Accelerometer-Derived Daily Life Movement Classified by Machine Learning and Incidence of Cardiovascular Disease in Older Women: The OPACH Study.
Ontology highlight
ABSTRACT:
SUBMITTER: Nguyen S
PROVIDER: S-EPMC9075073 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:PurposeAccelerometers are increasingly used to obtain valuable descriptors of physical activity for health research. The cut-points approach to segment accelerometer data is widely used in physical activity research but requires resource expensive calibration studies and does not make it easy to explore the information that can be gained for a variety of raw data metrics. To address these limitations, we present a data-driven approach for segmenting and clustering the accelerometer data using unsupervised machine learning.MethodsThe data used came from five hundred fourteen-year-old participants from the Millennium cohort study who wore an accelerometer (GENEActiv) on their wrist on one weekday and one weekend day. A Hidden Semi-Markov Model (HSMM), configured to identify a maximum of ten behavioral states from five second averaged acceleration with and without addition of x, y, and z-angles, was used for segmenting and clustering of the data. A cut-points approach was used as comparison.ResultsTime spent in behavioral states with or without angle metrics constituted eight and five principal components to reach 95% explained variance, respectively; in comparison four components were identified with the cut-points approach. In the HSMM with acceleration and angle as input, the distributions for acceleration in the states showed similar groupings as the cut-points categories, while more variety was seen in the distribution of angles.ConclusionOur unsupervised classification approach learns a construct of human behavior based on the data it observes, without the need for resource expensive calibration studies, has the ability to combine multiple data metrics, and offers a higher dimensional description of physical behavior. States are interpretable from the distributions of observations and by their duration.
| S-EPMC6326431 | biostudies-literature
Project description:BackgroundWe examined the associations between accelerometry-measured physical activity (PA) and incidence of 13 cancers among a cohort of postmenopausal women.MethodsIn this prospective study, 6382 women wore ActiGraph GT3X+ accelerometers at the hip for up to 7 days during 2012-2013, and were followed over a median of 4.7 years for diagnosis of 13 invasive cancers. Calibrated intensity cut points were used to define minutes per day of total, light and moderate-to-vigorous PA. We used multivariable Cox regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for tertiles, and one-standard deviation (SD) unit increments of PA exposures in relation to cancer incidence. We examined effect measure modification by age, race/ethnicity, body mass index and smoking history.ResultsThe highest (vs. lowest) tertiles of total, light and moderate-to-vigorous PA were associated with covariate-adjusted HRs of 0.72 (95% CI = 0.53-0.97), 0.81 (95% CI = 0.60-1.09) and 0.66 (95% CI = 0.48-0.91), respectively. In age-stratified analyses, HRs for total PA were lower among women <80 years (HRper one-SD = 0.75, 95% CI = 0.63-0.90) than among women ≥80 years (HRper one-SD = 0.99, 95% CI = 0.82-1.18) (PInteraction = 0.03). Race/ethnicity, BMI and smoking did not strongly modify these associations.ConclusionsEngaging in physical activity may play a beneficial role in the prevention of certain cancers in older women.
| S-EPMC7188876 | biostudies-literature
Project description:The detection of activities of daily living (ADL) and the detection of falls is of utmost importance for addressing the issue of serious injuries and death as a consequence of elderly people falling. Wearable sensors can provide a viable solution for monitoring people in danger of falls with minimal external involvement from health or care home workers. In this work, we recorded accelerometer data from 35 healthy individuals performing various ADLs, as well as falls. Spatial and frequency domain features were extracted and used for the training of machine learning models with the aim of distinguishing between fall and no fall events, as well as between falls and other ADLs. Supervised classification experiments demonstrated the efficiency of the proposed approach, achieving an F1-score of 98.41% for distinguishing between fall and no fall events, and an F1-score of 88.11% for distinguishing between various ADLs, including falls. Furthermore, the created dataset, named "ShimFall&ADL" will be publicly released to facilitate further research on the field.
