Statistical Machine Learning for Automatic Assessment of Physical Activity Intensity Using Multi-axial Accelerometry and Heart Rate

This work explores the automatic recognition of physical activity intensity patterns from multi-axial accelerometry and heart rate signals. Data collection was carried out in free-living conditions and in three controlled gymnasium circuits, for a total amount of 179.80 h of data divided into: sedentary situations (65.5%), light-to-moderate activity (17.6%) and vigorous exercise (16.9%). The proposed machine learning algorithms comprise the following steps: time-domain feature definition, standardization and PCA projection, unsupervised clustering (by k-means and GMM) and a HMM to account for long-term temporal trends. Performance was evaluated by 30 runs of a 10-fold cross-validation. Both k-means and GMM-based approaches yielded high overall accuracy (86.97% and 85.03%, respectively) and, given the imbalance of the dataset, meritorious F-measures (up to 77.88%) for non-sedentary cases. Classification errors tended to be concentrated around transients, what constrains their practical impact. Hence, we consider our proposal to be suitable for 24 h-based monitoring of physical activity in ambulatory scenarios and a first step towards intensity-specific energy expenditure estimators

Effectiveness of a School-Based Intervention on Physical Activity for High School Students in Brazil: The Saude na Boa Project

Background: We evaluated the effectiveness of a school-based intervention on the promotion of physical activity among high school students in Brazil: the Saude no Boa project. Methods: A school-based, randomized trial was carried out in 2 Brazilian cities: Recife (northeast) and Florianopolis (south). Ten schools in each city were matched by size and location, and randomized into intervention or control groups. The intervention included environmental/organizational changes, physical activity education, and personnel training and engagement. Students age 15 to 24 years were evaluated at baseline and 9 months later (end of school year). Results: Although similar at baseline, after the intervention, the control group reported significantly fewer d/wk accumulating 60 minutes+ moderate-to-vigorous physical activity (MVPA) in comparison with the intervention group (2.6 versus 3.3, P < .001). The prevalence of inactivity (0 days per week) rose in the control and decreased in the intervention group. The odds ratio for engaging at least once per week in physical activity associated with the intervention was 1.83 (95% CI = 1.24-2.71) in the unadjusted analysis and 1.88 (95% CI = 1.27-2.79) after controlling for gender. Conclusion: The Saude no Boa intervention was effective at reducing the prevalence of physical inactivity. The possibility of expanding the intervention to other locations should be considered.

Promoting physical activity to older adults: a preliminary evaluation of three general practice-based strategies

The aim of this study was to explore the feasibility of an exercise scientist (ES) working in general practice to promote physical activity (PA) to 55 to 70 year old adults. Participants were randomised into one of three groups: either brief verbal and written advice from a general practitioner (GP) (G1, N=9); or individualised counselling and follow-up telephone calls from an ES, either with (G3, N=8) or without a pedometer (G2, N=11). PA levels were assessed at week 1, after the 12-wk intervention and again at 24 weeks. After the 12-wk intervention, the average increase in PA was 116 (SD=237) min/wk; N=28, p < 0.001. Although there were no statistically significant between-group differences, the average increases in PA among G2 and G3 participants were 195 (SD=207) and 138 (SD=315) min/wk respectively, compared with no change (0.36, SD=157) in G1. After 24 weeks, average PA levels remained 56 (SD=129) min/wk higher than in week 1. The small numbers of participants in this feasibility study limit the power to detect significant differences between groups, but it would appear that individualised counselling and follow-up contact from an ES, with or without a pedometer, can result in substantial changes in PA levels. A larger study is now planned to confirm these findings.

Physical activity and sedentary behavior: A population-based study of barriers, enjoyment, and preference

The associations of physical activity and sedentary behavior with barriers, enjoyment, and preferences were examined in a population-based mail survey of 1,332 adults. Respondents reporting high enjoyment and preference for physical activity were more likely to report high levels of activity. Those reporting cost, the weather, and personal barriers to physical activity were less likely to be physically active. Preference for sedentary behavior was associated with the decreased likelihood of being physically active, and the weather as a barrier to physical activity was associated with the increased likelihood of sedentary behavior. These constructs can be used to examine individual and environmental influences on physical activity and sedentary behavior in specific populations and could inform the development of targeted interventions.

