Open AccessDesign and Evaluation of a Computer-Based 24-Hour Physical Activity Recall (cpar24) Instrument
Author(s) -
Simone Kohler,
Gundula Behrens,
Matthias Olden,
Sebastian E. Baumeister,
Alexander Horsch,
Beate Fischer,
Michael F. Leitzmann
Publication year - 2017
Publication title -
jmir. journal of medical internet research/journal of medical internet research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.446
H-Index - 142
eISSN - 1439-4456
pISSN - 1438-8871
DOI - 10.2196/jmir.7620
Subject(s) - the internet , physical activity , recall , world wide web , computer science , internet privacy , multimedia , psychology , medicine , physical therapy , cognitive psychology
Background Widespread access to the Internet and an increasing number of Internet users offers the opportunity of using Web-based recalls to collect detailed physical activity data in epidemiologic studies. Objective The aim of this investigation was to evaluate the validity and reliability of a computer-based 24-hour physical activity recall (cpar24) instrument with respect to the recalled 24-h period. Methods A random sample of 67 German residents aged 22 to 70 years was instructed to wear an ActiGraph GT3X+ accelerometer for 3 days. Accelerometer counts per min were used to classify activities as sedentary (<100 counts per min), light (100-1951 counts per min), and moderate to vigorous (≥1952 counts per min). On day 3, participants were also requested to specify the type, intensity, timing, and context of all activities performed during day 2 using the cpar24. Using metabolic equivalent of task (MET), the cpar24 activities were classified as sedentary (<1.5 MET), light (1.5-2.9 MET), and moderate to vigorous (≥3.0 MET). The cpar24 was administered twice at a 3-h interval. The Spearman correlation coefficient (r) was used as primary measure of concurrent validity and test-retest reliability. Results As compared with accelerometry, the cpar24 underestimated light activity by −123 min (median difference, P difference <.001) and overestimated moderate to vigorous activity by 89 min ( P difference <.001). By comparison, time spent sedentary assessed by the 2 methods was similar (median difference=+7 min, P difference=.39). There was modest agreement between the cpar24 and accelerometry regarding sedentary ( r =.54), light ( r =.46), and moderate to vigorous ( r =.50) activities. Reliability analyses revealed modest to high intraclass correlation coefficients for sedentary ( r =.75), light ( r =.65), and moderate to vigorous ( r =.92) activities and no statistically significant differences between replicate cpar24 measurements (median difference for sedentary activities=+10 min, for light activities=−5 min, for moderate to vigorous activities=0 min, all P difference ≥.60). Conclusion These data show that the cpar24 is a valid and reproducible Web-based measure of physical activity in adults.