Search

Morning, day, or evening chronotypes differ by the circadian timing of alertness and the preferred timing of sleep. It has been suggested that evening chronotype is associated with low physical activity (PA) and high sedentary time (SED). Our aim was to investigate whether such an association is confirmed by objectively measured PA and SED. In 46-year follow-up of the Northern Finland Birth Cohort 1966 study, total PA (MET min/day) and SED (min/day) among 5156 participants were determined using wrist-worn accelerometers for 14 days. We used the shortened Morningness-Eveningness Questionnaire to define participants' chronotypes. As covariates, we used self-reported physical strenuousness of work, health, and demographics, and clinical measures. We used adjusted general linear models (B coefficients with 95% confidence intervals, CI) to analyze how chronotype was related to total PA or SED. As compared to evening chronotype, men with day and morning chronotypes had higher total PA volumes (adjusted B 75.2, 95% CI [8.1, 142.4], P = .028, and 98.6, [30.2, 167.1], P = .005). Men with day and morning chronotypes had less SED (-35.8, [-53.8, 17.8], P < .0001, and - 38.6, [-56.9, -20.2], P < .0001). Among women, morning chronotype was associated with higher total PA (57.8, [10.5, 105.0], P = .017), whereas no association between chronotype and SED emerged. Evening chronotype was associated with low objectively measured PA in both sexes and with high SED in men, even after adjustments for established potential confounders. Chronotype should be considered in PA promotion.

Refereed/Peer-reviewed Purpose: This study aimed to identify and characterize joint profiles of sedentary time and physical activity among adults and to investigate how these profiles are associated with markers of cardiometabolic health. Methods: The participants included 3702 of the Northern Finland Birth Cohort 1966 at age 46 yr, who wore a hip-worn accelerometer during waking hours and provided seven consecutive days of valid data. Sedentary time, light-intensity physical activity, and moderate- to vigorous-intensity physical activity on each valid day were obtained, and a data-driven clustering approach (“KmL3D”) was used to characterize distinct joint profiles of sedentary time and physical activity intensities. Participants self-reported their sleep duration and performed a submaximal step test with continuous heart rate measurement to estimate their cardiorespiratory fitness (peak heart rate). Linear regression was used to determine the association between joint profiles of sedentary time and physical activities with cardiometabolic health markers, including adiposity markers and blood lipid, glucose, and insulin levels. Results: Four distinct groups were identified: “active couch potatoes” (n = 1173), “sedentary light movers” (n = 1199), “sedentary exercisers” (n = 694), and “movers” (n = 636). Although sufficiently active, active couch potatoes had the highest daily sedentary time (>10 h) and lowest light-intensity physical activity. Compared with active couch potatoes, sedentary light movers, sedentary exercisers, and movers spent less time in sedentary by performing more physical activity at light-intensity upward and had favorable differences in their cardiometabolic health markers after accounting for potential confounders (1.1%–25.0% lower values depending on the health marker and profile). Conclusions: After accounting for sleep duration and cardiorespiratory fitness, waking activity profiles characterized by performing more physical activity at light-intensity upward, resulting in less time spent in sedentary, were associated with better cardiometabolic health.

Introduction : Physical inactivity, excessive total time spent in sedentary behavior (SB) and prolonged sedentary bouts have been proposed to be risk factors for chronic disease morbidity and mortality worldwide. However, which patterns and postures of SB have the most negative impacts on health outcomes is still unclear. This population-based study aimed to investigate the independent associations of the patterns of accelerometer-based overall SB and sitting with serum lipid biomarkers at different moderate- to vigorous-intensity physical activity (MVPA) levels. Methods: Physical activity and SB were measured in a birth cohort sample (N = 3272) at 46 yr using a triaxial hip-worn accelerometer in free-living conditions for 14 d. Raw acceleration data were classified into SB and PA using a machine learning–based model, and the bouts of overall SB and sitting were identified from the classified data. The participants also answered health-related questionnaires and participated in clinical examinations. Associations of overall SB (lying and sitting) and sitting patterns with serum lipid biomarkers were investigated using linear regression. Results: The overall SB patterns were more consistently associated with serum lipid biomarkers than the sitting patterns after adjustments. Among the participants with the least and the most MVPA, high total time spent in SB and SB bouts of 15–29.99 and ≥30 min were associated with impaired lipid metabolism. Among those with moderate amount of MVPA, higher time spent in SB and SB bouts of 15–29.99 min was unfavorably associated with serum lipid biomarkers. Conclusions: The associations between SB patterns and serum lipid biomarkers were dependent on MVPA level, which should be considered when planning evidence-based interventions to decrease SB in midlife.

