Validating measures of free-living physical activity in overweight and obese subjects using an accelerometer

Data Analytics in Physical Activity Studies With Accelerometers: Scoping Review

BACKGROUND

Monitoring free-living physical activity (PA) through wearable devices enables the real-time assessment of activity features associated with health outcomes and provision of treatment recommendations and adjustments. The conclusions of studies on PA and health depend crucially on reliable statistical analyses of digital data. Data analytics, however, are challenging due to the various metrics adopted for measuring PA, different aims of studies, and complex temporal variations within variables. The application, interpretation, and appropriateness of these analytical tools have yet to be summarized.

OBJECTIVE

This research aimed to review studies that used analytical methods for analyzing PA monitored by accelerometers. Specifically, this review addressed three questions: (1) What metrics are used to describe an individual's free-living daily PA? (2) What are the current analytical tools for analyzing PA data, particularly under the aims of classification, association with health outcomes, and prediction of health events? and (3) What challenges exist in the analyses, and what recommendations for future research are suggested regarding the use of statistical methods in various research tasks?

METHODS

This scoping review was conducted following an existing framework to map research studies by exploring the information about PA. Three databases, PubMed, IEEE Xplore, and the ACM Digital Library, were searched in February 2024 to identify related publications. Eligible articles were classification, association, or prediction studies involving human PA monitored through wearable accelerometers.

RESULTS

After screening 1312 articles, 428 (32.62%) eligible studies were identified and categorized into at least 1 of the following 3 thematic categories: classification (75/428, 17.5%), association (342/428, 79.9%), and prediction (32/428, 7.5%). Most articles (414/428, 96.7%) derived PA variables from 3D acceleration, rather than 1D acceleration. All eligible articles (428/428, 100%) considered PA metrics represented in the time domain, while a small fraction (16/428, 3.7%) also considered PA metrics in the frequency domain. The number of studies evaluating the influence of PA on health conditions has increased greatly. Among the studies in our review, regression-type models were the most prevalent (373/428, 87.1%). The machine learning approach for classification research is also gaining popularity (32/75, 43%). In addition to summary statistics of PA, several recent studies used tools to incorporate PA trajectories and account for temporal patterns, including longitudinal data analysis with repeated PA measurements and functional data analysis with PA as a continuum for time-varying association (68/428, 15.9%).

CONCLUSIONS

Summary metrics can quickly provide descriptions of the strength, frequency, and duration of individuals' overall PA. When the distribution and profile of PA need to be evaluated or detected, considering PA metrics as longitudinal or functional data can provide detailed information and improve the understanding of the role PA plays in health. Depending on the research goal, appropriate analytical tools can ensure the reliability of the scientific findings.

Quality Evaluation of Free-living Validation Studies for the Assessment of 24-Hour Physical Behavior in Adults via Wearables: Systematic Review

Background

Wearable technology is a leading fitness trend in the growing commercial industry and an established method for collecting 24-hour physical behavior data in research studies. High-quality free-living validation studies are required to enable both researchers and consumers to make guided decisions on which study to rely on and which device to use. However, reviews focusing on the quality of free-living validation studies in adults are lacking.

Objective

This study aimed to raise researchers’ and consumers’ attention to the quality of published validation protocols while aiming to identify and compare specific consistencies or inconsistencies between protocols. We aimed to provide a comprehensive and historical overview of which wearable devices have been validated for which purpose and whether they show promise for use in further studies.

Methods

Peer-reviewed validation studies from electronic databases, as well as backward and forward citation searches (1970 to July 2021), with the following, required indicators were included: protocol must include real-life conditions, outcome must belong to one dimension of the 24-hour physical behavior construct (intensity, posture or activity type, and biological state), the protocol must include a criterion measure, and study results must be published in English-language journals. The risk of bias was evaluated using the Quality Assessment of Diagnostic Accuracy Studies-2 tool with 9 questions separated into 4 domains (patient selection or study design, index measure, criterion measure, and flow and time).

