Search

PurposeTo test the validity of the Ecological Video Identification of Physical Activity (EVIP) computer vision algorithms for automated video-based ecological assessment of physical activity in settings such as parks and schoolyards.MethodsTwenty-seven hours of video were collected from stationary overhead video cameras across 22 visits in nine sites capturing organized activities. Each person in the setting wore an accelerometer, and each second was classified as moderate-to-vigorous physical activity or sedentary/light activity. Data with 57,987 s were used to train and test computer vision algorithms for estimating the total number of people in the video and number of people active (in moderate-to-vigorous physical activity) each second. In the testing data set (38,658 s), video-based System for Observing Play and Recreation in Communities (SOPARC) observations were conducted every 5 min (130 observations). Concordance correlation coefficients (CCC) and mean absolute errors (MAE) assessed agreement between (1) EVIP and ground truth (people counts+accelerometry) and (2) SOPARC observation and ground truth. Site and scene-level correlates of error were investigated.ResultsAgreement between EVIP and ground truth was high for number of people in the scene (CCC = 0.88; MAE = 2.70) and moderate for number of people active (CCC = 0.55; MAE = 2.57). The EVIP error was uncorrelated with camera placement, presence of obstructions or shadows, and setting type. For both number in scene and number active, EVIP outperformed SOPARC observations in estimating ground truth values (CCC were larger by 0.11-0.12 and MAE smaller by 41%-48%).ConclusionsComputer vision algorithms are promising for automated assessment of setting-based physical activity. Such tools would require less manpower than human observation, produce more and potentially more accurate data, and allow for ongoing monitoring and feedback to inform interventions.

Few studies assess built environment correlates of active commuting in low-and-middle-income countries (LMICs), but the different context could yield distinct findings. Policies and investments to promote active commuting remain under-developed in LMICs like India, which grapples with traffic congestion, lack of activity-supportive infrastructure, poor enforcement of traffic rules and regulations, air pollution, and overcrowding. This cross-sectional study investigated associations between home neighborhood environment characteristics and active commuting in Chennai, India. Adults (N = 370, 47.2% female, mean age =37.9 years) were recruited from 155 wards in the metropolitan area of Chennai in southern India between January and June 2015. Participants self-reported their usual mode of commute to work, with responses recoded into three categories: (1) multi-modal or active commuting (walking and bicycling; n = 56); (2) public transit (n = 52); and (3) private transport (n = 111). Environmental attributes around participants' homes were assessed using the Neighborhood Environment Walkability Scale for India (NEWS-India). Associations between environmental characteristics and likelihood of active commuting and public transit use were modeled using logistic regression with private transport (driving alone or carpool) as the reference category, adjusting for age, gender, and household car ownership. Consistent with other international studies, participants living in neighborhoods with a mix of land uses and a transit stop within a 10-minute walk from home were more likely to use active commuting (both p < 0.01). Land-use mix was significantly associated with the use of public transit compared to private transport (adjusted odds ratio (aOR) =5.2, p = 0.002). Contrary to findings in high-income countries, the odds of active commuting were reduced with improved safety from crime (aOR =0.2, p = 0.003), aesthetics (aOR =0.2, p = 0.05), and street connectivity (aOR =0.2, p = 0.003). Different environmental attributes were associated with active commuting, suggesting that these relationships are complex and may distinctly differ from those in high-income countries. Unexpected inverse associations of perceived safety from crime and aesthetics with active commuting emphasize the need for high-quality epidemiologic studies with greater context specificity in the study of physical activity in LMICs. Findings have public health implications for India and suggest that caution should be taken when translating evidence across countries.

Technological advances provide opportunities for automating direct observations of physical activity, which allow for continuous monitoring and feedback. This pilot study evaluated the initial validity of computer vision algorithms for ecological assessment of physical activity. The sample comprised 6630 seconds per camera (three cameras in total) of video capturing up to nine participants engaged in sitting, standing, walking, and jogging in an open outdoor space while wearing accelerometers. Computer vision algorithms were developed to assess the number and proportion of people in sedentary, light, moderate, and vigorous activity, and group-based metabolic equivalents of tasks (MET)-minutes. Means and standard deviations (SD) of bias/difference values, and intraclass correlation coefficients (ICC) assessed the criterion validity compared to accelerometry separately for each camera. The number and proportion of participants sedentary and in moderate-to-vigorous physical activity (MVPA) had small biases (within 20% of the criterion mean) and the ICCs were excellent (0.82-0.98). Total MET-minutes were slightly underestimated by 9.3-17.1% and the ICCs were good (0.68-0.79). The standard deviations of the bias estimates were moderate-to-large relative to the means. The computer vision algorithms appeared to have acceptable sample-level validity (i.e., across a sample of time intervals) and are promising for automated ecological assessment of activity in open outdoor settings, but further development and testing is needed before such tools can be used in a diverse range of settings.