Tracking of physical activity from adolescence to adulthood: a population-based study

OBJECTIVE: To assess the association between regular physical activity in adolescence and leisure-time physical activity in adulthood, with emphasis on gender differences. METHODS: A population-based cross-sectional study was carried out in Pelotas, Southern Brazil, in 2003. A representative sample of households was selected in multiple stages and subjects aged 20-59 years were interviewed. Leisure-time physical activity was evaluated using the International Physical Activity Questionnaire. Data on adolescent physical activity were based on subjects' recall. RESULTS: Of 2,577 subjects interviewed, 27.5% were classified as adequately active, and 54.9% reported regular physical activity in adolescence. Subjects who engaged in regular physical activity during adolescence were more likely to be adequately active in adulthood (adjusted prevalence ratio 1.42; 95% CI: 1.23; 1.65). This effect was stronger in women (adjusted prevalence ratio: 1.51; 95% CI: 1.22; 1.86) than men (adjusted prevalence ratio: 1.35; 95% CI: 1.10; 1.67). CONCLUSIONS: Promoting physical activity in school age may be a successful intervention against the epidemic of adult inactivity. Although women were less likely to report regular physical activity in adolescence, the effect of this experience on adult behavior was stronger than in men.

Physical Activity, Obesity Status, and Blood Pressure in Preschool Children

Objective To examine the combined effects of physical activity and weight status on blood pressure (BP) in preschool-aged children. Study design The sample included 733 preschool-aged children (49% female). Physical activity was objectively assessed on 7 consecutive days by accelerometry. Children were categorized as sufficiently active if they met the recommendation of at least 60 minutes daily of moderate-to-vigorous physical activity (MVPA). Body mass index was used to categorize children as nonoverweight or overweight/obese, according to the International Obesity Task Force benchmarks. BP was measured using an automated BP monitor and categorized as elevated or normal using BP percentile-based cut-points for age, sex, and height. Results The prevalence of elevated systolic BP (SBP) and diastolic BP was 7.7% and 3.0%, respectively. The prevalence of overweight/obese was 32%, and about 15% of children did not accomplish the recommended 60 minutes of daily MVPA. After controlling for age and sex, overweight/obese children who did not meet the daily MVPA recommendation were 3 times more likely (OR 3.8; CI 1.6-8.6) to have elevated SBP than nonoverweight children who met the daily MVPA recommendation. Conclusions Overweight or obese preschool-aged children with insufficient levels of MVPA are at significantly greater risk for elevated SBP than their non overweight and sufficiently active counterparts. (J Pediatr 2015;167:98-102).

Everybody active, every day: An evidence-based approach to physical activity

‘Everybody active, every day’ is a national, evidence-based approach to support all sectors to embed physical activity into the fabric of daily life and make it an easy, cost-effective and ‘normal’ choice in every community in England.PHE has co-produced the framework with over 1,000 national and local leaders in physical activity and is calling for action from providers and commissioners in: health, social care, transportation, planning, education, sport and leisure, culture, the voluntary and community sector, as well as public and private employers.To make active lifestyles a reality for all, the framework’s 4 areas for action will:change the social ‘norm’ to make physical activity the expectationdevelop expertise and leadership within professionals and volunteerscreate environments to support active livesidentify and up-scale successful programmes nationwide‘Everybody active, every day’ is part of the cross-government ‘Moving More, Living More’ campaign for a more active nation as part of the 2012 Olympic and Paralympic Games legacy.

Assessment of free-living physical activity in humans: an overview of currently available and proposed new measures.

The number of physical activity measures and indexes used in the human literature is large and may result in some difficulty for the average investigator to choose the most appropriate measure. Accordingly, this review is intended to provide information on the utility and limitations of the various measures. Its primary focus is the objective assessment of free-living physical activity in humans based on physiological and biomechanical methods. The physical activity measures have been classified into three categories: Measures based on energy expenditure or oxygen uptake, such as activity energy expenditure, activity-related time equivalent, physical activity level, physical activity ratio, metabolic equivalent, and a new index of potential interest, daytime physical activity level. Measures based on heart rate monitoring, such as net heart rate, physical activity ratio heart rate, physical activity level heart rate, activity-related time equivalent, and daytime physical activity level heart rate. Measures based on whole-body accelerometry (counts/U time). Quantification of the velocity and duration of displacement in outdoor conditions by satellites using the Differential Global Positioning System may constitute a surrogate for physical activity, because walking is the primary activity of man in free-living conditions. A general outline of the measures and indexes described above is presented in tabular form, along with their respective definition, usual applications, advantages, and shortcomings. A practical example is given with typical values in obese and non-obese subjects. The various factors to be considered in the selection of physical activity methods include experimental goals, sample size, budget, cultural and social/environmental factors, physical burden for the subject, and statistical factors, such as accuracy and precision. It is concluded that no single current technique is able to quantify all aspects of physical activity under free-living conditions, requiring the use of complementary methods. In the future, physical activity sensors, which are of low-cost, small-sized, and convenient for subjects, investigators, and clinicians, are needed to reliably monitor, during extended periods in free-living situations, small changes in movements and grade as well as duration and intensity of typical physical activities.