The purpose of this study was to describe the accelerometry-based characteristics of overall sedentary behavior (SB) and sitting among adults under free-living conditions. Thirty-six (mean age 47.6 years) volunteers carried a waist-worn accelerometer for ≥4 days with data ≥600 min/d during 14 consecutive days. A machine learning (ML) based method was used to classify the patterns of SB and sitting from raw 3D acceleration. The participants spent most (69.3%) of their waking time in SB, and a half (52.2%) of the SB was performed in a sitting posture. Men broke their overall sedentary time less often (4.1 vs. 6.1 bouts/h), but women sat more; however, women broke their sitting time as often as men (7.6 bouts/h). This study confirms that SB and sitting can be distinguished using ML methods, and more information about SB can be achieved when overall SB and sitting are analyzed separately in free-living conditions.

Study design Population-based birth cohort study. Objective The aim of this study was to evaluate the relationship between vertebral dimensions and lumbar MC. Summary of background data Low back pain (LBP) has become the leading cause of disability worldwide. Modic changes (MC) of the lumbar spine are one potential LBP-associated etiological factor. Mechanical stress is considered to play a key role in the development of MC through damage to endplates. There is speculation that vertebral dimensions play a role in some degenerative changes in the spine. Previous studies have also shown a positive association between moderate-to-vigorous physical activity (MVPA) and both vertebral dimensions and MC. In this study, we aimed to evaluate the relationship between vertebral dimensions and MC. Methods The study population consisted of 1221 participants from the Northern Finland Birth Cohort 1966 who underwent lumbar magnetic resonance imaging (MRI) and physical activity measurements at the age of 46-48. The presence of Type 1 (MC1) and Type 2 (MC2) MC and the height, axial cross-sectional area (CSA), and volume of the L4 vertebra were determined from MRI scans. MVPA (≥3.5 metabolic equivalents) was measured by a wrist-worn accelerometer. We analyzed the association between lumbar MC and vertebral height, CSA, and volume using logistic regression models before and after adjustment for sex, height, weight, smoking, education level, and MVPA. Results Vertebral height was positively associated with the presence of MC2 (odds ratio [OR] 3.51; 95% confidence interval [CI] 1.43-8.65), whereas vertebral CSA was not associated with the presence of lumbar MC. Vertebral volume was positively associated with the presence of any MC (OR 1.04; 95% CI 1.00-1.07), but the association did not persist when analyzing MC1 and MC2 separately. Conclusion Vertebral height was associated with the presence of MC2. Further studies are needed to clarify the role of vertebral dimensions as independent risk factors for MC.Level of Evidence: 3.

This study investigated whether the timing of infant motor development is associated with self-reported and accelerometer-measured physical activity (PA) and sedentary time (ST) in midlife. This population-based study consisted of 4098 people born in 1966 in Northern Finland (NFBC 1966). Data on nine infant motor developmental milestones included making sounds, holding up the head, grabbing objects, turning from back to tummy, sitting without support, standing with support, walking with support, standing without support, and walking without support. At the age of 46, PA at leisure time and sitting time was self-reported. PA and ST were also measured with a wrist-worn Polar Active accelerometer that was instructed to be worn on the non-dominant hand 24 h/d for 14 days. A multiple linear regression analysis was used to analyze the association between infant motor development and PA and ST in midlife. Later infant motor development was weakly associated with higher accelerometer-measured light PA, but not with moderate-to-vigorous PA. Later infant locomotor development was associated with lower accelerometer-measured ST (β -0.07, p = 0.012) and lower self-reported sitting time at work (β -0.06, p = 0.004) in women. In conclusion, later infant motor development was associated with higher light PA and lower sedentary time at middle age. PA is a multifactorial behavior influenced by various factors from early childhood to midlife. Further research is required before more general conclusions can be drawn.