Results

Of the 13,285 unique search results, 222 (1.67%) articles were included. Most studies (153/237, 64.6%) validated an intensity measure outcome such as energy expenditure. However, only 19.8% (47/237) validated biological state and 15.6% (37/237) validated posture or activity-type outcomes. Across all studies, 163 different wearables were identified. Of these, 58.9% (96/163) were validated only once. ActiGraph GT3X/GT3X+ (36/163, 22.1%), Fitbit Flex (20/163, 12.3%), and ActivPAL (12/163, 7.4%) were used most often in the included studies. The percentage of participants meeting the quality criteria ranged from 38.8% (92/237) to 92.4% (219/237). On the basis of our classification tree to evaluate the overall study quality, 4.6% (11/237) of studies were classified as low risk. Furthermore, 16% (38/237) of studies were classified as having some concerns, and 72.9% (173/237) of studies were classified as high risk.

Conclusions

Overall, free-living validation studies of wearables are characterized by low methodological quality, large variability in design, and focus on intensity. Future research should strongly aim at biological state and posture or activity outcomes and strive for standardized protocols embedded in a validation framework. Standardized protocols for free-living validation embedded in a framework are urgently needed to inform and guide stakeholders (eg, manufacturers, scientists, and consumers) in selecting wearables for self-tracking purposes, applying wearables in health studies, and fostering innovation to achieve improved validity.

Validity of Accelerometers for the Evaluation of Energy Expenditure in Obese and Overweight Individuals: A Systematic Review

OBJECTIVE

Even though the validity of accelerometers for the measurement of energy expenditure (EE) has been demonstrated for normal-weight individuals, the applicability of this instrument in obese individuals remains controversial. This review aims to summarize the level of agreement between accelerometers and the gold standards (indirect calorimetry and doubly labelled water) for the measurement of energy expenditure (EE) in obese or overweight individuals.

METHODS

The literature search was limited to comparison studies assessing agreement in EE determination between accelerometers and indirect calorimetry (IC) or doubly labelled water (DLW). We searched in PubMed and in Scopus until March 1, 2019. The analysis was restricted to obese or overweight adult individuals. The following descriptive information was extracted for each study: sample size, characteristics of participants (sex, age, BMI, fat mass percentage, any pathological conditions, modality of recruitment in the study, and exclusion criteria), accelerometer description (model, type and body position), and type of gold standard and validity protocol (duration, conditions, and requirements during and before the experiment). Three review authors independently screened the obtained results, and the quality of the selected articles was assessed by the QUADAS-2 tool.

RESULTS

We obtained seventeen eligible articles, thirteen of which showed concerns for the applicability section, due to the patient selection. Regarding the accelerometers, nine devices were validated in the included studies with the BodyMedia SenseWear® (SWA) being the most frequently validated. Although correlations between accelerometers and the gold standard were high in some studies, agreement between the two methods was low, as shown by the Bland-Altman plots.

CONCLUSIONS

Most accelerometer estimations of EE were inaccurate for obese/overweight subjects, and authors advise to improve the accuracy of algorithms for SWA software, or the predicted equations for estimating EE from other accelerometers.

Calibration and validation of accelerometry to measure physical activity in adult clinical groups: A systematic review

A growing body of research calibrating and validating accelerometers to classify physical activity intensities has led to a range of cut-points. However, the applicability of current calibration protocols to clinical populations remains to be addressed. The aim of this review was to evaluate the accuracy of the methods for calibrating and validating of accelerometers to estimate physical activity intensity thresholds for clinical populations. Six databases were searched between March and July to 2017 using text words and subject headings. Studies developing moderate-to-vigorous intensity physical activity cut-points for adult clinical populations were included. The risk of bias was assessed using the health measurement instruments and a specific checklist for calibration studies. A total of 543,741 titles were found and 323 articles were selected for full-text assessment, with 11 meeting the inclusion criteria. Twenty-three different methods for calibration were identified using different models of ActiGraph and Actical accelerometers. Disease-specific cut-points ranged from 591 to 2717 counts·min-1 and were identified for two main groups of clinical conditions: neuromusculoskeletal disorders and metabolic diseases. The heterogeneity in the available clinical protocols hinders the applicability and comparison of the developed cut-points. As such, a mixed protocol containing a controlled laboratory exercise test and activities of daily-life is suggested. It is recommended that this be combined with a statistical approach that allows for adjustments according to disease severity or the use of machine learning models. Finally, this review highlights the generalisation of cut-points developed on healthy populations to clinical populations is inappropriate.

