in this sample and/or setting.…”

Section: Confidence Interval For Equivalence Testmentioning

confidence: 93%

“…Monitors were initialized to collect raw acceleration data at a sampling rate of 100 Hz and data were downloaded as .gt3x files using ActiLife software (version 6.13.4; firmware 1.2.0). In accord with previous research (Brønd et al 2017, Clevenger et al 2021, the 100 Hz data were resampled to 30, 60, 80, and 90 Hz using Java software (Oracle Corp., Redwood Shores, CA) and the resample function available in MATLAB (MathWorks Inc., Natwick, MA) which uses a polyphase antialiasing filter. These sampling rates were chosen because of their frequent use in the literature (particularly for 30 and 100 Hz; Migueles et al 2017), use in NHANES (80 Hz; Troiano et al 2014), and Brønd and Arvidsson's (2016) prior supposition that counts are unaffected by sampling rates that are multiples of 30 (i.e.…”

Section: Methods Data Collection and Processingmentioning

confidence: 99%

“…If the confidence interval did not overlap or exceed the equivalence bounds, then the monitors were considered equivalent (p<0.05). Equivalence bounds were set as 5% of the mean value for each variable except MVPA, for which the equivalence bounds were modified 0.5 percentage points, as using the 5% of the mean criterion resulted in extremely narrow bounds that have little practical meaning (Clevenger et al 2020(Clevenger et al , 2021. Equivalence bounds are reported in supplementary table 1.…”

Section: Methods Data Collection and Processingmentioning

confidence: 99%

See 3 more Smart Citations

Impact of ActiGraph sampling rate on free-living physical activity measurement in youth

Clevenger¹,

Brønd²,

Mackintosh³

et al. 2022

Physiol. Meas.

View full text Add to dashboard Cite

ActiGraph sampling frequencies of more than 30 Hz may result in overestimation of activity counts in both children and adults, but research on free-living individuals has not included the range of sampling frequencies used by researchers. Objective: We compared count- and raw-acceleration-based metrics from free-living children and adolescents across a range of sampling frequencies. Approach: Participants (n=445; 10-15 y) wore an ActiGraph accelerometer for at least one 10-h day. Vector Magnitude counts, Mean Amplitude Deviation, Monitor-Independent Movement Summary units, and activity intensity classified using six methods (four cut-points, two-regression model, and artificial neural network) were compared between 30 Hz and 60, 80, 90, and 100 Hz sampling frequencies using mean absolute differences, correlations, and equivalence testing. Main results: All outcomes were statistically equivalent, and correlation coefficients were ≥0.970. Absolute differences were largest for the 30 vs. 80 and 30 vs. 100 Hz count comparisons. For comparisons of 30 with 60, 80, 90, or 100 Hz, mean (and maximum) absolute differences in minutes of moderate-to-vigorous physical activity per day ranged from 0.1 to 0.3 (0.4 to 1.5), 0.3 to 1.3 (1.6 to 8.6), 0.1 to 0.3 (1.1 to 2.5), and 0.3 to 2.5 (1.6 to 14.3) across the six classification methods. Significance: Acceleration-based outcomes are comparable across the full range of sampling rates and therefore recommended for future research. If using counts, we recommend a multiple of 30 Hz because using a 100 Hz sampling rate resulted in large maximum individual differences and epoch-level differences, and increasing differences with activity level.

show abstract

Section: Confidence Interval For Equivalence Testcontrasting

confidence: 89%

Section: Confidence Interval For Equivalence Testmentioning

confidence: 93%

Section: Methods Data Collection and Processingmentioning

confidence: 99%

Section: Methods Data Collection and Processingmentioning

confidence: 99%

See 2 more Smart Citations

Impact of ActiGraph sampling rate on free-living physical activity measurement in youth

Clevenger¹,

Brønd²,

Mackintosh³

et al. 2022

Physiol. Meas.