Background: Recently, the importance of light physical activity (LPA) for health has been emphasized, and residential greenness has been positively linked to the level of LPA and a variety of positive health outcomes. However, people spend less time in green environments because of urbanization and modern sedentary leisure activities. Aims: In this population-based study, we investigated the association between objectively measured residential greenness and accelerometry measured physical activity (PA), with a special interest in LPA and gender differences. Methods: The study was based on the Northern Finland Birth Cohort 1966 (5433 members). Participants filled in a postal questionnaire and underwent clinical examinations and wore a continuous measurement of PA with wrist-worn Polar Active Activity Monitor accelerometers for two weeks. The volume of PA (metabolic equivalent of task or MET) was used to describe the participant&rsquo, s total daily activity (light: 2&ndash, 3.49 MET, moderate: 3.5&ndash, 4.99 MET, vigorous: 5&ndash, 7.99 MET, very vigorous: &ge, 8 MET). A geographic information system (GIS) was used to assess the features of each individual&rsquo, s residential environment. The normalized difference vegetation index (NDVI) was used for the objective quantification of residential greenness. Multiple linear regression and a generalized additive model (GAM) were used to analyze the association between residential greenness and the amount of PA at different intensity levels. Results: Residential greenness (NDVI) was independently associated with LPA (unadjusted &beta, = 174, CI = 140, 209) and moderate physical activity (MPA) (unadjusted &beta, = 75, CI = 48, 101). In the adjusted model, residential greenness was positively and significantly associated with LPA (adjusted &beta, = 70, CI = 26, 114). In men, residential greenness was positively and significantly associated with LPA (unadjusted &beta, = 224, CI = 173, 275), MPA (unadjusted &beta, CI = 48, 101), and moderate to vigorous physical activity (MVPA) (unadjusted &beta, = 89, CI = 25, 152). In women, residential greenness was positively related to LPA (unadjusted &beta, = 142, CI = 96, 188) and inversely associated with MPA (unadjusted &beta, = &minus, 22, CI = &minus, 36, &minus, 8), vigorous/very vigorous physical activity (VPA/VVPA) (unadjusted &beta, 49, 84, &minus, 14), and MVPA (unadjusted &beta, 71, 113, &minus, 29). In the final adjusted models, residential greenness was significantly associated only with the amount of LPA in men (adjusted &beta, = 140, CI = 75, 204). Conclusions: Residential greenness was positively associated with LPA in both genders, but the association remained significant after adjustments only in men. Residential greenness may provide a supportive environment for promoting LPA.

Vertebral fractures reduce the quality of life and are a major burden to the health care sector. Small vertebral size is associated with increased vertebral fracture risk. Previous studies have investigated the relationship between physical activity (PA) and vertebral size, but their results seem somewhat contradictory. In this population-based birth cohort study, we aimed to evaluate the relationship between objectively measured PA and vertebral size.The study population consisted of 1202 cohort participants who underwent PA and vertebral size measurements at the age of 46 to 48 yr. Moderate-to-vigorous PA (MVPA, ≥3.5 METs) was measured by a wrist-worn accelerometer (Polar Active; Polar Electro, Finland) for 14 d. The vertebral axial cross-sectional area (CSA) of the L4 vertebra was measured and calculated from lumbar magnetic resonance imaging scans at 46 to 48 yr. We analyzed the association between the daily amount of MVPA (min·d) and vertebral CSA using multivariable linear regression analysis.The daily amount of MVPA was significantly and positively associated with CSA in both sexes. For every minute per day of MVPA, men had 0.71 mm (95% confidence interval, 0.36-1.06) and women 0.90 mm (95% confidence interval, 0.58-1.21) larger CSA.Physical activity of at least moderate intensity is positively associated with vertebral size and may thus prevent future vertebral fractures.

Cardiorespiratory fitness (CRF) and physical activity (PA) are associated with autonomic function, but their associations to orthostatic autonomic responses are unclear in epidemiological setting. We hypothesized that higher CRF and PA would associate with higher immediate vagal responses and lower incidence of adverse findings during orthostatic test. At age of 46, 787 men and 938 women without cardiorespiratory diseases and diabetes underwent an orthostatic test (3-minutes sitting, 3-minutes standing) with recording of RR intervals (RRi) and blood pressure (BP) by finger plethysmography. Acute responses of RRi (30:15 ratio) and BP were calculated. CRF was measured by a submaximal step test and daily amount of moderate-to-vigorous PA (MVPA) for 2 weeks by wrist-worn accelerometer. Lifelong PA was based on questionnaires at ages of 14, 31, and 46. High CRF was significantly associated with higher RRi 30:15 ratio (adjusted standardized β = 0.17, P < 0.001) and milder acute decrease of systolic BP while standing (β = 0.10, P = 0.001), while MVPA was not (β = 0.04 for RRi 30:15 ratio and β = 0.05 for systolic BP acute response). High lifelong PA was significantly associated with higher RRi 30:15 ratio (β = 0.08, P = 0.002) but not with acute systolic BP response. Those in the lowest tertile of CRF had 9.2-fold risk (P = 0.002) of having postural orthostatic tachycardia syndrome compared to more fit. Cardiorespiratory fitness and lifelong physical activity, but not current physical activity, were independently associated with higher cardiac vagal response to orthostasis. The present results underscore the importance fitness and lifelong physical activity in prevention of abnormal autonomic function and related cardiovascular risk.

