Level of agreement between physical activity levels determined by accelerometry (ActiGraph) and PAQ-C in 8-year-old children from a black South African population: BC-IT study

Oral Presentation B13.2

Authors

  • Caroline M. Sedumedi North-West University
  • Makama A. Monyeki North-West University
  • Xanne Janssen University of Strathclyde
  • John J. Reilly University of Strathclyde

DOI:

https://doi.org/10.14288/hfjc.v14i3.577

Keywords:

Physical Activity, Objective, Subjective, Deuterium Dilution Method, South African Children

Abstract

Background: Misclassification of physical activity (PA) levels in children due the use of invalid methods may heighten related adverse health risks. Objective methods for assessing PA are preferable but are not always feasible due to costs and related logistics, and questionnaires may be more suitable for national surveillance purposes. However, their validity and reliability vary in different populations. Purpose: To determine the level of agreement between PA levels measured by accelerometry (ActiGraph Model GT3X-BT) and the Physical Activity Questionnaire for Older Children (PAQ-C) in 8-year-old children from a black South African population. Methods: Eight-year-old children (8.37±0.73 years) from a larger study were participants. PA was determined using the objective and subjective methods of accelerometer (ActiGraph) and PAQ-C, respectively. Bland–Altman plots were used to assess the level of agreement between ActiGraph and PAQ-C. Results: No significant gender difference in PA score determined by PAQ-C was found. When ActiGraph was used, boys reported significantly (p<0.05) higher moderate PA, vigorous PA and moderate-to-vigorous PA than girls. The results show a significant (p<0.001) mean bias of -0.50, and equivalent (r2) of -0.01%, with limits of agreement ranges from -0.23 to -0.68, SEE of 0.14, equivalent (r2) to -0.1% for total sample. Furthermore Bland–Altman plots showed poor level of agreement between PA levels by ActiGraph and PAQ-C (p>0.05). The results further reveal that PAQ-C underestimated the PA of the sample population. Conclusion: Poor level of agreement existed between the two methods used to assess PA. PAQ-C underestimated PA levels in a sample of South Africa black children. PAQ-C, when used as a lone method for assessment of PA levels, should be approached with caution and where possible, the use of both objective measures and self-report is recommended. Funding: Research supported by the South African Medical Research Council (MRC) under the Self-Initiated Research Grants Programme. Additionally, funds received from the NRF for rated researchers are acknowledged. Financial support received from North-West University in the establishment of the stable isotope laboratory for the determination of body composition within the PhASRec research entity is greatly appreciated. Additionally, we would like to thank the IAEA for financial support (TC project SAF6020).

Published

2021-09-30

How to Cite

Sedumedi, C., Monyeki, M., Janssen, X., & Reilly, J. (2021). Level of agreement between physical activity levels determined by accelerometry (ActiGraph) and PAQ-C in 8-year-old children from a black South African population: BC-IT study: Oral Presentation B13.2. The Health & Fitness Journal of Canada, 14(3). https://doi.org/10.14288/hfjc.v14i3.577