Physical activity and time spent outdoors may be important non-pharmacological approaches to improve sleep quality and duration (or sleep patterns) but there is little empirical research evaluating the two simultaneously. The current study assesses the role of physical activity and time outdoors in predicting sleep health by using objective measurement of the three variables. A convenience sample of 360 adult women (mean age = 55.38 ±9.89 years; mean body mass index = 27.74 ±6.12) was recruited from different regions of the U.S. Participants wore a Global Positioning System device and ActiGraph GT3X+ accelerometers on the hip for 7 days and on the wrist for 7 days and 7 nights to assess total time and time of day spent outdoors, total minutes in moderate-to-vigorous physical activity per day, and 4 measures of sleep health, respectively. A generalized mixed-effects model was used to assess temporal associations between moderate-to-vigorous physical activity, outdoor time, and sleep at the daily level (days = 1931) within individuals. There was a significant interaction (p = 0.04) between moderate-to-vigorous physical activity and time spent outdoors in predicting total sleep time but not for predicting sleep efficiency. Increasing time outdoors in the afternoon (versus morning) predicted lower sleep efficiency, but had no effect on total sleep time. Time spent outdoors and the time of day spent outdoors may be important moderators in assessing the relation between physical activity and sleep. More research is needed in larger populations using experimental designs.

Recent technological, geophysical, and societal forces have fundamentally altered the structure and functioning of human environments. Prominent among these forces are the rise of the Internet; rapid rates of global environmental change; and widening rifts among different socioeconomic, racial, religious, and ethnic groups. The present article traces the influence of these conditions on individuals' cognition, behavior, and well-being. New theoretical questions are raised and conceptual frameworks proposed to understand how global conditions are restructuring people's relationships with their everyday environments. New directions for psychological research and practice aimed at reducing global threats to personal and societal well-being are discussed.

Background: Neighborhood parks are recognized as important spaces for facilitating physical activity (PA); however, it remains unclear how the frequency of park use is associated with PA. The purpose of this study was to examine associations between minutes of moderate to vigorous PA and multiple park use indicators: (1) use of a neighborhood park, (2) unique number of neighborhood parks used, and (3) frequency of neighborhood park use. Methods: Adults were surveyed from 4 US cities (Brooklyn, NY; Greenville County, SC; Raleigh, NC; and Seattle, WA). Using a map-based survey platform, participants indicated all neighborhood parks they used and the frequency of use in the past 30 days. Participants self-reported their weekly moderate to vigorous PA. Quantile regression was used to examine associations between PA and park use indicators. Results: Of all respondents (N = 360), 60% indicated visiting a neighborhood park in the past 30 days, with an average of about 13 total neighborhood park visits (SD = 17.5). Significant, positive associations were found between moderate to vigorous PA and both unique neighborhood park visits and total number of neighborhood parks visits. Conclusions: Frequency of park visitation is associated with PA among US adults. Ensuring equitable and safe access to neighborhood parks has the potential for population-level PA health benefits. [ABSTRACT FROM AUTHOR]

Much of the current neighborhood-based research uses variables aggregated on administrative boundaries such as zip codes, census tracts, and block groups. However, other methods using current technological advances in geographic sciences may broaden our ability to explore the spatial concentration of neighborhood factors affecting individuals and groups. This article builds on extant neighborhood-based research methods and proposes using a geographic information system (GIS) to explore a conceptually new method of defining neighborhood boundaries for social measurement. Geospatial factors are important in the formation of neighborhoods, and a GIS can be used to define neighborhood boundaries that account for the spatial factors influencing the natural flow and pattern of neighbor interactions. Georeferenced data from the 2000 U.S. census for a midwestern city are presented to elucidate the potential aggregation problems and theoretical arguments. Moran's I statistics show that administrative boundaries may not be valid aggregate measures of neighborhoods or their effects.