Machine Learning Models for Weight-Bearing Activity Type Recognition Based on Accelerometry in Postmenopausal Women.

Hip-worn triaxial accelerometers are widely used to assess physical activity in terms of energy expenditure. Methods for classification in terms of different types of activity of relevance to the skeleton in populations at risk of osteoporosis are not currently available. This publication aims to assess the accuracy of four machine learning models on binary (standing and walking) and tertiary (standing, walking, and jogging) classification tasks in postmenopausal women. Eighty women performed a shuttle test on an indoor track, of which thirty performed the same test on an indoor treadmill. The raw accelerometer data were pre-processed, converted into eighteen different features and then combined into nine unique feature sets. The four machine learning models were evaluated using three different validation methods. Using the leave-one-out validation method, the highest average accuracy for the binary classification model, 99.61%, was produced by a k-NN Manhattan classifier using a basic statistical feature set. For the tertiary classification model, the highest average accuracy, 94.04%, was produced by a k-NN Manhattan classifier using a feature set that included all 18 features. The methods and classifiers within this study can be applied to accelerometer data to more accurately characterize weight-bearing activity which are important to skeletal health.

Implementing Exercise in Standard Cancer Care (Bizi Orain Hybrid Exercise Program): Protocol for a Randomized Controlled Trial.

BACKGROUND: Despite the established benefits of regular exercise for patients with cancer to counteract the deleterious effects of the disease itself and treatment-related adverse effects, most of them do not engage in sufficient levels of physical activity and there is a paucity of data on the integration of efficacious exercise programs that are accessible and generalizable to a large proportion of patients with cancer into routine cancer care. OBJECTIVE: We intend to examine the effects attributable to the implementation of a community-based exercise program on cardiorespiratory functional capacity and quality of life for patients with cancer. METHODS: This will be a hybrid study. In the first experimental phase, patients diagnosed with any type of cancer will be randomized into two parallel groups. One group immediately performs Bizi Orain, a 3-month supervised exercise program (3 times a week), in addition to behavioral counseling in a primary health care setting; the other is a reference group that starts the exercise program 3 months later (delayed treatment). In the second observational phase, the entire cohort of participants will be followed-up for 5 years. Any person diagnosed with cancer in the previous 2 years is eligible for the program. The program evaluation involves the uptake, safety, adherence, and effectiveness assessed after completion of the program and with follow-ups at 3, 6, 12, 24, 36, 48, and 60 months. The primary outcomes of the experimental study, to be compared between groups, are improved physical function and quality of life, whereas overall survival is the main objective of the prospective study. To analyze the association between changes in physical activity levels and overall survival, longitudinal mixed-effects models will be used for repeated follow-up measures. RESULTS: A total of 265 patients have been enrolled into the study since January 2019, with 42 patients from the hematology service and 223 from the oncology service. CONCLUSIONS: Bizi Orain is the first population-based exercise program in Spain that will offer more insight into the implementation of feasible, generalizable, and sustainable supportive care services involving structured exercise to extend survival of patients with cancer, improve their physical function and quality of life, and reverse the adverse effects of their disease and related treatments, thereby reducing the clinical burden. TRIAL REGISTRATION: ClinicalTrials.gov NCT03819595; http://clinicaltrials.gov/ct2/show/NCT03819595. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/24835.

Investigation of the Relationship Between Peak Vertical Accelerations and Aerobic Exercise Intensity During Graded Walking and Running in Postmenopausal Women.

How exercise intensity targets, calibrated according to oxygen consumption, relate to vertical impacts during weight-bearing exercise is currently unknown. The authors investigated the relationship between vertical peaks (VPs) and metabolic equivalents (METs) of oxygen consumption in 82 women during walking and running. The magnitude of VPs, measured using a hip-worn triaxial accelerometer, was derived from recommended aerobic exercise intensity targets. VPs were 0.63 ± 0.18g at the lower recommended absolute exercise intensity target (3 METs) but >1.5g at the upper end of moderate-intensity activities (1.90 ± 1.13g at 6 METs). Multilevel linear regression analyses identified speed and type of locomotion as the strongest independent predictors of VPs, explaining 54% and 11% of variance, respectively. The authors conclude that, in contrast to lower intensities, exercising close to or above the 6-MET threshold generates VPs of osteogenic potential, suggesting this could provide simultaneous benefits to decrease all-cause mortality and osteoporosis risk.

Exercise-Intensity Adherence During Aerobic Training and Cardiovascular Response During Resistance Training in Cancer Survivors.

ABSTRACT: Quevedo-Jerez, K, Gil-Rey, E, Maldonado-Martín, S, and Herrero-Román, F. Exercise-intensity adherence during aerobic training and cardiovascular response during resistance training in cancer survivors. J Strength Cond Res 35(8): 2338-2345, 2021-Combined aerobic-resistance training has shown the best benefits has proved beneficial for cancer survivors (CS). To understand the adherence to the aerobic training program (in terms of the intensity and duration of the sessions) and the cardiovascular response to the resistance training program, heart rate (HR) of 48 CS was monitored in each training session with an HR monitor for a 2-year period. During aerobic training, CS had to maintain the intensity in zone 2 (Z2) (between the ventilatory threshold and respiratory compensation point). The time spent below Z2 (Z1), in Z2, and above Z2 (Z3) was assessed in both aerobic and resistance training. The exercise-intensity distribution (aerobic vs. resistance training) was as follows: Z1 6.6 ± 12.8% vs. 34.3 ± 29.9% (p < 0.001); Z2 66.6 ± 29.3% vs. 54.5 ± 27.6% (p < 0.05); and Z3 26.9 ± 29.9% vs. 11.2 ± 20.6% (p < 0.001). The most deconditioned CS (<4.5 metabolic equivalents [METs]) presented the poorest adherence in Z2 and spent the most time in Z3. A significant positive moderate-high correlation was found for the percentage of time in Z3 between resistance and aerobic exercise (r = 0.75, p < 0.001). In conclusion, the individualization of exercise intensity resulted in good adherence to the prescribed intensity. Less fit CS needed more supervision in their training sessions. Resistance training allowed the CS to train in moderate-vigorous intensities of cardiovascular response. Resistance training should have more scope in exercise prescriptions, particularly in deconditioned CS and in the first steps of exercise programs.