View full text Add to dashboard Cite

show abstract

“…As a result, the volume of VPA might also be underestimated. However, MAD algorithm has demonstrated a high comparability for the ActiGraph counts [ 50 ] and has shown a very strong association with VO 2 during ambulatory activities [ 33 ]; moreover, similar problems with the measurement of VPA have been highlighted in other studies that used other algorithms for accelerometer data [ 51 , 52 ].…”

Section: Strengths and Limitationsmentioning

confidence: 99%

Reliability and Validity of Slovenian Versions of IPAQ-SF, GPAQ, and EHIS-PAQ for Assessing Physical Activity and Sedentarism of Adults

Meh

Sember

Đurić

et al. 2021

IJERPH

View full text Add to dashboard Cite

Health policies rely on physical activity (PA) and sedentary behavior data collected through PA questionnaires (PAQs). Validity of international PAQs varies among countries. Therefore, it is important to know the validity of the national versions of the PAQs to properly evaluate the results. We conducted a validation study of the Slovenian versions of the International PAQ Short Form (IPAQ-SF), the Global PAQ (GPAQ), and the PAQ used in the European Health Interview Survey (EHIS-PAQ) on 306 healthy adults. The most valid and reliable constructs in all tested were sedentary behavior and vigorous PA (VPA), however the criterion validity of these constructs was low (Spearman’s ρ 0.38–0.45 for sedentary behavior and 0.34–0.42 for VPA). Moderate to vigorous PA (MVPA) had low validity (0.26–0.29) despite being used as a standard measure of PA behavior. Participants over-reported MVPA for 17 to 156 min and underreported the sedentary behavior for more than two hours. The test-retest study found high reliability for sedentary behavior (0.69–0.81) and low to moderate reliability for PA behavior (0.42–0.76). The Slovenian versions of the observed PAQs are a useful tool for national PA surveillance, but for qualitative assessment of individual health-related PA behavior they should be combined with accelerometer-based devices.

show abstract

Wrist‐worn accelerometers: Influence of decisions during data collection and processing: A cross‐sectional study

Eke,

Bonn,

Trolle Lagerros

2024

Health Science Reports

View full text Add to dashboard Cite

Background and AimsAccelerometers collect data in an objective way, however, a number of decisions must be done during data collection, processing and output‐interpretation. The influence of those decisions is seldom investigated, reported, or discussed. Herein, we examined the influence of different decisions on the outcomes: daily minutes of moderate‐to‐vigorous physical activity (MVPA), inactivity and light physical activity (LPA).MethodsIn total, 156 participants wore an accelerometer (ActiGraph wGT3X‐BT) on their nondominant wrist for 7 days. Data collection was conducted from February 2017 to June 2018. Data was processed using the R‐package GGIR and default settings were compared to by‐the‐literature‐suggested options. The output was examined using paired t‐tests.ResultsWhen comparing two commonly used MVPA‐cut‐points, default and Hildebrand et al. we found a marginal difference (0.4 min, 1.0%, p < 0.001) in MVPA/day. When no bout criteria for MVPA/day was applied, MVPA/day was twice as high as bouted MVPA/day. Further, when we changed the epoch‐length from 5 to 1 s, statistically significant changes were seen for MVPA/day (−6.6 min, 19%, p < 0.001), inactivity/day (−22 min, 3.0%, p < 0.001) and LPA/day (28 min, 81%, p < 0.001).ConclusionDecisions made during data processing of wrist‐worn accelerometers has an influence on the output and thus, may influence the conclusions drawn. However, there may be situations when these settings are changed. If so, we recommend examining if the variables of interest are affected. We encourage researchers to report decisions made during data collection, processing and output‐interpretation, to facilitate comparisons between different studies.

show abstract

Impact of ActiGraph Sampling Rate and Intermonitor Comparability on Measures of Physical Activity in Adults

Cited by 6 publications

References 45 publications

Impact of ActiGraph sampling rate on free-living physical activity measurement in youth

Impact of ActiGraph sampling rate on free-living physical activity measurement in youth

Reliability and Validity of Slovenian Versions of IPAQ-SF, GPAQ, and EHIS-PAQ for Assessing Physical Activity and Sedentarism of Adults

Wrist‐worn accelerometers: Influence of decisions during data collection and processing: A cross‐sectional study

Contact Info

Product

Resources

About