Browse

Purpose: To assess the interdevice reliability and validity of the repetition-counting feature of the Push Band 2.0. Methods: Thirty college-aged participants (aged 18–24 years) simultaneously wore two Push Band 2.0 devices and performed 10 common resistance training exercises at four different tempos over the course of two testing sessions. Twelve repetitions were completed with visual confirmation for each set of exercises and compared with repetition estimates from the Push Band 2.0. Interdevice reliability was quantified using single measures intraclass correlation coefficients with 95% confidence intervals while validity was assessed via mean absolute percent error and mean percent error. Results: Interdevice reliability was found to be good to very good regardless of exercise type or tempo, as all intraclass correlation coefficients were >.770. Validity of the repetition-counting feature of the device was dependent on both exercise type and tempo, as exercises that did not involve rotation of the device throughout the movement demonstrated greater mean absolute percent error (31.0% average of all four tempos) and mean percent error (−29.9% average of all four tempos) than those that required such rotation (average mean absolute percent error of 3.5% and mean percent error of −1.6% across all four tempos). Conclusions: This study supports the reliability of the repetition-counting feature of the Push Band 2.0. However, device accuracy appears to be dependent on the type of movement and the speed at which the movement is performed, with greater accuracy observed during faster exercise tempos and exercises involving rotation of the device during movement execution.

Accelerometers are increasingly used to measure 24-hr movement behaviors but are sometimes removed intermittently (e.g., for sleep or bathing), resulting in missing data. This study compared physical behaviors between times a hip-placed accelerometer was worn versus not worn in a college student sample. Participants (n = 115) wore a hip-placed ActiGraph during waking times and a thigh-placed activPAL continuously for at least 7 days (mean ± SD 7.5 ± 1.1 days). Thirteen nonwear algorithms determined ActiGraph nonwear; days included in the analysis had to have at least 1 min where the ActiGraph classified nonwear while participant was classified as awake by the activPAL. activPAL data for steps, time in sedentary behaviors (SB), light-intensity physical activity (LPA), and moderate- to vigorous-intensity physical activity (MVPA) from ActiGraph wear times were then compared with activPAL data from ActiGraph nonwear times. Participants took more steps (10.2–11.8 steps/min) and had higher proportions of MVPA (5.0%–5.9%) during ActiGraph wear time than nonwear time (3.1–8.0 steps/min, 0.8%–1.3% in MVPA). Effects were variable for SB (62.6%–66.9% of wear, 45.5%–76.2% of nonwear) and LPA (28.2%–31.5% of wear, 23.0%–53.2% of nonwear) depending on nonwear algorithm. Rescaling to a 12-hr day reduced SB and LPA error but increased MVPA error. Requiring minimum wear time (e.g., 600 min/day) reduced error but resulted in 10%–22% of days removed as invalid. In conclusion, missing data had minimal effect on MVPA but resulted in underestimation of SB and LPA. Strategies like scaling SB and LPA, but not MVPA, may improve physical behavior estimates from incomplete accelerometer data.

The eighth wave of the Survey of Health, Aging and Retirement in Europe comprises a subsample of respondents who were asked to participate in a measurement of physical activity using thigh-worn accelerometers. This paper describes the process for obtaining consent, identifies determinants of consent, and analyzes the aggregated results of the accelerometer measurements for bias due to sample selection. The overall consent rate in the Survey of Health, Aging and Retirement in Europe accelerometer study was 54%, with variations between countries ranging from 34% to 70%. Multivariate logistic regressions show that various factors are correlated with consent such as respondents’ age, self-reported moderate activity, computer literacy, willingness to answer questions, and the interviewers’ age. After introducing inverse probability weights, there appears to be only a small and insignificant influence of participant selection and consent.

Measurement of 24-hr movement behaviors is important for assessing adherence to guidelines, participation trends over time, group differences, and whether health-promoting interventions are successful. For a measurement tool to be useful, it must be valid, reliable, and able to detect change, the latter being a measurement property called responsiveness, sensitivity to change, or longitudinal validity. We systematically reviewed literature on the responsiveness of accelerometers to detect change in 24-hr movement behaviors. Databases (PubMed, Scopus, and EBSCOHost) were searched for peer-reviewed papers published in English between 1998 and 2023. Quality/risk of bias was assessed using a customized tool. This study is registered at https://osf.io/qrn8a. Twenty-six papers met the inclusion/exclusion criteria with an overall sample of 1,939 participants. Narrative synthesis was used. Most studies focused on adults (n = 21), and almost half (n = 12) included individuals with specific medical conditions. Studies primarily took place in free-living settings (n = 21) and used research-grade accelerometers (n = 24) worn on the hip (n = 18), thigh (n = 7), or wrist (n = 9). Outcomes included physical activity (n = 19), sedentary time/behavior (n = 12), or sleep (n = 2) and were calculated using proprietary formulas (e.g., Fitbit algorithm), cut points, and/or count-based methods. Most studies calculated responsiveness by comparing before versus after an intervention (n = 16). Six studies included a criterion measure to confirm that changes occurred. Limited research is available on the responsiveness of accelerometers for detecting change in 24-hr movement behaviors, particularly in youth populations, for sleep outcomes, and for commercial and thigh- or wrist-worn devices. Lack of a criterion measure precludes conclusions about the responsiveness even in more frequently studied outcomes/populations.

Actiwatch devices are often used to estimate time in bed (TIB) but recently became commercially unavailable. Thigh-worn activPAL devices could be a viable alternative. We tested convergent validity between activPAL (CREA algorithm) and Actiwatch devices. Data were from free-living samples comprising 47 youth (3–16 valid nights/participant) and 42 adults (6–26 valid nights/participant) who wore both devices concurrently. On average, activPAL predicted earlier bedtimes and later risetimes compared with Actiwatch, resulting in longer overnight intervals (by 1.49 hr/night for youth and 0.67 hr/night for adults; both p < .001). TIB interruptions were predicted less commonly by activPAL (mean <2 interruptions/night for both youth and adults) than Actiwatch (mean of 24–26 interruptions/night in both groups; both p < .001). Overnight intervals for both devices tended to overlap for lengthy periods (mean of 7.38 hr/night for youth and 7.69 hr/night for adults). Within these overlapping periods, the devices gave matching epoch-level TIB predictions an average of 87.9% of the time for youth and 84.3% of the time for adults. Most remaining epochs (11.8% and 15.1%, respectively) were classified as TIB by activPAL, but not Actiwatch. Overall, the devices had fair agreement during the overlapping periods but limited agreement when predicting interruptions, bedtime, or risetime. Future work should assess the criterion validity of activPAL devices to understand implications for health research. The present findings demonstrate that activPAL is not interchangeable with Actiwatch, which is consistent with their differing foundations (thigh inclination for activPAL vs. wrist movement for Actiwatch).