Computer-based assessments of expected satiety predict behavioural measures of portion-size selection and food intake

Abstract

Previously, expected satiety (ES) has been measured using software and two-dimensional pictures presented on a computer screen. In this context, ES is an excellent predictor of self-selected portions, when quantified using similar images and similar software. In the present study we sought to establish the veracity of ES as a predictor of behaviours associated with real foods. Participants (N=30) used computer software to assess their ES and ideal portion of three familiar foods. A real bowl of one food (pasta and sauce) was then presented and participants self-selected an ideal portion size. They then consumed the portion ad libitum. Additional measures of appetite, expected and actual liking, novelty, and reward, were also taken. Importantly, our screen-based measures of expected satiety and ideal portion size were both significantly related to intake (p<.05). By contrast, measures of liking were relatively poor predictors (p>.05). In addition, consistent with previous studies, the majority (90%) of participants engaged in plate cleaning. Of these, 29.6% consumed more when prompted by the experimenter. Together, these findings further validate the use of screen-based measures to explore determinants of portion-size selection and energy intake in humans.

Preventing the pack size effect: Exploring the effectiveness of pictorial and non-pictorial serving size recommendations

“…We replicated the portion size effect that typically occurs in actual eating situations in an online paradigm with food pictures. Our findings correspond with previous research showing that the portion size preferences that people provide while using food pictures (Wilkinson et al, 2012) or food replicas (Bucher, van der Horst, & Siegrist, 2012) align well with actual consumption amounts. Furthermore, our results showed that the pack size effect is also visible when measuring expected consumption.…”

“…As in Experiment 1 and 2, we again found that the evaluation of the size of the recommended serving had a significant influence on the amount of M&M's taken and consumed. The amount of snack food that people take thus clearly depends on people's general notions of what a small, reasonable or large portion size is (Brunstrom & Shakeshaft, 2009;Wilkinson et al, 2012) and is not only influenced by environmental factors.…”

“…Earlier research has suggested that people typically represent the portions they eat visually (Wilkinson et al, 2012) or in easily countable units (Geier, Rozin, & Doros, 2006;Marchiori, Waroquier, & Klein, 2011). Similarly, people often have difficulty understanding serving size recommendations in grams (Faulkner et al, 2012).…”

Socioeconomic status and the susceptibility to overeat from excessive food portions

“…Importantly, intended consumption in this study positively correlated with factors known to predict actual eating, such as hunger and liking (e.g., Robinson et al, 2017). In line with previous work (e.g., Wilkinson et al, 2012), it is argued that the measure of intended consumption in the present study is likely to map well onto actual consumption. In addition, a recent meta-analysis demonstrated a larger effect of portion size on actual consumption (d = 0.45) than on intended consumption (d = 0.18; Zlatevska, Dubelaar, & Holden, 2014).…”

Expected satiation is correlated with amplitude of intake but does not fit well with actual consumption of desserts

“…Our 387 findings were partly in contradiction with previous results which had shown that expected 388 satiation was the only significant predictor of intake (Wilkinson et al, 2012). Such a 389 difference in the results might have been caused by our choice to work with desserts, products 390 that are more likely to be consumed for their hedonic value than the pasta with sauce used in 391 Wilkinson's study.…”