scholarly journals Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK Biobank participants

2017 ◽  
Author(s):  
Matthew Willetts ◽  
Sven Hollowell ◽  
Louis Aslett ◽  
Chris Holmes ◽  
Aiden Doherty
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Machine Learning ◽  
Markov Models ◽  
Activity Patterns ◽  
Sleep Time ◽  
Sensor Data ◽  
Statistical Machine Learning ◽  
Using Data ◽  
Objectively Measured

ABSTRACTCurrent public health guidelines on physical activity and sleep duration are limited by a reliance on subjective self-reported evidence. Using data from simple wrist-worn activity monitors, we developed a tailored machine learning model, using balanced random forests with Hidden Markov Models, to reliably detect a number of activity modes. We show that physical activity and sleep behaviours can be classified with 87% accuracy in 159,504 minutes of recorded free-living behaviours from 132 adults. These trained models can be used to infer fine resolution activity patterns at the population scale in 96,220 participants. For example, we find that men spend more time in both low- and high-intensity behaviours, while women spend more time in mixed behaviours. Walking time is highest in spring and sleep time lowest during the summer. This work opens the possibility of future public health guidelines informed by the health consequences associated with specific, objectively measured, physical activity and sleep behaviours.

scholarly journals Association between physical activity patterns and anthropometric characteristics of adults: an issue of public health?

2017 ◽  
Vol 9 (1) ◽  
pp. 124-132 ◽  
Author(s):  
Filipe Manuel Clemente ◽  
Fernando Manuel Lourenço Martins ◽  
Pantelis Theodoros Nikolaidis ◽  
Rui Sousa Mendes
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Activity Patterns ◽  
Binary Logistic Regression ◽  
P Value ◽  
Body Fat Mass ◽  
Full Model ◽  
Cross Sectional ◽  
Anthropometric Characteristics ◽  
Objectively Measured

Summary Study aim: The aim of this study was to evaluate the association between objectively measured daily physical activity (PA) and body fat mass (BF) and body mass index (BMI). A further aim was to analyse the variance of PA between quartiles of BF and BMI. Material and methods: A cross-sectional, observational study of 126 university students (53 males aged 20.46 ± 2.04 years and 73 female aged 19.69 ± 1.32 years) was conducted. Results: The female participants and PA characteristics explain 57.10% of BF variance and the model was statistically signifi­cant (F(6, 875) = 196.38; p = 0.001). BMI was also included in the model. Standard binary logistic regression was used to test the hypothesis that female sex and PA characteristics can influence overweight. The full model containing all variables was statistically significant (G2(6) = 58.598, p-value = 0.001). Analysis of variance between BF quartiles revealed statistically sig­nificant differences in male participants in light PA (p = 0.001; ES = 0.09), moderate PA (p = 0.001; ES = 0.042) and vigorous PA (p = 0.001; ES = 0.130). Conclusions: The statistical model in the regression analysis suggests that low and vigorous levels of PA explain 57% of BF variance in female participants.

Physical activity patterns among walkers and compliance with public health recommendations

2002 ◽  
Vol 34 (8) ◽  
pp. 1255-1261 ◽  
Author(s):  
ANN P. RAFFERTY ◽  
MATHEW J. REEVES ◽  
HARRY B. MCGEE ◽  
JAMES M. PIVARNIK
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Activity Patterns ◽  
Physical Activity Patterns

scholarly journals Limitations of Canada’s physical activity data: implications for monitoring trendsThis article is part of a supplement entitled Advancing physical activity measurement and guidelines in Canada: a scientific review and evidence-based foundation for the future of Canadian physical activity guidelines co-published by Applied Physiology, Nutrition, and Metabolism and the Canadian Journal of Public Health. It may be cited as Appl. Physiol. Nutr. Metab. 32(Suppl. 2E) or as Can. J. Public Health 98(Suppl. 2).

2007 ◽  
Vol 32 (S2E) ◽  
pp. S185-S194 ◽  
Author(s):  
Peter T. Katzmarzyk ◽  
Mark S. Tremblay
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Activity Patterns ◽  
Health Concern ◽  
Activity Measurement ◽  
Clear Understanding ◽  
Activity Data ◽  
Scientific Review ◽  
Physical Activity Guidelines ◽  
Applied Physiology

The current low level of physical activity among Canadians is a dominant public health concern. Accordingly, a clear understanding of physical activity patterns and trends is of paramount importance. Irregularities in monitoring, analysis, and reporting procedures create potential confusion among researchers, policy-makers, and the public alike. The purpose of this paper is to consolidate reported findings and provide a critical assessment of the physical activity surveillance procedures, analytical practices, and reporting protocols currently employed in Canada to provide insights for accurate and consistent interpretation of data, as well as recommendations for future surveillance efforts.

Towards Behaviour Recognition with Unlabelled Sensor Data

2013 ◽  
pp. 86-110
Author(s):  
Sook-Ling Chua ◽  
Stephen Marsland ◽  
Hans W. Guesgen
Keyword(s):  
Machine Learning ◽  
Data Mining ◽  
Inverse Problem ◽  
Sensor Data ◽  
Training Set ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Using Data ◽  
Symbolic Approach ◽  
Behaviour Recognition

The problem of behaviour recognition based on data from sensors is essentially an inverse problem: given a set of sensor observations, identify the sequence of behaviours that gave rise to them. In a smart home, the behaviours are likely to be the standard human behaviours of living, and the observations will depend upon the sensors that the house is equipped with. There are two main approaches to identifying behaviours from the sensor stream. One is to use a symbolic approach, which explicitly models the recognition process. Another is to use a sub-symbolic approach to behaviour recognition, which is the focus in this chapter, using data mining and machine learning methods. While there have been many machine learning methods of identifying behaviours from the sensor stream, they have generally relied upon a labelled dataset, where a person has manually identified their behaviour at each time. This is particularly tedious to do, resulting in relatively small datasets, and is also prone to significant errors as people do not pinpoint the end of one behaviour and commencement of the next correctly. In this chapter, the authors consider methods to deal with unlabelled sensor data for behaviour recognition, and investigate their use. They then consider whether they are best used in isolation, or should be used as preprocessing to provide a training set for a supervised method.