Do moderate- to vigorous-intensity accelerometer count thresholds correspond to relative moderate- to vigorous-intensity physical activity?

We aimed to predict % maximal oxygen consumption at absolute accelerometer thresholds and to estimate and compare durations of objective physical activity (PA) among body mass index (BMI) categories using thresholds that account for cardiorespiratory fitness. Eight hundred twenty-eight adults (53.5% male; age, 33.9 ± 0.3 years) from the National Health and Nutrition Examination Survey 2003-2004 were analyzed. Metabolic equivalent values at absolute thresholds were converted to percentage of maximal oxygen consumption, and accelerometer counts corresponding to 40% or 60% maximal oxygen consumption were determined using 4 energy expenditure prediction equations. Absolute thresholds underestimated PA intensity for all adults; however, because of lower fitness, individuals with overweight and obesity work at significantly higher percentage of maximal oxygen consumption at the absolute thresholds and require significantly lower accelerometer counts to reach relative moderate and vigorous PA intensities compared with those with normal weight (P < 0.05). However, moderate-to-vigorous physical activity (MVPA) durations were shorter when using relative thresholds compared with absolute thresholds (in all BMI groups, P < 0.05), and they were shorter among individuals with obesity compared with those with normal weight when using relative thresholds (P < 0.05). Regardless of the thresholds used, a greater proportion of individuals with normal weight met the PA guideline of 150 min·week^-1 of MVPA compared with individuals with obesity (absolute: 21.3% vs 6.7%; Yngve: 4.0% vs 0.2%; Swartz: 10.7% vs 3.9%; Hendelman: 4.7% vs 0.2%; Freedson: 6.4% vs 0.5%; P < 0.05). Current absolute thresholds of accelerometry-derived PA may overestimate MVPA for all BMI categories when compared with relative thresholds that account for cardiorespiratory fitness. Given the large variability in our results, more work is needed to better understand how to use accelerometers for evaluating PA at the population level.

Determining the Validity and Accuracy of Multiple Activity-Tracking Devices in Controlled and Free-Walking Conditions

PURPOSE

Clinicians and fitness professionals are increasingly recommending the use of activity trackers. This study compares commercially available activity tracking devices for step and distance accuracy in common exercise settings.

DESIGN

Cross sectional.

SETTING

Rochester, Minnesota.

PARTICIPANTS

Thirty-two men (n = 10) and women (n = 22) participated in the study.

MEASURES

Researchers manually counted steps and measured distance for all trials, while participants wore 6 activity tracking devices that measured steps and distance.

ANALYSIS

We computed the difference between the number of steps measured by the device and the actual number of steps recorded by the observers, as well as the distance displayed by the device and the actual distance measured.

RESULTS

The analyses showed that both the device and walking trials affected the accuracy of the results (steps or distance, P < .001). Hip-based devices were more accurate and consistent for measuring step count. No significant differences were found among devices or locations for the distance measured.

CONCLUSIONS

Hip-based activity tracking devices varied in accuracy but performed better than their wrist-based counterparts for step accuracy. Distance measurements for both types of devices were more consistent but lacked accuracy.

Validation of Consumer-Based Hip and Wrist Activity Monitors in Older Adults With Varied Ambulatory Abilities

BACKGROUND

The accuracy of step detection in consumer-based wearable activity monitors in older adults with varied ambulatory abilities is not known.