Could a satellite-based navigation system (GPS) be used to assess the physical activity of individuals on earth?

OBJECTIVES: To test whether the Global Positioning System (GPS) could be potentially useful to assess the velocity of walking and running in humans. SUBJECT: A young man was equipped with a GPS receptor while walking running and cycling at various velocity on an athletic track. The speed of displacement assessed by GPS, was compared to that directly measured by chronometry (76 tests). RESULTS: In walking and running conditions (from 2-20 km/h) as well as cycling conditions (from 20-40 km/h), there was a significant relationship between the speed assessed by GPS and that actually measured (r = 0.99, P < 0.0001) with little bias in the prediction of velocity. The overall error of prediction (s.d. of difference) averaged +/-0.8 km/h. CONCLUSION: The GPS technique appears very promising for speed assessment although the relative accuracy at walking speed is still insufficient for research purposes. It may be improved by using differential GPS measurement.

Physical activity, bodyweight, health and fear of negative evaluation in primary school children.

Fear of negative evaluation (FNE) is regarded as being the core feature of social anxiety. The present study examined how FNE is associated with physical activity (PA), body mass index (BMI) and perceived physical health (PPH) in children. Data were collected in a sample of 502 primary school children in first and fifth grades taking part in a randomized-controlled trial ("Kinder-Sportstudie KISS") aimed at increasing PA and health. PA was assessed by accelerometry over 7 days, PPH by the Child Health Questionnaire and FNE by the Social Anxiety Scale for Children--Revised. BMI z-scores were calculated based on Swiss norms. Cross-sectional analyses indicated that children high in FNE exercised less, reported lower levels of PPH and had higher BMI z-scores (P&lt;0.01). Using mixed linear models, the school-based PA intervention did not manage to reduce FNE scores. Overweight children demonstrated a greater increase in FNE (P&lt;0.05) indicating that enhanced weight may be a risk factor for FNE. In conclusion, the associations among high FNE, low PA and increased BMI should be considered when promoting an active lifestyle in children.

Assessment of Physical Activity and Energy Expenditure by GPS Combined With Accelerometry in Real-Life Conditions.

Background: Physical activity (PA) and related energy expenditure (EE) is often assessed by means of a single technique. Because of inherent limitations, single techniques may not allow for an accurate assessment both PA and related EE. The aim of this study was to develop a model to accurately assess common PA types and durations and thus EE in free-living conditions, combining data from global positioning system (GPS) and 2 accelerometers. Methods: Forty-one volunteers participated in the study. First, a model was developed and adjusted to measured EE with a first group of subjects (Protocol I, n = 12) who performed 6 structured and supervised PA. Then, the model was validated over 2 experimental phases with 2 groups (n = 12 and n = 17) performing scheduled (Protocol I) and spontaneous common activities in real-life condition (Protocol II). Predicted EE was compared with actual EE as measured by portable indirect calorimetry. Results: In protocol I, performed PA types could be recognized with little error. The duration of each PA type could be predicted with an accuracy below 1 minute. Measured and predicted EE were strongly associated (r = .97, P < .001). Conclusion: Combining GPS and 2 accelerometers allows for an accurate assessment of PA and EE in free-living situations.

Data-based approach for developing a physical activity frequency questionnaire.

Measurement of total energy expenditure may be crucial to an understanding of the relation between physical activity and disease and in order to frame public health intervention. To devise a self-administered physical activity frequency questionnaire (PAFQ), the following data-based approach was used. A 24-hour recall was administered to a random sample of 919 adult residents of Geneva, Switzerland. The data obtained were used to establish the list of activities (and their median duration) that contributed to 95% of the energy expended, separately for men and women. Activities that were trivial for the whole sample but that contributed to > or = 10% of an individual's energy expenditure were also selected. The final PAFQ lists 70 activities or group of activities with their typical duration. About 20 minutes are required for respondents to indicate the number of days and the number of hours per day that they performed each activity. The PAFQ method was validated against a heart rate monitor, a more objective method. The total energy estimated by the PAFQ in 41 volunteers correlated well (r = 0.76) with estimates using a heart rate monitor. The authors conclude that the design of their self-administered physical activity frequency questionnaire based on data from 24-hour recall appeared to accurately estimate energy expenditure.