Correlates of physical activity behavior in adults: a data mining approach

Abstract

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:

Joint Profiles of Sedentary Time and Physical Activity in Adults and Their Associations with Cardiometabolic Health

“…Although no time-based limit exists for sedentary time, efforts have been made to provide more detailed recommendations for sedentary time ( 8 , 14 , 40 ), including encouraging adults to limit the daily time spent sedentary to 8 h·d −1 while accumulating at least 150–300 min of MVPA each week ( 40 ). However, sedentary time and physical activity intensities could be constrained by nonnegotiable factors such as occupation and environment ( 10 , 12 , 41 ). A few studies have recently argued that one-size-fits-all thresholds for sedentary time and physical activities may not always be adequate for promoting healthier daily activity behaviors ( 41 , 42 ).…”

“…It followed the legislation, decrees, and ethical principles concerning medical research on humans in Finland. Further information about the NFBC1966 study, recruitment, and follow-ups is available elsewhere ( 10 ). This cross-sectional study included members of the NFBC1966 who participated in the latest follow-up performed at the age of 46 yr (during 2012–2014) and who agreed to wear an accelerometer for measuring daily activity.…”

“…Sedentary time and physical activity intensities are influenced by occupation, environment, and physical capacity (e.g., fitness level), over which individuals have minimal control ( 9 , 10 ). Substantial variations may occur in diurnal levels of sedentary time and physical activity intensities across weekdays and weekend days ( 11 ), which are shown to be related to cardiometabolic health ( 12 , 13 ) in adults.…”

Device-measured physical activity and sedentary time in a national sample of Luxembourg residents: the ORISCAV-LUX 2 study

“…This is the first device-based study to show that the Luxembourg adult population spend on average more than 12 h/day sedentary, about 160 min/d in light PA, and 84 min/d in MVPA. Other device-based estimates captured across Europe typically show more daily time in light PA (ranging from 200 to 416 min/d), less sedentary time (7.3 to 10.2 h/day) and MVPA (26 to 69 min/d) [ 36 – 45 ]. Few investigations have reported equivalent or more daily MVPA [ 7 , 46 , 47 ].…”

Conceptualizing and Measuring Appetite Self-Regulation Phenotypes and Trajectories in Childhood: A Review of Person-Centered Strategies

“…However, the inclusion of many behaviors, characteristics, and traits to determine ASR phenotypes may be too computationally burdensome for the model-based person-centered approaches discussed in the present review. Future ASR research may need to employ machine learning, or data mining, methods to determine phenotypes from large and complex datasets ( 113 – 115 ).…”