METHODS

We assessed the validity of two hip-worn (Fitbit One and Omron HJ-112) and two wrist-worn (Fitbit Flex and Jawbone UP) activity monitors in 99 older adults of varying ambulatory abilities and also included the validity results from the ankle-worn StepWatch as a comparison device. Nonimpaired, impaired (Short Physical Performance Battery Score < 9), cane-using, or walker-using older adults (62 and older) ambulated at a self-selected pace for 100 m wearing all activity monitors simultaneously. The criterion measure was directly observed steps. Intraclass correlation coefficients (ICC), mean percent error and mean absolute percent error, equivalency, and Bland-Altman plots were used to assess accuracy.

RESULTS

Nonimpaired adults steps were underestimated by 4.4% for StepWatch (ICC = 0.87), 2.6% for Fitbit One (ICC = 0.80), 4.5% for Omron HJ-112 (ICC = 0.72), 26.9% for Fitbit Flex (ICC = 0.15), and 2.9% for Jawbone UP (ICC = 0.55). Impaired adults steps were underestimated by 3.5% for StepWatch (ICC = 0.91), 1.7% for Fitbit One (ICC = 0.96), 3.2% for Omron HJ-112 (ICC = 0.89), 16.3% for Fitbit Flex (ICC = 0.25), and 8.4% for Jawbone UP (ICC = 0.50). Cane-user and walker-user steps were underestimated by StepWatch by 1.8% (ICC = 0.98) and 1.3% (ICC = 0.99), respectively, where all other monitors underestimated steps by >11.5% (ICCs < 0.05).

CONCLUSIONS

StepWatch, Omron HJ-112, Fitbit One, and Jawbone UP appeared accurate at measuring steps in older adults with nonimpaired and impaired ambulation during a self-paced walking test. StepWatch also appeared accurate at measuring steps in cane-users.

Objectively Measured Physical Activity in European Adults: Cross-Sectional Findings from the Food4Me Study

Background

Comparisons of objectively measured physical activity (PA) between residents of European countries measured concurrently with the same protocol are lacking. We aimed to compare PA between the seven European countries involved in the Food4Me Study, using accelerometer data collected remotely via the Internet.

Methods

Of the 1607 participants recruited, 1287 (539 men and 748 women) provided at least 3 weekdays and 2 weekend days of valid accelerometer data (TracmorD) at baseline and were included in the present analyses.

Results

Men were significantly more active than women (physical activity level = 1.74 vs. 1.70, p < 0.001). Time spent in light PA and moderate PA differed significantly between countries but only for women. Adherence to the World Health Organization recommendation to accumulate at least 150 min of moderate-equivalent PA weekly was similar between countries for men (range: 54–65%) but differed significantly between countries for women (range: 26–49%). Prevalence estimates decreased substantially for men and women in all seven countries when PA guidelines were defined as achieving 30 min of moderate and vigorous PA per day.

Conclusions

We were able to obtain valid accelerometer data in real time via the Internet from 80% of participants. Although our estimates are higher compared with data from Sweden, Norway, Portugal and the US, there is room for improvement in PA for all countries involved in the Food4Me Study.

Changes in Physical Activity Following a Genetic-Based Internet-Delivered Personalized Intervention: Randomized Controlled Trial (Food4Me)

Background

There is evidence that physical activity (PA) can attenuate the influence of the fat mass- and obesity-associated (FTO) genotype on the risk to develop obesity. However, whether providing personalized information on FTO genotype leads to changes in PA is unknown.

Objective

The purpose of this study was to determine if disclosing FTO risk had an impact on change in PA following a 6-month intervention.