Individualized Accelerometer Activity Cut-Points for the Measurement of Relative Physical Activity Intensity Levels.

Purpose: The aim of this study was to compare the widely used accelerometer activity cut-points derived from the absolute moderate intensity recommendation (3‒6 METs), with relative intensity cut-points according to maximal cardiorespiratory fitness (46%‒63% V˙O2max ) and to individual lactate thresholds (LT1 and LT2) in postmenopausal women. Method: Thirty postmenopausal women performed several exercise tests with measures of heart rate, blood lactate, accelerometer activity counts and oxygen consumption. Individual regressions were developed to derive the accelerometer activity counts at absolute and relative moderate intensity recommendations and at individual LTs. Results: The activity counts calculated at the lower moderate intensity boundary were lower for the absolute 3 METs threshold (2026 ± 808 ct·min-1) compared to relative 46 % V˙O2max intensity threshold (p < .01, ES: 1.95) and LT1 (p < .01, ES: 2.27), which corresponded to 4.6 ± 0.7 METs. The activity counts at the upper moderate intensity boundary were higher for LT2 (7249 ± 2499 ct·min-1) compared to the absolute 6 METs threshold (p < .01, ES: 0.72) and relative 63% V˙O2max intensity threshold (p < .01, ES: 0.55). The interindividual variability in activity counts at relative intensity thresholds was high (CV = 30-34%), and was largely explained by cardiorespiratory fitness level (R2 = ~ 50%). Conclusion: Individually tailored (relative to V˙O2max or submaximal LTs) rather than fixed accelerometer intensity cut-points derived from the classic absolute moderate physical activity intensity (3-6 METs) would result in a more accurate measurement of an individual´s activity levels and reduce the risk of overestimating or underestimating physical activity.

Objectively measured absolute and relative physical activity intensity levels in postmenopausal women.

OBJECTIVES: To investigate how objectively measured physical activity (PA) levels differ according to absolute moderate intensity recommendation (3-6 METs) and relative to individual lactate thresholds (LT1 and LT2), and to verify if high-fit women record higher PA levels compared to women with lower aerobic fitness. METHODS: Seventy-five postmenopausal women performed an incremental exercise test and several constant-velocity tests wearing an accelerometer to identify the activity counts (ct min-1) corresponding to LT1 and LT2. Individual linear regression determined activity counts cut-points for each intensity: (1) sedentary (<200 ct min-1), (2) light (from 200 ct min-1 to ct min-1 at LT1), (3) moderate (ct min-1 between LT1 and LT2) and (4) vigorous (ct min-1 > LT2). Participants then wore an accelerometer during a week to measure the time spent at each PA intensity level. RESULTS: According to absolute intensity categorisation, high-fit postmenopausal women recorded twice as much time at moderate-to-vigorous PA (MVPA) (P < 0.01) than low-fit women. However, when PA intensity was calculated relative to individual lactate thresholds, MVPA was significantly reduced by half (P < 0.01) and the data revealed no differences (P = 0.62) between groups (∼20 min day-1 at MVPA). CONCLUSIONS: Accelerometer cut-points derived from absolute moderate-intensity recommendations overestimated MVPA. Similar time at MVPA was recorded by high- and low-fit postmenopausal women when the cut-points were tailored to individual lactate thresholds. A more accurate estimation of PA behaviour could be provided with the use of individually tailored accelerometer cut-points.

Quantification of the perceived training load and its relationship with changes in physical fitness performance in junior soccer players.

The aim of this study was to determine the relationship between perceived respiratory and muscular training load (TL) and changes in physical fitness in elite and non-elite junior soccer players. Twenty-eight elite (n = 14, 17.6 ± 0.6 years, 70.3 ± 4.4 kg, 179.7 ± 5.6 cm) and non-elite (n = 14, 17.5 ± 0.5 years, 71.1 ± 6.5 kg, 178.1 ± 5.6 cm) soccer players belonging to a Spanish first and third division football academies and competing in junior Spanish first division (2012-2013) participated in the study. Countermovement jump (CMJ), CMJ arm swing, 5 and 15 m sprints and the Université de Montreal endurance test were performed in January and 9 weeks later in March. In order to quantify TLs, after each training session and match, players reported their session rating of perceived exertion (sRPE) separately for respiratory (sRPEres) and leg musculature (sRPEmus). Elite players accumulated greater weekly training volume (361 ± 14 vs. 280 ± 48 min; effect sizes (ES) = 5.23 ± 1.74; most likely), and perceived respiratory (1460 ± 184 vs. 1223 ± 260 AU; ES = 1.12 ± 0.79; very likely) and muscular (1548 ± 216 vs. 1318 ± 308 AU; ES = 0.99 ± 0.84; likely) TL than did non-elite players. Training volume, sRPEres-TL and sRPEmus-TL were positively and largely correlated (r = 0.67-0.71) with the changes in aerobic fitness. The present results suggest that a low training volume and TL can impair improvement in aerobic fitness in junior soccer players during the in-season period.