| S-EPMC7378757 | biostudies-literature
Project description:Limited evidence exists to inform physical activity (PA) and sedentary behavior guidelines for older people, especially women. Rigorous evidence on the amounts, intensities, and movement patterns associated with better health in later life is needed.The Objective PA and Cardiovascular Health (OPACH) Study is an ancillary study to the Women's Health Initiative (WHI) Program that examines associations of accelerometer-assessed PA and sedentary behavior with cardiovascular and fall events. Between 2012 and 2014, 7048 women aged 63-99 were provided with an ActiGraph GT3X+ (Pensacola, Florida) triaxial accelerometer, a sleep log, and an OPACH PA Questionnaire; 6489 have accelerometer data. Most women were in their 70s (40%) or 80s (46%), while approximately 10% were in their 60s and 4% were age 90 years or older. Non-Hispanic Black or Hispanic/Latina women comprise half of the cohort. Follow-up includes 1-year of falls surveillance with monthly calendars and telephone interviews of fallers, and annual follow-up for outcomes with adjudication of incident cardiovascular disease (CVD) events through 2020. Over 63,600 months of calendar pages were returned by 5,776 women, who reported 5,980 falls. Telephone interviews were completed for 1,492 women to ascertain the circumstances, injuries and medical care associated with falling. The dataset contains extensive information on phenotypes related to healthy aging, including inflammatory and CVD biomarkers, breast and colon cancer, hip and other fractures, diabetes, and physical disability.This paper describes the study design, methods, and baseline data for a diverse cohort of postmenopausal women who wore accelerometers under free-living conditions as part of the OPACH Study. By using accelerometers to collect more precise and complete data on PA and sedentary behavior in a large cohort of older women, this study will contribute crucial new evidence about how much, how vigorous, and what patterns of PA are necessary to maintain optimal cardiovascular health and to avoid falls in later life.ClinicalTrials.gov identifier NCT00000611 . Registered 27 October 1999.
| S-EPMC5307783 | biostudies-literature
Project description:BackgroundEvidence that higher sedentary time is associated with higher risk for cardiovascular disease (CVD) is based mainly on self-reported measures. Few studies have examined whether patterns of sedentary time are associated with higher risk for CVD.MethodsWomen from the Objective Physical Activity and Cardiovascular Health (OPACH) Study (n=5638, aged 63-97, mean age=79±7) with no history of myocardial infarction (MI) or stroke wore accelerometers for 4-to-7 days and were followed for up to 4.9 years for CVD events. Average daily sedentary time and mean sedentary bout duration were the exposures of interest. Cox regression models estimated hazard ratios (HR) and 95% confidence intervals (CI) for CVD using models adjusted for covariates and subsequently adjusted for potential mediators (body mass index (BMI), diabetes, hypertension, and CVD-risk biomarkers [fasting glucose, high-density lipoprotein, triglycerides, and systolic blood pressure]). Restricted cubic spline regression characterized dose-response relationships.ResultsThere were 545 CVD events during 19,350 person-years. Adjusting for covariates, women in the highest (≥ ~11 hr/day) vs. the lowest (≤ ~9 hr/day) quartile of sedentary time had higher risk for CVD (HR=1.62; CI=1.21-2.17; p-trend <0.001). Further adjustment for potential mediators attenuated but did not eliminate significance of these associations (p-trend<.05, each). Longer vs. shorter mean bout duration was associated with higher risks for CVD (HR=1.54; CI=1.27-2.02; p-trend=0.003) after adjustment for covariates. Additional adjustment for CVD-risk biomarkers attenuated associations resulting in a quartile 4 vs. quartile 1 HR=1.36; CI=1.01-1.83; p-trend=0.10). Dose-response associations of sedentary time and bout duration with CVD were linear (P-nonlinear >0.05, each). Women jointly classified as having high sedentary time and long bout durations had significantly higher risk for CVD (HR=1.34; CI=1.08-1.65) than women with both low sedentary time and short bout duration. All analyses were repeated for incident coronary heart disease (MI or CVD death) and associations were similar with notably stronger hazard ratios.ConclusionsBoth high sedentary time and long mean bout durations were associated in a dose-response manner with increased CVD risk in older women, suggesting that efforts to reduce CVD burden may benefit from addressing either or both component(s) of sedentary behavior.