Methods

The single nucleotide polymorphism (SNP) rs9939609 in the FTO gene was genotyped in 1279 participants of the Food4Me study, a four-arm, Web-based randomized controlled trial (RCT) in 7 European countries on the effects of personalized advice on nutrition and PA. PA was measured objectively using a TracmorD accelerometer and was self-reported using the Baecke questionnaire at baseline and 6 months. Differences in baseline PA variables between risk (AA and AT genotypes) and nonrisk (TT genotype) carriers were tested using multiple linear regression. Impact of FTO risk disclosure on PA change at 6 months was assessed among participants with inadequate PA, by including an interaction term in the model: disclosure (yes/no) × FTO risk (yes/no).

Results

At baseline, data on PA were available for 874 and 405 participants with the risk and nonrisk FTO genotypes, respectively. There were no significant differences in objectively measured or self-reported baseline PA between risk and nonrisk carriers. A total of 807 (72.05%) of the participants out of 1120 in the personalized groups were encouraged to increase PA at baseline. Knowledge of FTO risk had no impact on PA in either risk or nonrisk carriers after the 6-month intervention. Attrition was higher in nonrisk participants for whom genotype was disclosed (P=.01) compared with their at-risk counterparts.

Conclusions

No association between baseline PA and FTO risk genotype was observed. There was no added benefit of disclosing FTO risk on changes in PA in this personalized intervention. Further RCT studies are warranted to confirm whether disclosure of nonrisk genetic test results has adverse effects on engagement in behavior change.

Trial Registration

ClinicalTrials.gov NCT01530139; http://clinicaltrials.gov/show/NCT01530139 (Archived by WebCite at: http://www.webcitation.org/6XII1QwHz)

Prediction of activity-related energy expenditure using accelerometer-derived physical activity under free-living conditions: a systematic review

BACKGROUND/OBJECTIVES

Activity-related energy expenditure (AEE) might be an important factor in the etiology of chronic diseases. However, measurement of free-living AEE is usually not feasible in large-scale epidemiological studies but instead has traditionally been estimated based on self-reported physical activity. Recently, accelerometry has been proposed for objective assessment of physical activity, but it is unclear to what extent this methods explains the variance in AEE.

SUBJECTS/METHODS

We conducted a systematic review searching MEDLINE database (until 2014) on studies that estimated AEE based on accelerometry-assessed physical activity in adults under free-living conditions (using doubly labeled water method). Extracted study characteristics were sample size, accelerometer (type (uniaxial, triaxial), metrics (for example, activity counts, steps, acceleration), recording period, body position, wear time), explained variance of AEE (R(2)) and number of additional predictors. The relation of univariate and multivariate R(2) with study characteristics was analyzed using nonparametric tests.

RESULTS

Nineteen articles were identified. Examination of various accelerometers or subpopulations in one article was treated separately, resulting in 28 studies. Sample sizes ranged from 10 to 149. In most studies the accelerometer was triaxial, worn at the trunk, during waking hours and reported activity counts as output metric. Recording periods ranged from 5 to 15 days. The variance of AEE explained by accelerometer-assessed physical activity ranged from 4 to 80% (median crude R(2)=26%). Sample size was inversely related to the explained variance. Inclusion of 1 to 3 other predictors in addition to accelerometer output significantly increased the explained variance to a range of 12.5-86% (median total R(2)=41%). The increase did not depend on the number of added predictors.

CONCLUSIONS

We conclude that there is large heterogeneity across studies in the explained variance of AEE when estimated based on accelerometry. Thus, data on predicted AEE based on accelerometry-assessed physical activity need to be interpreted cautiously.

Effects of a Web-Based Personalized Intervention on Physical Activity in European Adults: A Randomized Controlled Trial

BACKGROUND

The high prevalence of physical inactivity worldwide calls for innovative and more effective ways to promote physical activity (PA). There are limited objective data on the effectiveness of Web-based personalized feedback on increasing PA in adults.

OBJECTIVE

It is hypothesized that providing personalized advice based on PA measured objectively alongside diet, phenotype, or genotype information would lead to larger and more sustained changes in PA, compared with nonpersonalized advice.

