How do people cognitively represent appetitive stimuli? Do interactions with appetitive stimuli shape how we think about them, and do such representations affect motivation to consume? Although much is known about how people respond to appetitive stimuli, little is known about how they are represented. We examine this in the domain of sugar-sweetened drinks, which constitute a significant self-control problem for many people. Given people’s rich and diverse learning histories of consuming them, we propose that representations of these stimuli will show high variability, and that they will reflect idiosyncratic simulations, or re-enactments, of previous consumption experiences. Representing drinks in terms of consuming and enjoying them may predict the motivation to consume. In three experiments (total N = 457), participants described non-alcoholic drinks in a “feature listing task”, a free production task to assess cognitive representations of concepts through natural language. We also measured consumption frequency, desire to drink, and intake (Exp. 3), and we measured (Exp. 1 and 2) or manipulated (Exp. 3) thirst. Illustrating the variability of participants’ representations of drinks, participants reported a large number of different features (210-331 unique features per drink). Drinks were described heavily with words related to consumption and reward experiences, especially sugary drinks, and especially when consumed frequently. Consumption and reward features predicted desire and intake, more strongly than thirst. These findings suggest that simulations of previous rewarding interactions play a key role in representations of appetitive stimuli, and that understanding these representations may be useful across domains of appetitive behaviour.What is the motivation for consuming sugary drinks? Why do some people choose Coke, and others water, to accompany their dinner or to quench their thirst? We know very little about the psychological processes underlying these behaviours. While the motivation for unhealthy food has been researched extensively, the motivation for sugary drinks remains understudied, despite their negative health implications. Up to 19% of daily calorie intake consists of sugar from drinks, and the consumption of sugary drinks contributes to weight gain. The consumption of sugary drinks is a main contributor to poor dental health and to overweight, which cost the NHS 3.4 billon and 4.7 billion a year in England alone (Public Health England, 2014). Especially given the recent media attention, many consumers are aware of the health implications of sugary drinks, but struggle to successfully reduce their intake. Therefore, it is important to understand what underlies the motivation for sugary drinks, and how we can effectively assist consumers in replacing sugary drinks with healthier alternatives such as water. We propose that sugary drinks gain their attractiveness through consumption and reward simulations. In other words, when people see or think about a sugary drink, they spontaneously simulate (i.e., re-experience) the sensation and the reward of consuming it, such as its taste, the resulting energy boost, and the quenching of thirst, based on their previous, rewarding experiences. These simulations trigger a desire to consume sugary drinks, particularly when feeling thirsty. Although evidence exists for the role of such simulations in the motivation for food, no previous studies have applied this account to drinks. Our research will first systematically test this simulation account of the motivation for sugary drinks, and then use it to stimulate healthier choices in innovative ways. In Subproject 1, we will investigate the specific simulations that are triggered by sugary drinks and by water. Building on recent pilot data that we have collected, we expect that sugary drinks will trigger more consumption and reward simulations than water, particularly among high consumers of sugary drinks, and particularly when thirsty. In Subproject 2, we will link these consumption and reward simulations to the motivation to consume sugary drinks and water. To this end, we will use a novel method to assess motivation unobtrusively: we will measure the degree to which participants slightly lean forward on a Wii balance board when viewing images of drinks. Such subtle approach movements have been shown to reflect motivation and desire. We predict that more consumption and reward simulations will be associated with leaning forward more toward sugary drinks images, especially among high consumers of sugary drinks and especially when thirsty. Finally, in Subproject 3, we will use these findings to develop an intervention approach to help consumers replace sugary drinks with water. Typically, advertisements for sugary drinks focus heavily on consumption and reward, whereas advertisements for water focus on purity and health benefits. We propose that motivation for consuming water can be increased by boosting consumption and reward simulations, in a similar way as for sugary drinks. Thus, we will test whether using images and words that trigger consumption and reward simulations for drinking water makes water more attractive and increases water choices, and reduces choices for sugary drinks. We will test this in both online and field experiments with actual consumers in naturalistic settings. Together, these experiments will help us understand what makes sugary drinks so difficult to resist, and how health practitioners, intervention developers, and industry can boost the motivational appeal of healthier alternatives to stimulate healthier beverage choices.

Two experiments were conducted online using Prolific, with UK populations. Participants used short phrases or words to describe sugary drinks and water, and rated all drinks on a number of dimensions (e.g., frequency of consumption, desire to consume). In Exp. 3, this was replicated with Glasgow University participants in a lab setting, and actual intake behaviour was measured as well (amount drunk). The words/phrases were coded into categories, and the use of the categories was analysed statistically.