Portion size selection in children: Effect of sensory imagery for snacks varying in energy density

Abstract

Food sensory imagery – creating a vivid mental image of the sensory experience of eating – can lead to the selection of smaller portions because it serves as a reminder that eating enjoyment does not necessarily increase with portion size. The evidence is mostly limited to adults and to energy-dense foods for which it is particularly difficult to predict the satiating effects of consumption quantity. The objective was to study how food sensory imagery influences portion size selection of foods varying in energy density (brownie and applesauce) by 7- to 11-year-old children. During after-school time, 171 children were randomized into two conditions. Children in the food sensory imagery condition were asked to imagine the taste, smell, and texture of eating palatable foods, i.e., chocolate cereal, chocolate waffle, and chocolate candies. Children in the control condition performed a similar sensory imagery task for non-food-related activities. Children were then asked to choose between the recommended serving size, a 50% larger portion, and a 125% larger portion of either brownie or applesauce. One week later, they were placed in the same condition for the other food. Compared to the control condition, food sensory imagery led children to choose 7.1% less brownie but had no effect on applesauce portion selection. Exploratory findings showed that the food sensory imagery intervention was especially effective at reducing brownie portion size selection among children who were moderately hungry, who usually eat fast, and whose parents pressure them to eat. In conclusion, food sensory imagery may be a useful intervention to nudge children towards healthier portion size choices because it reduces the selected portion size of an energy-dense snack without reducing the selected portion size of a healthier snack.

Introduction

The size of food portions has increased over the past forty years, so has childhood obesity. When they are served larger portions, children (like adults) eat more, a phenomenon known as the Portion Size Effect (PSE) (Kral & Hetherington, 2015; Reale et al., 2019). A recent meta-analysis found that PSE significantly increased food intake among children aged 2 to 12 (Standardized Mean Differences = 0.47), for both unit and amorphous foods (Reale et al., 2019) and also underscored the need for more research to understand the PSE mechanism in children. Reducing portion sizes of nutrition-poor foods consumed by children is a key strategy to fight obesity (Hetherington et al., 2018; Hollands et al., 2015). Among promising solutions, research on mindful eating has shown that training people to pay attention to their sensations while eating can reduce food cravings, as well as ad libitum food intake (Chang, Mulders, Cserjesi, Cleeremans, & Klein, 2018; Kristeller & Bolinskey, 2010; Marchiori & Papies, 2014; Muñoz-Vilches, van Trijp, & Piqueras-Fiszman, 2019; Petit, Spence, Velasco, Woods, & Cheok, 2017; Tiggemann & Kemps, 2005). Part of the explanation for this effect is that a heightened focus on eating sensations makes people better aware of the decline in food enjoyment as intake increases, a phenomenon called sensory-specific satiation (Rolls, Rolls, Rowe, & Sweeney, 1981; Tapper, 2017). Recent research has also investigated how this heightened focus can influence portion size selection. Cornil & Chandon developed an intervention called “food sensory imagery”, which involves creating a vivid mental image of the multi-sensory experience of eating, prior to choosing a portion size of food (Cornil & Chandon, 2016a). They found that compared to a control condition, sensory imagery led participants to choose smaller portions of brownies or chocolate cakes. This could occur because sensory imagery increased the relative importance of anticipated sensory pleasure in portion size selection and helped people better anticipate that sensory pleasure tends to decline with increasing portion size. This effect was observed for the portion size choice of palatable, energy-dense foods, and it was stronger among hungry rather than among sated participants, who tend to choose small portions regardless of the sensory condition. Another recent study examined the effect of sensory imagery on the consumption of unsalted almonds and M&Ms (Chang et al., 2018). Compared to a control condition, sensory imagery led to a reduction in the consumption of both types of food among hungry adults. Importantly, even though unsalted almonds are healthier than M&Ms, both foods were perceived as equally liked by the study participants.

The current research aimed to investigate further the beneficial effect of food sensory imagery in children. With the exception of one study conducted with a small sample of 4- to 5-year-old children (N = 42) and which did not examine the moderating role of initial hunger (Cornil & Chandon, 2016a), all studies on sensory imagery were conducted among adults. Testing the effect of a sensory imagery intervention among slightly older children is important because it is an age at which children are forming their own perceptions of food, with potentially long-lasting impacts (Schultz & Danford, 2016). Hence, this study aimed to test the effectiveness of sensory imagery among 7- to 11-year-old children. In addition, this research aimed to extend past research by testing the effects of sensory imagery across types of foods that vary in energy density (chocolate brownie vs. applesauce). If the effects of sensory imagery are due to its ability to improve people's expectations of the effects of portion size on satiation and eating enjoyment (Cornil & Chandon, 2016a), it should only influence portion size choices when people overestimate the true effects of quantity on satiation and eating enjoyment. Recent research has shown that adults accurately predict the satiating effects of increasing quantities only for low-energy dense foods, such as fruits or vegetables. For energy-dense foods (above 1.75 kcal/g), however, expected satiation barely increases with energy density, showing a failure to recognize their high energy content (Brunstrom, Drake, Forde, & Rogers, 2018). We therefore hypothesize that sensory imagery would decrease chosen portion sizes for high calorie-dense food like brownies (calorie density = 4.6 kcal/g), consistent with existing research, but not for less calorie-dense food like applesauce (calorie density = 0.7 kcal/g).

In an exploratory way, this research also investigated the moderating role of individual characteristics related to children's eating traits (slowness in eating, restrained eating, food responsiveness, and liking for sweetness and fattiness sensations) and to parental feeding practices (parental food restriction and parental pressure to eat). These factors were chosen because of their likely interaction with sensory imagery (Birch et al., 2001; Fogel et al., 2018; Lange et al., 2018; Monnery-Patris et al., 2011; van Jaarsveld, Llewellyn, Johnson, & Wardle, 2011; Wardle, Guthrie, Sanderson, & Rapoport, 2001). For example, a recent review paper has argued that faster eating, by delaying sensory-specific satiation, may be one of the mechanisms of the portion size effect (Zuraikat, Smethers, & Rolls, 2019). This suggests that sensory imagery may be particularly effective for children who are fast eaters.