| S-EPMC6481298 | biostudies-literature
Project description:BackgroundModerate-to-vigorous physical activity (MVPA) is proposed as key for cardiovascular diseases (CVD) prevention. At older ages, the role of sedentary behaviour (SB) and light intensity physical activity (LIPA) remains unclear. Evidence so far is based on studies examining movement behaviours as independent entities ignoring their co-dependency. This study examines the association between daily composition of objectively-assessed movement behaviours (MVPA, LIPA, SB) and incident CVD in older adults.MethodsWhitehall II accelerometer sub-study participants free of CVD at baseline (N = 3319, 26.7% women, mean age = 68.9 years in 2012-2013) wore a wrist-accelerometer from which times in SB, LIPA, and MVPA during waking period were extracted over 7 days. Compositional Cox regression was used to estimate the hazard ratio (HR) for incident CVD for daily compositions of movement behaviours characterized by 10 (20 or 30) minutes greater duration in one movement behaviour accompanied by decrease in another behaviour, while keeping the third behaviour constant, compared to reference composition. Analyses were adjusted for sociodemographic, lifestyle, cardiometabolic risk factors and multimorbidity index.ResultsOf the 3319 participants, 299 had an incident CVD over a mean (SD) follow-up of 6.2 (1.3) years. Compared to daily movement behaviour composition with MVPA at recommended 21 min per day (150 min/week), composition with additional 10 min of MVPA and 10 min less SB was associated with smaller risk reduction - 8% (HR, 0.92; 95% CI, 0.87-0.99) - than the 14% increase in risk associated with a composition of similarly reduced time in MVPA and more time in SB (HR, 1.14; 95% CI, 1.02-1.27). For a given MVPA duration, the CVD risk did not differ as a function of LIPA and SB durations.ConclusionsAmong older adults, an increase in MVPA duration at the expense of time in either SB or LIPA was found associated with lower incidence of CVD. This study lends support to public health guidelines encouraging increase in MVPA or at least maintain MVPA at current duration.
| S-EPMC8273960 | biostudies-literature
Project description:BackgroundThere is growing interest in the area of "wearable tech" and its relationship to health. A common element of many of these devices is a triaxial accelerometer that can yield continuous information on gross motor activity levels; how such data might predict changes in health is less clear.MethodsWe examined accelerometry data from 2,976 older men who were part of the Osteoporotic Fractures in Men (MrOS) study. Using a shape-naive technique, functional principal component analysis, we examined the patterns of motor activity over the course of 4-7 days and determined whether these patterns were associated with changes in polysomnographic-determined sleep and cognitive function (Trail Making Test-Part B [Trails B], Modified Mini-Mental State Examination [3MS]), as well as mortality over 6.5-8 years of follow-up.ResultsIn comparing baseline to 6.5 years later, multivariate modeling indicated that low daytime activity at baseline was associated with worsening of sleep efficiency (p < .05), more wake after sleep onset (p < .05), and a decrease in cognition (Trails B; p < .001), as well as a 1.6-fold higher rate of all-cause mortality (hazard ratio = 1.64 [1.34-2.00]). Earlier wake and bed times were associated with a decrease in cognition (3MS; p < .05). Having a late afternoon peak in activity was associated with a 1.4-fold higher rate of all-cause mortality (hazard ratio = 1.46 [1.21-1.77]). Those having a longer duration of their daytime activity with a bimodal activity pattern also had over a 1.4-fold higher rate of cardiovascular-related mortality (hazard ratio = 1.42 [1.02-1.98]).ConclusionsPatterns of daily activity may be useful as predictive biomarkers for changes in clinically relevant outcomes, including mortality and changes in sleep and cognition in older men.
| S-EPMC5905637 | biostudies-literature
Project description:BackgroundSelf-reported time spent standing has been associated with lower risk of mortality. No previous studies have examined this association using device-measured standing.MethodThis was a prospective cohort study of 5878 older (median age = 80 years), racial/ethnically diverse, community-dwelling women in the WHI Objective Physical Activity and Cardiovascular Health Study (OPACH). Women wore accelerometers for 1 week and were followed for mortality. The study applied previously validated machine learning algorithms to ActiGraph GT3X+ accelerometer data to separately measure time spent standing with and without ambulation. Cox proportional hazards models were used to estimate mortality risk adjusting for potential confounders. Effect modification by age, body mass index, moderate-to-vigorous physical activity, sedentary time, physical functioning, and race/ethnicity was evaluated.ResultsThere were 691 deaths during 26 649 person-years of follow-up through March 31, 2018 (mean follow-up = 4.8 years). In fully adjusted models, all-cause mortality risk was lower among those with more standing without ambulation (quartile [Q] 4 vs Q1 HR = 0.63; 95% CI = 0.49-0.81, p-trend = .003) and more standing with ambulation (Q4 vs Q1 HR = 0.50; 95% CI = 0.35-0.71, p-trend < .001). Associations of standing with ambulation and mortality were stronger among women with above-median sedentary time (HR = 0.51; 95% CI = 0.38-0.68) compared to women with below-median sedentary time (HR = 0.80; 95% CI = 0.59-1.07; p-interaction = .02).ConclusionsIn this prospective study among older women, higher levels of accelerometer-measured standing were associated with lower risks of all-cause mortality. Standing is an achievable approach to interrupting prolonged sedentary time, and if not contraindicated, is a safe and feasible behavior that appears to benefit health in older ages.
