Validity and repeatability of the EPIC physical activity questionnaire

a validation study using accelerometers as an objective measure

Anne E. Cust*, Ben J. Smith, Josephine Chau, Hidde P. van der Ploeg, Christine M. Friedenreich, Bruce K. Armstrong, Adrian Bauman

*Corresponding author for this work

Research output: Contribution to journalArticle

120 Citations (Scopus)

Abstract

Background: A primary aim of the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort study is to examine the association between total physical activity levels (comprising occupational, household and recreational activity) and the incidence of cancer. We examined the validity and long-term repeatability of total physical activity measurements estimated from the past-year recall EPIC questionnaire, using accelerometers as an objective reference measure. Methods: Participants included 100 men and 82 women aged 50-65 years. Criterion validity was assessed by comparing the physical activity estimates from the EPIC questionnaire with total activity estimated from the average of three separate 7-day accelerometer periods during the same (past-year) period. Long-term repeatability of the EPIC questionnaire was assessed by comparing the responses from the baseline and 10-month administrations. Past-year EPIC estimates were also compared with the Friedenreich Lifetime Total Physical Activity Questionnaire to examine whether recent activity reflected lifetime activity. Results: Accelerometer total metabolic equivalent (MET)-hours/week were positively associated with a total physical activity index (Spearman rank correlation ρ = 0.29, 95% confidence interval (CI) 0.15, 0.42) and with non-occupational activity estimated in MET-hours/week (ρ = 0.21, 95% CI 0.07, 0.35). Stratified analyses suggested stronger correlations for non-occupational activity for participants who were male, had a lower BMI, were younger, or were not full-time workers, although the differences in correlations between groups were not statistically significant. The weighted kappa coefficient for repeatability of the total physical activity index was 0.62 (95% CI 0.53, 0.71). Spearman correlations for repeatability of components of activity were 0.65 (95% CI 0.55, 0.72) for total non-occupational, 0.58 (95% CI 0.48, 0.67) for recreational and 0.73 (95% CI 0.66, 0.79) for household activity. When past-year activity was compared to lifetime estimates of activity, there was fair agreement for non-occupational (ρ = 0.26) activity, which was greater for household activity (ρ = 0.46) than for recreational activity (ρ = 0.21). Conclusion: Our findings suggest that the EPIC questionnaire has acceptable measurement characteristics for ranking participants according to their level of total physical activity. The questionnaire should be able to identify the presence or absence of reasonably strong aetiological associations when either recent or long-term activity is the responsible factor.

Original languageEnglish
Article number33
Pages (from-to)1-13
Number of pages13
JournalInternational Journal of Behavioral Nutrition and Physical Activity
Volume5
DOIs
Publication statusPublished - 2 Jun 2008
Externally publishedYes

Bibliographical note

Copyright the Author(s) 2008. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Fingerprint Dive into the research topics of 'Validity and repeatability of the EPIC physical activity questionnaire: a validation study using accelerometers as an objective measure'. Together they form a unique fingerprint.

  • Cite this