Nearly Three in Four Popular Comfort Foods Are Hyperpalatable, Study Finds

Between December 2019 and January 2020, a national online survey asked American adults a simple question: what do you eat when you want to feel better? The answers, now analyzed in a paper in the journal Nutrients, give an unusually broad picture of what comfort food actually looks like in the United States, and what is in it.

The survey, called the UCLA Eating in America Study, recruited 2,022 adults aged 18 or older through Qualtrics, using quotas matched to the US census. Each person could name up to three comfort foods and report their age, gender, and region. Of the 1,760 respondents who supplied usable comfort-food answers, 1,228, or 69.8%, named at least one food that could be nutritionally assessed.

Those responses added up to 300 distinct foods and beverages, which the researchers folded into 132 groups based on ingredients and preparation. Two trained research assistants sorted the answers independently, using a standard protocol. Nutrition data came from USDA databases and, when needed, brand-specific labels. The most frequently named groups were chips, ice cream, pizza, chocolate, and cookies. Candy, hamburgers, cakes, pasta, bread, cheese, macaroni and cheese, fries, sodas, and chicken dishes also appeared. So did nuts, fruits, eggs, vegetables, and salads, which made the top 50 groups.

Across the 300 foods, the unweighted average was 223.1 calories per estimated serving, with every food counted equally regardless of how often it was named. On average, 39.2% of calories came from fat and 18.3% from sugar, and fiber averaged 1.5 grams per serving. By the study's measures, 60.7% of the foods were low in fiber, 29.7% were high in saturated fat, 31.7% were high in sodium, 23.3% were high in total fat, and 17.0% were high in total sugar. The sugar figure reflects total sugar, not added sugar, measured against a UK reference intake of 90 grams per day.

The headline finding concerns hyperpalatability, a classification based on particular combinations of fat, sugar, sodium, and carbohydrates. In all, 72.7% of the foods met it. The most common type was the high-fat, high-sodium meat or meal category, at 54.3%. Foods high in fat and sugar made up 21.3%, and those high in carbohydrates and sodium 11.0%. Some foods qualified under more than one criterion.

For ultra-processing, the team used the Nova classification and applied a strict test: a food counted only if 100% of its calories came from Nova Group 4. By that standard, 42.7% of foods were ultra-processed, and the authors note the threshold likely understates how widespread ultra-processing is. Some 36.0% of foods were both hyperpalatable and ultra-processed. Every food within the four most popular groups, chips, ice cream, pizza, and chocolate, met both labels. Pasta was the only top-10 item that met neither.

Demographic differences were statistically significant but mostly small. Men's choices averaged about 33 more calories per serving than women's, and were higher in protein, saturated fat, fiber, and sodium, while women's choices contained more total sugar. Adults 55 and older named foods lower in calories, protein, fat, carbohydrates, fiber, and sodium than those aged 18 to 34. Northeasterners named foods higher in several nutrients than Southerners and Midwesterners. Most gaps amounted to less than 5% of the recommended daily intake.

The authors flag limits. The survey recorded which foods people named, not how much or how often they ate them, and it did not measure stress. Nutrition values relied on standard food descriptions. The sample was non-random and was collected before the COVID-19 pandemic and before GLP-1 medications became widely used. The authors suggest that promoting healthier versions of popular comfort foods may be more practical than targeting demographic groups, though the study did not test that idea.