A guide to analysing Universal Eating Monitor data: Assessing the impact of different analysis techniques

Cumulative intake curves and within-meal changes in subjective feelings of fullness can provide researchers
with detailed data on the effects of psychological, nutritional or pharmacological manipulations on the
expression of human appetite. However, a number of different approaches to the collection and analysis of
within-meal data exist resulting in potential to produce contrasting findings. The current study measured
cumulative intake and change in appetite using a Universal Eating Monitor (UEM). Three different techniques
(area under the curve, visual ascription, and a coefficient approach) were used to analyse the same cumulative
intake curves produced in a study of stress on food intake. Twenty-three adult participants (mean age 21 years)
consumed ameal comprised of pasta and marinara sauce and, with the aid of the Sussex Meal Pattern Monitor
(SMPM), were periodically interrupted to measure subjective feelings of fullness. As hypothesised, analysing
cumulative intake curves with differing techniques affected the overall study findings. No significant betweencondition
differences in the cumulative intake or fullness curves were found using either the visual ascription
or the area under the curve approaches. In contrast, the coefficient approach found a significant difference in
the fullness curves between relaxation and cold pressor conditions (p=0.012). This discrepancy in findingswas
due to the presence of a quadratic component in the cumulative intake curve in the stress condition whichwas
not present in control (p=0.017). Whilst the relative merits of various approaches to microstructural analysis of
eating behaviour remain to be fully evaluated, the case for some form of standardised analytic approach may
need to be addressed.