| S-EPMC7756713 | biostudies-literature
Project description:BackgroundEvidence suggests that short and long sleep durations are potential lifestyle factors associated with cardiovascular disease (CVD). Research on sleep duration and CVD risk is limited by use of self-report sleep measures, homogeneous populations, and studies on individual CVD risk factors. For women, risk of CVD and inadequate sleep duration increases with age. We hypothesized that accelerometer-measured sleep duration was associated with 10-year predicted probability of future CVD risk in a cohort of aging women.MethodsThis cross-sectional analysis included 3,367 older women (mean age 78.9 years; 53.3% White), from the Objective Physical Activity and Cardiovascular Health Study, ancillary study to the Women's Health Initiative. Women wore ActiGraph GT3X+ accelerometers on the hip for 24 hours/7 days. A 10-year predicted probability of future CVD risk, the Reynolds Risk Score (RRS), was computed using age, systolic blood pressure, high-sensitivity C-reactive protein (CRP), total and HDL cholesterol, diabetes mellitus status, smoking status, and family history of CVD. Average nightly sleep duration was derived from accelerometer data. Adjusted linear regression models investigated the association between sleep duration and RRS.ResultsResults suggested a U-shaped relationship between sleep duration and RRS, with both short and long sleep associated with higher RRS (p < .001). The association remained significant after adjustments for race/ethnicity, education, lifestyle factors, and health status indicators.ConclusionIn older women, actigraphy-ascertained sleep duration was associated with a 10-year predicted probability of future CVD risk. This study supports sleep duration as a modifiable risk factor for CVD in older women.
| S-EPMC7494034 | biostudies-literature
Project description:BACKGROUND:Stroke survivors are commonly left with disabilities that impair activities of daily living. The main objective of their rehabilitation program is to maximize the functional performance at home. However, the actual performance of patients in their home environment is unknown. Therefore, objective evaluation of daily life activities of stroke survivors in their physical interaction with the environment is essential for optimal guidance of rehabilitation therapy. Monitoring daily life movements could be very challenging, as it may result in large amounts of data, without any context. Therefore, suitable metrics are necessary to quantify relevant aspects of movement performance during daily life. The objective of this study is to develop data processing methods, which can be used to process movement data into relevant metrics for the evaluation of intra-patient differences in quality of movements in a daily life setting. METHODS:Based on an iterative requirement process, functional and technical requirements were formulated. These were prioritized resulting in a coherent set of metrics. An activity monitor was developed to give context to captured movement data at home. Finally, the metrics will be demonstrated in two stroke participants during and after their rehabilitation phases. RESULTS:By using the final set of metrics, quality of movement can be evaluated in a daily life setting. As example to demonstrate potential of presented methods, data of two stroke patients were successfully analyzed. Differences between in-clinic measurements and measurements during daily life are observed by applying the presented metrics and visualization methods. Heel height profiles show intra-patient differences in height, distance, stride profile, and variability between strides during a 10-m walk test in the clinic and walking at home. Differences in distance and stride profile between both feet were larger at home, than in clinic. For the upper extremities, the participant was able to reach further away from the pelvis and cover a larger area. DISCUSSION:Presented methods can be used for the objective evaluation of intra-patient differences in movement quality between in-clinic and daily life measurements. Any observed progression or deterioration of movement quality could be used to decide on continuing, stopping, or adjusting rehabilitation programs.
| S-EPMC4710748 | biostudies-literature