Purpose A data mining approach was applied to establish a multilevel hierarchy predicting physical activity (PA) behavior, and to methodologically identify the correlates of PA behavior. Methods Cross-sectional data from the population-based Northern Finland Birth Cohort 1966 study, collected in the most recent follow-up at age 46, were used to create a hierarchy using the chi-square automatic interaction detection (CHAID) decision tree technique for predicting PA behavior. PA behavior is defined as active or inactive based on machine-learned activity profiles, which were previously created through a multidimensional (clustering) approach on continuous accelerometer-measured activity intensities in one week. The input variables (predictors) used for decision tree fitting consisted of individual, demographical, psychological, behavioral, environmental, and physical factors. Using generalized linear mixed models, we also analyzed how factors emerging from the model were associated with three PA metrics, including daily time (minutes per day) in sedentary (SED), light PA (LPA), and moderate-to-vigorous PA (MVPA), to assure the relative importance of methodologically identified factors. Results Of the 4582 participants with valid accelerometer data at the latest follow-up, 2701 and 1881 had active and inactive profiles, respectively. We used a total of 168 factors as input variables to classify these two PA behaviors. Out of these 168 factors, the decision tree selected 36 factors of different domains from which 54 subgroups of participants were formed. The emerging factors from the model explained minutes per day in SED, LPA, and/or MVPA, including body fat percentage (SED: B = 26.5, LPA: B = − 16.1, and MVPA: B = − 11.7), normalized heart rate recovery 60 s after exercise (SED: B = -16.1, LPA: B = 9.9, and MVPA: B = 9.6), average weekday total sitting time (SED: B = 34.1, LPA: B = -25.3, and MVPA: B = -5.8), and extravagance score (SED: B = 6.3 and LPA: B = − 3.7). Conclusions Using data mining, we established a data-driven model composed of 36 different factors of relative importance from empirical data. This model may be used to identify subgroups for multilevel intervention allocation and design. Additionally, this study methodologically discovered an extensive set of factors that can be a basis for additional hypothesis testing in PA correlates research.

Background Regular physical activity (PA) promotes health and decreases mortality. The positive relationship between PA and perceived health (PH) is well known. However, previous research in the field has often used self-reported PA measures. The aim of this population-based NFBC1966 birth cohort study was to assess the relationship between both self-reported and objectively measured PA and PH in midlife. Methods A sample group of 6384 participants (2878 men, 3506 women, response rate 62%) aged 46 completed a questionnaire on PH and health behaviors, including items on weekly leisure time physical activity (LTPA) and daily sitting time (ST). PH was dichotomized as good (very good or good) and other (fair, poor, or very poor). PA was measured with a wrist-worn Polar Active (Polar Electro, Finland) accelerometer for 14 days (n = 5481, 98%) and expressed as daily average time spent in moderate to vigorous intensity PA (MVPA). Odds ratios (OR) and 95% confidence intervals for good PH were calculated using binary logistic regression and adjusted for relevant demographic, socioeconomic, and health characteristics, and ST. Results The level of PA was positively associated with PH after adjustments with covariates and ST. There was a dose-response relationship across the PA quartiles according to the adjusted multivariable models. Self-reported LTPA was more strongly associated with good PH (OR from 1.72 to 4.33 compared to lowest PA quartile) than objectively measured PA (OR from 1.37 to 1.66 compared to lowest PA quartile). Conclusions In this large population-based birth cohort study, we for the first time show a positive dose-response relationship of both self-reported and objectively measured PA to PH, the relationship being stronger for self-reported LTPA. Despite the cross-sectional design of this study, the results from this large sample suggest that both self-reported and objectively measured physical activity are strongly associated with PH, which is a predictor of morbidity and mortality, and regular PA should be encouraged in midlife.

