Association between self-reported and objective activity levels by demographic factors: Ecological momentary assessment study in children

Jennifer Zink, Britni R. Belcher, Eldin Dzubur, Wangjing Ke, Sydney O'connor, Jimi Huh, Nanette Lopez, Jaclyn P. Maher, Genevieve F. Dunton

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

Background: To address the limitations of the retrospective self-reports of activity, such as its susceptibility to recall bias, researchers have shifted toward collecting real-time activity data on mobile devices via ecological momentary assessment (EMA). Although EMA is becoming increasingly common, it is not known how EMA self-reports of physical activity and sedentary behaviors relate to the objective measures of activity or whether there are factors that may influence the strength of association between these two measures. Understanding the relationship between EMA and accelerometry can optimize future instrument selection in studies assessing activity and health outcomes. Objective: The aim of this study was to examine the associations between EMA-reported sports or exercise using the accelerometer-measured moderate-to-vigorous physical activity (MVPA) and EMA-reported TV, videos, or video games with the accelerometer-measured sedentary time (ST) in children during matched 2-h windows and test potential moderators. Methods: Children (N=192; mean age 9.6 years; 94/192, 49.0% male; 104/192, 54.2% Hispanic; and 73/192, 38.0% overweight or obese) wore an accelerometer and completed up to 7 EMA prompts per day for 8 days during nonschool time, reporting on past 2-h sports or exercise and TV, videos, or video games. Multilevel models were used to assess the relationship between the accelerometer-measured ST and EMA-reported TV, videos, or video games. Given the zero-inflated distribution of MVPA, 2-part models were used assess the relationship between the accelerometer-measured MVPA and EMA-reported sports or exercise. Results: EMA-reported TV, videos, or video games were associated with a greater accelerometer-measured ST (beta=7.3, 95% CI 5.5 to 9.0, P<.001). This relationship was stronger in boys (beta=9.9, 95% CI 7.2 to 12.6, P<.001) than that in girls (beta=4.9, 95% CI 2.6 to 7.2, P≤.001). EMA-reported sports or exercise was associated with a greater accelerometer-measured MVPA (zero portion P<.001; positive portion P<.001). This relationship was stronger on weekends, in older children, and in non-Hispanic children (zero portion all P values<.001; positive portion all P values<.001). Conclusions: EMA reports highly relate to accelerometer measures. However, the differences in the strength of association depending on various demographic characteristics suggest that future research should use both EMA and accelerometers to measure activity to collect complementary activity data.

Original languageEnglish (US)
Article numbere150
JournalJMIR mHealth and uHealth
Volume6
Issue number6
DOIs
StatePublished - Jun 2018
Externally publishedYes

Keywords

  • Children
  • Measurement
  • Mobile devices
  • Physical activity
  • Sedentary behavior

ASJC Scopus subject areas

  • Health Informatics

Fingerprint

Dive into the research topics of 'Association between self-reported and objective activity levels by demographic factors: Ecological momentary assessment study in children'. Together they form a unique fingerprint.

Cite this