METHODS

A total of 1607 adults in seven European countries were randomized to either a control group (nonpersonalized advice, Level 0, L0) or to one of three personalized groups receiving personalized advice via the Internet based on current PA plus diet (Level 1, L1), PA plus diet and phenotype (Level 2, L2), or PA plus diet, phenotype, and genotype (Level 3, L3). PA was measured for 6 months using triaxial accelerometers, and self-reported using the Baecke questionnaire. Outcomes were objective and self-reported PA after 3 and 6 months.

RESULTS

While 1270 participants (85.81% of 1480 actual starters) completed the 6-month trial, 1233 (83.31%) self-reported PA at both baseline and month 6, but only 730 (49.32%) had sufficient objective PA data at both time points. For the total cohort after 6 months, a greater improvement in self-reported total PA (P=.02) and PA during leisure (nonsport) (P=.03) was observed in personalized groups compared with the control group. For individuals advised to increase PA, we also observed greater improvements in those two self-reported indices (P=.006 and P=.008, respectively) with increased personalization of the advice (L2 and L3 vs L1). However, there were no significant differences in accelerometer results between personalized and control groups, and no significant effect of adding phenotypic or genotypic information to the tailored feedback at month 3 or 6. After 6 months, there were small but significant improvements in the objectively measured physical activity level (P<.05), moderate PA (P<.01), and sedentary time (P<.001) for individuals advised to increase PA, but these changes were similar across all groups.

CONCLUSIONS

Different levels of personalization produced similar small changes in objective PA. We found no evidence that personalized advice is more effective than conventional "one size fits all" guidelines to promote changes in PA in our Web-based intervention when PA was measured objectively. Based on self-reports, PA increased to a greater extent with more personalized advice. Thus, it is crucial to measure PA objectively in any PA intervention study.

TRIAL REGISTRATION

ClinicalTrials.gov NCT01530139; http://clinicaltrials.gov/show/NCT01530139 (Archived by WebCite at: http://www.webcitation.org/6XII1QwHz).

Objectively Measured Sedentary Time and Cardiometabolic Biomarkers in US Hispanic/Latino Adults: The Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

BACKGROUND

Sedentary behavior is recognized as a distinct construct from lack of moderate-vigorous physical activity and is associated with deleterious health outcomes. Previous studies have primarily relied on self-reported data, whereas data on the relationship between objectively measured sedentary time and cardiometabolic biomarkers are sparse, especially among US Hispanics/Latinos.

METHODS AND RESULTS

We examined associations of objectively measured sedentary time (via Actical accelerometers for 7 days) and multiple cardiometabolic biomarkers among 12 083 participants, aged 18 to 74 years, from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). Hispanics/Latinos of diverse backgrounds (Central American, Cuban, Dominican, Mexican, Puerto Rican, and South American) were recruited from 4 US cities between 2008 and 2011. Sedentary time (<100 counts/min) was standardized to 16 hours/d of wear time. The mean sedentary time was 11.9 hours/d (74% of accelerometer wear time). After adjustment for moderate-vigorous physical activity and confounding variables, prolonged sedentary time was associated with decreased high-density lipoprotein cholesterol (P=0.04), and increased triglycerides, 2-hour glucose, fasting insulin, and homeostatic model assessment of insulin resistance (all P<0.0001). These associations were generally consistent across age, sex, Hispanic/Latino backgrounds, and physical activity levels. Even among individuals meeting physical activity guidelines, sedentary time was detrimentally associated with several cardiometabolic biomarkers (diastolic blood pressure, high-density lipoprotein cholesterol, fasting and 2-hour glucose, fasting insulin and homeostatic model assessment of insulin resistance; all P<0.05).

CONCLUSIONS

Our large population-based, objectively derived data showed deleterious associations between sedentary time and cardiometabolic biomarkers, independent of physical activity, in US Hispanics/Latinos. Our findings emphasize the importance of reducing sedentary behavior for the prevention of cardiometabolic diseases, even in those who meet physical activity recommendations.