Differences in Adults’ Sedentary and Physical Activity Levels Across Social and Physical Contexts: An Ecological Momentary Assessment and Accelerometer Study

Background and Purpose Research has shown that features of the home, work, and neighborhood environments are associated with sedentary and physical activity in adults. However, little is known about the extent to which objectively measured levels of sedentary and physical activity differ across these settings. According to behavior setting theory, characteristics of specific context can influence the nature of the behavior occurring in that setting. Preliminary studies using Global Positioning System (GPS) methodologies have been able to identify some contexts where adults‘ physical activity occurs (e.g., home, work, school). However, these strategies are weak at differentiating between indoor and outdoor settings at the same location, micro settings within the same location, and types of social contexts. These limitations may be partially addressed using real-time self-report Ecological Momentary Assessment (EMA), which can simultaneously measure where and with whom behaviors occur.

Objectives This study used EMA paired with accelerometers in low-to-middle income, ethnically-diverse adults to compare the levels of sedentary and physical activity occurring across specific social and physical contexts. A second objective was to determine whether context-specific patterns differed for men and women.

Methods A sample of 110 adults (73% female, 30% Hispanic, 66% married, 62% overweight or obese, mean age = 40.4 years) participated in four consecutive days (Saturday – Tuesday) of EMA monitoring between 6:30am and 10:00pm each day. Eight EMA surveys were prompted per day. EMA data were collected through a mobile phone with custom software installed. Participants responded to EMA questions about their current social context (e.g., spouse, children, friends, alone), general physical context (e.g., home [indoors], home [outdoors], outdoors [not at home], work), and specific physical context (e.g., kitchen, living room, yard, driveway, park/trail, sidewalk). Adults simultaneously wore an Actigraph GT2M accelerometer. EMA responses were time-matched to the number of minutes in moderate-to-vigorous physical activity (MVPA) (> 2,020 counts/minute) and sedentary activity (SA) (< 100 counts/minute) in the ± 15 minute window surrounding that EMA prompt. Multilevel linear regressions compared minutes of MVPA and SA across social and physical contexts. Interactions terms were entered into the models to test effect modification by gender.

Results On average, participants answered 82% (range 25% – 100%) of EMA prompts. There was a statistically significant gender by general physical context interaction for MVPA (F = 3.31, p = .006). For men, MVPA in the ± 15 minute window surrounding that EMA prompt was greater when participants reporting being outdoors (not at home) (M = 1.53 min., SE = 0.26) as compared with at home (outdoors) (M = 0.24 min., SE = 0.41) (p < .05) whereas for women, MVPA was greater at home (outdoors) (M = 1.48 min., SE = 0.25) than outdoors (not at home) (M = 0.99 min., SE = 0.19) (p < .05). For men and women, SA was greater at home (indoors) (M = 22.81 min., SE = 0.22) and at work (indoors) (M = 21.40 min., SE = 0.48) as compared with at home (outdoors) (M = 19.44 min., SE = 0.75) and outdoors (not at home) (M = 19.0 min., SE = 0.54) (p‘s < .05). When at home (indoors), SA was greater in the bedroom (M = 25.72 min., SE = 0.33) and living/family room (M = 22.11 min., SE = 0.33) than in the kitchen (M = 19.60 min., SE = 0.44) or garage (M = 17.61 min., SE = 2.04) (p‘s < .05) for men and women. Neither MVPA nor SA differed across outdoor locations away from home (i.e., park trail, road, sidewalk, parking lot). SA was significantly greater when participants were with a spouse (M = 24.01 min., SE = 0.51) than with a child (M = 21.64, SE = 0.47) or alone (M = 21.62 min., SE = 0.28) (p‘s < .05) for men and women.

Conclusions The current study used a novel research methodology to compare adults‘ objectively-measured sedentary and physical activity levels across social and physical contexts. Results indicated that women and men may differ in the types of physical settings that are most conducive to physical activity. Women‘s physical activity levels were higher when they reported being in the outdoor spaces of their home (e.g., yard, patio, driveway) whereas men‘s physical activity levels were higher in outdoor settings away from home (e.g., park/trail, road, sidewalk). Further research is needed to understand whether these gender differences are due to women‘s greater engagement in home-based physical activities such as yard work, gardening, house chores, and childcare. Further, the higher level of sedentary behavior performed in the company of a spouse as compared with children or being alone indicates a need for behavioral interventions promoting joint physical activity among married couples. Results from this study support the viability of using EMA methodology in larger scale research projects that could form the basis for context-specific interventions.

Support/Funding Source American Cancer Society 118283-MRSGT-10-012-01-CPPB