Background Objective measures using accelerometer-based activity monitors have been extensively used in physical activity (PA) and sedentary behavior (SB) research. To measure PA and SB precisely, the field is shifting towards machine learning-based (ML) approaches for calibration and validation of accelerometer-based activity monitors. Nevertheless, various parameters regarding the use and development of ML-based models, including data type (raw acceleration data versus activity counts), sampling frequency, window size, input features, ML technique, accelerometer placement, and free-living settings, affect the predictive ability of ML-based models. The effects of these parameters on ML-based models have remained elusive, and will be systematically reviewed here. The open challenges were identified and recommendations are made for future studies and directions. Method We conducted a systematic search of PubMed and Scopus databases to identify studies published before July 2017 that used ML-based techniques for calibration and validation of accelerometer-based activity monitors. Additional articles were manually identified from references in the identified articles. Results: A total of 62 studies were eligible to be included in the review, comprising 48 studies that calibrated and validated ML-based models for predicting the type and intensity of activities, and 22 studies for predicting activity energy expenditure. Conclusions It appears that various ML-based techniques together with raw acceleration data sampled at 20–30 Hz provide the opportunity of predicting the type and intensity of activities, as well as activity energy expenditure with comparable overall predictive accuracies regardless of accelerometer placement. However, the high predictive accuracy of laboratory-calibrated models is not reproducible in free-living settings, due to transitive and unseen activities together with differences in acceleration signals.

Purpose: Although low cardiorespiratory fitness (CRF), physical inactivity, and obesity are associated with impaired autonomic function, they are also extensively interrelated. The present study aimed to assess the extent to which they contribute to autonomic function independently of each other. Methods: At the age of 46 yr, 1383 men and 1761 women without cardiorespiratory diseases and diabetes underwent assessments of vagally mediated heart rate (HR) variability (root mean square of successive differences in R-R interval (rMMSD)), peak HR during a submaximal step test (CRF), and 60-s HR recovery (HRR). Moderate-to-vigorous physical activity (MVPA; ≥3.5 METs, 2 wk) was measured by wrist-worn accelerometer and body fat percentage (Fat%) by bioimpedance Results: In men, CRF and Fat% were significantly associated with higher rMSSD (standardized β = 0.31 and -0.16) and HRR (β = 0.19 and -0.18), whereas higher MVPA was linked with higher HRR (β = 0.13) when including CRF, MVPA, and Fat% in the initial regression. After adjustments for other lifestyle and cardiometabolic factors, CRF remained significantly associated with rMMSD (β = 0.24) and HRR (β = 0.14), as did MVPA with HRR (β = 0.11). In women, CRF was associated with rMSSD (β = 0.23) and HRR (β = 0.15), and MVPA (β = 0.17) and Fat% (β = -0.07) with HRR, when CRF, MVPA, and Fat% were adjusted for each other. After further adjustments, CRF remained a significant determinant of rMSSD (β = 0.20) and HRR (β = 0.13), as did MVPA with HRR (β = 0.15). The final models explained 23% and 21% of variation in rMSSD and HRR in men, and 10% and 12% in women, respectively. Conclusions: CRF was a more important determinant of cardiac autonomic function than MVPA and body fat. Furthermore, MVPA but not body fat was independently associated with cardiac autonomic function in both men and women.

The purpose of this study was to develop and validate an algorithm for classifying physical activity (PA) classes and sedentary behavior (SED) from raw acceleration signal measured from hip. Twenty-two adult volunteers completed a pre-defined set of controlled and supervised activities. The activities included nine daily PAs. The participants performed PA trials while wearing a hip-worn 3D accelerometer. Indirect calorimetry was used for measuring energy expenditure. The raw acceleration data were used for training and testing a prediction model in MATLAB environment. The prediction model was built using bagged trees classifier and the most suitable extracted features (mean, maximum, minimum, zero crossing rate, and mean amplitude deviation) were selected using a sequential forward selection method. Leave-one-out cross validation was used for validation. Activities were classified as lying, sitting, light PA (standing, table wiping, floor cleaning, slow walking), moderate PA (fast walking) and vigorous PA (soccer and jogging). The oxygen consumption data were used for estimating the intensity of measured PA. Total accuracy of the prediction model was 96.5%. Mean sensitivity of the model was 95.5% (SD 3.5) and mean specificity 99.1% (SD 0.5). Based on the results PA types can be classified from raw data of the hip-worn 3D accelerometer using supervised machine learning techniques with a high sensitivity and specificity. The developed algorithm has a potential for objective evaluations of PA and SED.