Physical Activity Surveillance Using Wearable Activity Monitors: Are US Adults Willing to Share Their Data?

2020 ◽  
Vol 34 (6) ◽  
pp. 672-676 ◽  
Author(s):  
Eric T. Hyde ◽  
John D. Omura ◽  
Janet E. Fulton ◽  
Andre Weldy ◽  
Susan A. Carlson
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Health Agency ◽  
Cross Sectional ◽  
Physically Active ◽  
Web Based ◽  
Activity Monitors ◽  
Share Data ◽  
Using Data ◽  
National Physical Activity

Purpose: Wearable activity monitors (wearables) have generated interest for national physical activity (PA) surveillance; however, concerns exist related to estimates obtained from current users willing to share data. We examined how limiting data to current users who are willing to share data associated with PA estimates in a nationwide sample. Design: Cross-sectional web-based survey. Setting: US adults. Subjects: In total, 942 respondents. Measures: The 2018 Government & Academic Omnibus Survey assessing current wearable use, willingness to share data with various people or organizations, and PA levels. Analysis: Estimated the prevalence of current wearable use; current users’ willingness to share data with various people or organizations; and PA levels overall, among current users, and among current users willing to share their data. Results: Overall, 21.7% (95% confidence interval [CI]: 19.1-24.5) of US adults reported currently using a wearable. Among current users, willingness to share ranged from 40.1% with a public health agency to 76.3% with their health-care provider. Overall, 62.2% (95% CI: 58.9-65.3) of adults were physically active. These levels were similar between current users (75.0%, 95% CI: 68.3-80.7) and current users willing to share their data (75.3%, 95% CI: 67.9-81.5). Conclusion: Our findings suggest that using data from wearable users may overestimate PA levels, although reported willingness to share the data may not compound this issue.

scholarly journals A Structured and Flexible Language for Physical Activity Assessment and Characterization

2013 ◽  
Vol 2013 ◽  
pp. 1-9 ◽  
Author(s):  
Pedro Silva ◽  
Maria Teresa Andrade ◽  
Pedro Carvalho ◽  
Jorge Mota
Keyword(s):  
Public Health ◽  
Physical Activity ◽  
Sedentary Behavior ◽  
Activity Patterns ◽  
Public Health Research ◽  
Cognitive Limitations ◽  
Public Health Professionals ◽  
Activity Assessment ◽  
Base Language ◽  
Flexible Base

Developing more accurate assessments of physical activity (PA) and sedentary behavior (SB) is an important public health research priority. Assessing PA and SB is challenging in all segments of the population, but it is especially difficult in children due to cognitive limitations and more sporadic and intermittent activity patterns. Moreover, they are influenced by several factors including temporal-spatial constraints and social conditions. To accurately assess PA and SB, it is essential to clearly define methods for describing all these factors. The goal of this paper is to potentiate advances in the field by proposing a base ontology for characterizing physical activity, sedentary behavior, and the context in which it occurs. The ontology would establish a flexible base language to facilitate standardized descriptions of these behaviors for researchers and public health professionals.

scholarly journals Continuous, objective measurement of physical activity during chemotherapy for breast cancer: the Activity in Treatment pilot study

2019 ◽  
Vol 10 (4) ◽  
pp. 1031-1038 ◽  
Author(s):  
Sandahl H Nelson ◽  
Lauren S Weiner ◽  
Loki Natarajan ◽  
Barbara A Parker ◽  
Ruth E Patterson ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Physical Activity ◽  
Average Rate ◽  
Vigorous Physical Activity ◽  
Activity Patterns ◽  
Early Period ◽  
Activity Levels ◽  
Potential Benefits ◽  
Objectively Measured ◽  
Restricted Cubic Splines

Abstract Despite many potential benefits of physical activity during and after breast cancer treatment, activity levels typically decline from pre- to posttreatment. Most previous research has relied on self-reported activity. The purpose of this study were to assess patterns of daily, to objectively measured physical activity throughout chemotherapy for breast cancer, and to identify predictors of physical activity patterns. Participants were given a Fitbit before starting chemotherapy and asked to wear it throughout chemotherapy. Restricted cubic splines assessed nonlinear patterns of Fitbit measured total physical activity (TPA) and moderate-to-vigorous physical activity (MVPA) throughout the duration of chemotherapy (mean = 17 weeks, standard deviation [SD] = 6.3). Mixed-effects regression models assessed the rate of physical activity decline. Regressions of subject-level random slope assessed predictors of the rate of physical activity decline on participant and cancer characteristics and self-reported physical and cognitive functioning. Participants (n = 32) were on average 50 years old; the majority had stage II breast cancer. MVPA declined linearly at a mean rate of 1.4 min/day (p = .002) for every 10% of chemotherapy completed, whereas TPA declined linearly at an average rate of 13.4 min/day (p = .0007) for every 10% of chemotherapy completed, until around halfway through chemotherapy, when activity rates leveled off. HER+ receptor status was associated with a greater rate of MVPA decline, β = 13.3, p = .04. This novel study of objectively measured daily MVPA throughout chemotherapy showed that most reductions in activity occurred during the first half of a course of chemotherapy. Targeting this early period of chemotherapy may be important for preventing declines in activity levels throughout chemotherapy.

Sign in / Sign up

Export Citation Format

Share Document