Examination of environmental, affective and cognitive drivers of food intake

Rates of overweight and obesity have dramatically increased over the past 30 years, placing individuals at higher risk of hypertension, type 2 diabetes, heart disease and stroke. As such, overweight and obesity are a public health issue posing significant risk to the individuals as well as burdening the broader community with associated financial costs. To address this public health concern, it is crucial to understand factors contributing to weight gain.
The idea that people only eat to restore homeostasis has been replaced with the understanding that a large proportion of food intake is driven by what is referred to as ‘hedonic hunger’: eating in the absence of true hunger, driven by reward and pleasure associated with food intake. Hedonic hunger is driven by both external (i.e., location, food availability) and internal (i.e., affect) factors. However, there are many individual factors (i.e., BMI, impulsivity, eating style, food-related attentional bias) that impact how one responds to eating cues.
Therefore, the objective of the present thesis was to examine how individual characteristics such as eating style, BMI and attentional bias interacted with momentary cues (i.e., affect, food outlet presence, location) to influence food intake. Food intake was assessed using Ecological Momentary Assessment (EMA), which allowed food intake to be measured in naturalistic environments as participants went about their daily lives. This, in conjunction with baseline assessments of individuals’ BMI, eating style, impulsivity, as well as behavioural tasks assessing attentional bias, enabled a greater understanding of how individual characteristics such as eating style, BMI and attentional bias interacted with momentary cues to influence food intake.
Four complimentary studies were conducted to address the overall thesis objective. This thesis begins by examining the impact of stimulus control on food intake, and whether this differed based on BMI status. In this study, participants recorded their food and drink intake in real-time and responded to randomly timed assessments assessing stimulus control (food cues: location, affective state, food availability and the presence of food outlets). The findings suggest that individuals were indeed under the influence of stimulus-controlled eating, and that BMI played a minor role. The stimulus control domain which could differentiate between eating and non-eating instances with statistically significant accuracy between those in the healthy-weight range compared to those with high BMIs was food outlet presence.
Study 2 (Chapter 3) explored relationship between affect and food intake in greater detail. Specifically, Study 2 explored the affective trajectories in the hours leading up to, and following, healthy and unhealthy snacking instances, and whether this differed based on eating style (restrained or emotional eating). Data from two similar EMA studies was pooled to explore this relationship. In both studies, participants completed baseline assessments measuring eating style, followed by 14 days of monitoring their food and drink intake. The findings from this study support the notion of ‘comfort eating’, such that individuals were more likely to choose an unhealthy snack following a drop in positive affect. This effect was the same across both eating styles.
Study 3 (Chapter 4) explored whether a relationship between increased food-related attentional bias and increased cue-related eating in the real-world (stimulus control) existed. Given the possibility that some individuals may experience heightened levels of attentional bias but do not “act” on it, impulsivity was explored as a moderator of this relationship. Our findings suggest there to be no relationship between laboratory measures of attentional bias and stimulus-controlled eating in the real-word on an individual level. While impulsivity predicted levels of stimulus-controlled eating in the way it would be expected (increased impulsivity predicted increased stimulus control), impulsivity did not moderate the relationship between attentional bias and stimulus control.
Finally, Study 4 (Chapter 5) explored the reliability of the visual probe task in assessing food-related attentional bias. Specifically, we investigated the reliability of the visual probe in capturing food-related attentional bias measured under the same conditions 14 days apart, as well as how the different versions of the visual probe task (stimuli presented in words vs. stimuli presented in pictures) and different scoring methods impacted the results. The findings suggest the reliability of the task to capture attentional bias 14 days apart to be poor, regardless of the scoring method used. This finding may explain the lack of association between attentional bias and stimulus control found in Study 3 (Chapter 4).
Taken together, the findings from this thesis highlight potential interventions for dietary change. These include reducing comfort eating by assisting individuals to develop emotion regulation and distress tolerance strategies, as well as implementing intuitive eating approaches to dietary intake. External cues to eating could be curbed by shaping the food environment. Within the public sphere, access to healthy food choices needs to be at the forefront of any dietary change initiative as individuals are most often guided by the food that is available to them. Similarly, reducing the number of fast-food outlets, and discretionary food / advertising will reduce environmental triggers nudging individuals towards unhealthy food intake. Importantly, both measurement of real-world food intake and attentional bias require refinement if we are to better understand cognitive, affective, and external drivers of food intake.