Supplemental digital content is available in the text., Purpose This study aimed to examine how compositions of 24-h time use and time reallocations between movement behaviors are associated with cardiometabolic health in a population-based sample of middle-age Finnish adults. Methods Participants were 3443 adults 46 yr of age from the Northern Finland Birth Cohort 1966 study. Participants wore a hip-worn accelerometer for 14 d from which time spent in sedentary behavior (SB), light-intensity physical activity (LPA), and moderate- to vigorous-intensity physical activity (MVPA) were determined. These data were combined with self-reported sleep to obtain the 24-h time-use composition. Cardiometabolic outcomes included adiposity markers, blood lipid levels, and markers of glucose control and insulin sensitivity. Multivariable-adjusted regression analysis, using a compositional data analysis approach based on isometric log-ratio transformation, was used to examine associations between movement behaviors with cardiometabolic outcomes. Results More daily time in MVPA and LPA, relative to other movement behaviors, was consistently favorably associated with all cardiometabolic outcomes. For example, relative to time spent in other behaviors, 30 min·d−1 more MVPA and LPA were both associated with lower 2-h post–glucose load insulin level (−11.8% and −2.7%, respectively). Relative to other movement behaviors, more daily time in SB was adversely associated with adiposity measures, lipid levels, and markers of insulin sensitivity, and more daily time asleep was adversely associated with adiposity measures, blood lipid, fasting plasma glucose, and 2-h insulin. For example, 60 min·d−1 more SB and sleep relative to the remaining behaviors were both associated with higher 2-h insulin (3.5% and 5.7%, respectively). Conclusions Altering daily movement behavior compositions to incorporate more MVPA at the expense of any other movement behavior, or more LPA at the expense of SB or sleep, could help to improve cardiometabolic health in midadulthood.

Physical activity (PA) and sedentary time (SED) are associated with the risk of cardiovascular disease (CVD), but the temporal patterns of these behaviors most beneficial for cardiovascular health remain unknown. We aimed to identify the intensity and temporal patterns of PA and SED measured continuously by an accelerometer and their relationship with CVD risk. At the age of 46 years, 4582 members (1916 men; 2666 women) of the Northern Finland Birth Cohort 1966 study underwent continuous measurement of PA with Polar Active (Polar Electro, Finland) accelerometers for one week. X-means clustering was applied based on 10 min average MET (metabolic equivalent) values during the measurement period. Ten-year risk of CVD was estimated using the Framingham risk model. Most of the participants had low risk for CVD. Four distinct PA clusters were identified that were well differentiable by the intensity and temporal patterns of activity (inactive, evening active, moderately active, very active). A significant difference in 10-year CVD risk across the clusters was found in men (p = 0.028) and women (p

Purpose Machine-learning (ML) approaches have been repeatedly coupled with raw accelerometry to classify physical activity classes, but the features required to optimize their predictive performance are still unknown. Our aim was to identify appropriate combination of feature subsets and prediction algorithms for activity class prediction from hip-based raw acceleration data. Methods The hip-based raw acceleration data collected from 27 participants was split into training (70 %) and validation (30 %) subsets. A total of 206 time- (TD) and frequencydomain (FD) features were extracted from 6-second non-overlapping windows of the signal. Feature selection was done using seven filter-based, two wrapper-based, and one embedded algorithm, and classification was performed with artificial neural network (ANN), support vector machine (SVM), and random forest (RF). For every combination between the feature selection method and the classifiers, the most appropriate feature subsets were found and used for model training within the training set. These models were then validated with the left-out validation set. Results The appropriate number of features for the ANN, SVM, and RF ranged from 20 to 45. Overall, the accuracy of all the three classifiers was higher when trained with feature subsets generated using filter-based methods compared with when they were trained with wrapper-based methods (range: 78.1 %–88 % vs. 66 %–83.5 %). TD features that reflect how signals vary around the mean, how they differ with one another, and how much and how often they change were more frequently selected via the feature selection methods. Conclusions A subset of TD features from raw accelerometry could be sufficient for ML-based activity classification if properly selected from different axes.