Original Article
European Journal of Clinical Nutrition advance online publication 16 September 2009; doi: 10.1038/ejcn.2009.107
Accuracy of children's school-breakfast reports and school-lunch reports (in 24-h dietary recalls) differs by retention interval
S D Baxter1, C H Guinn1, J A Royer1, J W Hardin2, A J Mackelprang1 and A F Smith3
- 1Institute for Families in Society, University of South Carolina, Columbia, SC, USA
- 2Department of Epidemiology and Biostatistics and Center for Health Services and Policy Research, University of South Carolina, Columbia, SC, USA
- 3Department of Psychology, Cleveland State University, Cleveland, OH, USA
Correspondence: Dr SD Baxter, Institute for Families in Society, University of South Carolina, 1600 Hampton Street, Suite 507, Columbia, SC 29208, USA. E-mail: sbaxter@mailbox.sc.edu
Received 3 March 2009; Revised 15 July 2009; Accepted 20 July 2009; Published online 16 September 2009.
Abstract
Background/Objectives:
Validation-study data were analyzed to investigate the effect of retention interval (time between the to-be-reported meal and interview) on accuracy of children's school-breakfast reports and school-lunch reports in 24-h recalls, and to compare accuracy of children's school-breakfast reports for two breakfast locations (classroom; cafeteria).
Subjects/Methods:
Each of 374 fourth-grade children was interviewed to obtain a 24-h recall using one of six conditions from crossing two target periods (prior 24 h; previous day) with three interview times (morning; afternoon; evening). Each condition had 62 or 64 children (half boys). A recall's target period included one school breakfast and one school lunch, for which the child had been observed. Food-item variables (observed number; reported number; omission rate; intrusion rate) and energy variables (observed; reported; report rate; correspondence rate; inflation ratio) were calculated for each child for school breakfast and school lunch separately.
Results:
Accuracy for school-breakfast reports and school-lunch reports was inversely related to retention interval. Specifically, as indicated by smaller omission rates, smaller intrusion rates, larger correspondence rates and smaller inflation ratios, accuracy for school-breakfast reports was best for prior-24-h recalls in the morning, and accuracy for school-lunch reports was best for prior-24-h recalls in the afternoon. For neither school meal was a significant sex effect found for any variable. For school-breakfast reports, there was no significant school-breakfast location effect for any variable.
Conclusions:
By shortening the retention interval, accuracy can be improved for school-breakfast reports and school-lunch reports in children's 24-h recalls.
Keywords:
validation study, school meals, retention interval
