Effect of training on the reliability of satiety evaluation and use of trained panellists to determine the satiety effect of dietary fibre: a randomised controlled trial

Vicky A Solah; Xingqiong Meng; Simon Wood; Roland J Gahler; Deborah A Kerr; Anthony P James; Sebely Pal; Haelee K Fenton; Stuart K Johnson

doi:10.1371/journal.pone.0126202

Effect of training on the reliability of satiety evaluation and use of trained panellists to determine the satiety effect of dietary fibre: a randomised controlled trial

PLoS One. 2015 May 15;10(5):e0126202. doi: 10.1371/journal.pone.0126202. eCollection 2015.

Authors

Vicky A Solah¹, Xingqiong Meng², Simon Wood³, Roland J Gahler⁴, Deborah A Kerr¹, Anthony P James¹, Sebely Pal¹, Haelee K Fenton¹, Stuart K Johnson¹

Affiliations

¹ Nutrition, Dietetics and Food Science, School of Public Health, Faculty of Health Sciences, Curtin University, Perth, WA, Australia.
² Nutrition, Dietetics and Food Science, School of Public Health, Faculty of Health Sciences, Curtin University, Perth, WA, Australia; Flinders Centre for Innovation in Cancer, Flinders University, Adelaide, Australia.
³ Nutrition, Dietetics and Food Science, School of Public Health, Faculty of Health Sciences, Curtin University, Perth, WA, Australia; Food, Nutrition and Health Program, University of British Columbia, Vancouver, BC, Canada; InovoBiologic Inc., Calgary, AB, Canada.
⁴ Factors Group R & D, Burnaby, BC, Canada.

Abstract

Background: The assessment of satiety effects on foods is commonly performed by untrained volunteers marking their perceived hunger or fullness on line scales, marked with pre-set descriptors. The lack of reproducibility of satiety measurement using this approach however results in the tool being unable to distinguish between foods that have small, but possibly important, differences in their satiety effects. An alternate approach is used in sensory evaluation; panellists can be trained in the correct use of the assessment line-scale and brought to consensus on the meanings of descriptors used for food quality attributes to improve the panel reliability. The effect of training on the reliability of a satiety panel has not previously been reported.

Method: In a randomised controlled parallel intervention, the effect of training in the correct use of a satiety labelled magnitude scale (LMS) was assessed versus no-training. The test-retest precision and reliability of two hour postprandial satiety evaluation after consumption of a standard breakfast was compared. The trained panel then compared the satiety effect of two breakfast meals containing either a viscous or a non-viscous dietary fibre in a crossover trial.

Results: A subgroup of the 23 panellists (n = 5) improved their test re-test precision after training. Panel satiety area under the curve, "after the training" intervention was significantly different to "before training" (p < 0.001). Reliability of the panel determined by intraclass correlation (ICC) of test and retest showed improved strength of the correlation from 0.70 pre-intervention to 0.95 post intervention. The trained "satiety expert panel" determined that a standard breakfast with 5g of viscous fibre gave significantly higher satiety than with 5g non-viscous fibre (area under curve (AUC) of 478.2, 334.4 respectively) (p ≤ 0.002).

Conclusion: Training reduced between panellist variability. The improved strength of test-retest ICC as a result of the training intervention suggests that training satiety panellists can improve the discriminating power of satiety evaluation.

Publication types

Randomized Controlled Trial
Research Support, Non-U.S. Gov't

MeSH terms

Adult
Cross-Over Studies
Dietary Fiber / pharmacology*
Female
Humans
Male
Middle Aged
Reproducibility of Results
Research Subjects / education*
Satiety Response / drug effects*
Young Adult

Grants and funding

Financial support for the study and supply of PGX(r) were provided by InovoBiologic, Inc. SW receives consulting fees from InovoBiologic, Inc. VAS received the funding. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.