Why do we turn to comfort food? A survey of 1,203 adults offers clues

Chocolate, strawberries and roast dinners all featured as researchers examined how anticipated feelings about food relate to what people eat..

An observational study of food-related emotional expectancies, diet quality and comfort eating in the UK adult population. Image Credit: Master1305 / Shutterstock

In a recent study published in the British Journal of Nutrition, researchers conducted a cross-sectional analysis to examine associations between food-related emotional expectations, diet quality, and comfort eating among UK adults.

The study sample comprised 1,203 adults whose dietary and sociodemographic data were assessed alongside their responses to food images, allowing the researchers to examine how branding and nutritional classification relate to anticipated emotional effects. The researchers also examined expectations of negative emotions and their relationship with comfort eating.

The study found that participants, on average, expected high-fat, high-salt, and/or high-sugar (HFSS) foods to elicit more positive emotions than non-HFSS foods, but branding did not consistently increase these expectations. The analyses also identified small associations between positive expectancies for HFSS foods and poorer diet quality and more frequent comfort eating.

Background

Poor diet quality remains a major public health concern in the United Kingdom (UK). According to national dietary data cited by the researchers, an estimated 82% of UK adults consume excessive amounts of fat, 70% consume excessive amounts of salt, and 81% consume excessive amounts of sugar. In the present sample, 58.9% had diet-quality scores below 12, which the researchers classified as inadequate.

More recent research has investigated the relationship between comfort eating, defined as eating in response to negative emotions, and weight-related outcomes. Previous evidence cited by the authors estimates that around one in six adults comfort eat often. Among these participants, 54% reported comfort eating more frequently than “hardly ever”, including 26% who did so fairly or very often.

While various psychological models have been developed to describe emotional eating, expectancy theory offers an explanation for why people may continue learned eating patterns. It proposes that memories of earlier emotional experiences influence what people expect from eating particular foods and, potentially, their later choices.

The interactive effects of branding and HFSS classification on expectancies for individual foods had not previously been examined in UK adults.

About the study

The present study aimed to address this knowledge gap and inform dietary public health policies by quantifying food-related emotional expectancies in a UK adult sample. The study was conducted using the Prolific platform and enrolled 1,203 UK adults (mean age 46.7 years; 52.0% women; 86.6% White), with recruitment designed to reflect the national population by age, sex and race/ethnicity.

The survey asked each participant to view 18 food and non-alcoholic beverage images from a collection of 36, including 12 generic and six branded items per participant, covering fast food, sweets, savory snacks, traditional and cultural meals, fruits, vegetables and drinks. Most images came from the Restrain Food Database, while beverage images were sourced from Freepik and Coca-Cola.

The nutritional profiles of the selected items were classified according to HFSS status, nutritional quality and level of processing.

The researchers quantified participant-specific expectancies using the Anticipated Effects of Food Scale-Brief (AEFS-Brief), a validated questionnaire. The researchers also assessed participants’ overall diet quality using the 20-item Short-form Food Frequency Questionnaire (SFFFQ), and comfort eating was measured using a single question rated on a five-point frequency scale.

Study findings

The study’s mean positive expectancy scores were highest for ice cream sundaes (mean 4.21), chocolate (4.08), strawberries (4.07), cake (3.99), and roast dinners (3.93), while the lowest scores were observed for prawn curry (2.91), salad (2.93), and carrots (2.96). The high rankings of strawberries and roast dinners showed that positive expectations were not restricted to HFSS foods.

Multilevel modeling identified a small interaction between HFSS classification and branding, but branding did not consistently increase anticipated positive emotions. The researchers cautioned that the comparison was limited by the small number of branded non-HFSS foods in the image set.

Main effects showed that HFSS foods were associated with slightly higher positive expectancy scores than non-HFSS foods, whereas branding was associated with slightly lower positive expectancy scores overall in the statistical model. Both effects were small.

Correlation analyses showed small associations between higher positive expectancies for HFSS items and poorer diet quality, particularly for sundaes, fizzy drinks, and chocolate. Among non-HFSS foods, salad was the only item for which greater positive expectancies were significantly associated with better diet quality.

Finally, more frequent comfort eating was associated with stronger positive expectancies for many HFSS foods and negative expectancies for nearly all foods, including non-HFSS foods. Associations involving negative expectancies were generally stronger than those involving positive HFSS expectancies.

Conclusions

The present study is among the first to quantitatively examine how food-related emotional expectancies are associated with poor diet quality and comfort eating in UK adults.

These results suggest that anticipated emotional rewards may be related to eating behaviors, alongside taste preferences and other influences. The researchers did not measure actual mood changes after eating, so the findings cannot show whether particular foods relieve distress. Because the study was cross-sectional and relied on self-reported responses to food photographs, it cannot establish cause and effect. The limited range of branded non-HFSS foods also restricts interpretation of the branding results.

Clinical practitioners could evaluate tailored “expectancy challenge” behavioral interventions, guiding patients to re-evaluate whether comfort foods genuinely alleviate distress. Policymakers could also consider restricting HFSS food advertising and promotional cues, but further research is needed to test whether either approach improves eating behavior.