Reproducibility, power and validity of visual analogue scares in assessment of appetite sensations in single test meal studies

OBJECTIVE: To examine reproducibility and validity of visual analogue scale s (VAS) for measurement of appetite sensations, with and without a diet sta ndardization prior to the test days. DESIGN: On two different Test days the subjects recorded their appetite sen sations before breakfast and every 30 min during the 4.5 h postprandial per iod under exactly the same conditions. SUBJECTS: 55 healthy men (age 25.6+/-0.6 y, BMI 22.6+/-0.3 kg/m(2)). MEASUREMENTS: VAS were used to record hunger, satiety, fullness, prospectiv e food consumption, desire to eat something fatty, salty, sweet or savoury, and palatability of the meals. Subsequently an ad libitum lunch was served and energy intake was recorded. Reproducibility was assessed by the coeffi cient of repeatability (CR) of fasting, mean 4.5 h and peak/nadir values. RESULTS: CRs (range 20-61 mm) were larger for fasting and peak/nadir values compared with mean 4.5 h values. No parameter seemed tea be improved by di et standardization. Using a paired design and a study power of 0.8, a diffe rence of 10 mm on fasting and 5 mm on mean 4.5 h ratings can be detected wi th 18 subjects. When using desires to eat specific types of food or an unpa ired design, more subjects ape needed due to considerable variation. The be st correlations of validity were found between 4.5 h mean VAS of the appeti te parameters and subsequent energy intake (r=+/-0.50 - 0.53, P<0.001). CONCLUSION: VAS scores are reliable for appetite research and do not: seem to be influenced by prior diet standardization. However, consideration shou ld be given to the specific parameters being measured, their sensitivity an d study power.