This Addiction study linked company ownership records with U.S. food data from 1988 to 2001 to test whether food design changed under tobacco ownership. Researchers compared 105 tobacco-owned foods with 587 non-tobacco-owned foods using a fixed nutrition rule for "hyper-palatable" design. Tobacco-owned products were 29% more likely to fit the fat+salt pattern and 80% more likely to fit the carb+salt pattern (p values 0.005 to 0.009), which supports deliberate formulation for stronger reward response.

This Obesity journal paper created a numeric definition for foods built to be especially rewarding to eat, then applied it to the U.S. national food database. Instead of subjective labels, the method used nutrient thresholds for three engineered combinations: fat+salt, fat+sugar, and carb+salt. Using that rule, 62% of foods met hyper-palatability criteria, showing that these high-reward formulations are common across the food supply.

This PLOS ONE paper tested which foods people connect most with addictive-like eating patterns. Study 1 used 120 university participants to rate 35 foods, and Study 2 used 384 adults to model which food features predicted problematic ratings. Processed foods high in fat and fast-absorbing carbs ranked highest, and the statistical model showed food processing as a large positive predictor of addictive-like ratings.

This Appetite study combined two large U.S. nurse cohorts and examined whether common processed foods were linked to food-addiction symptoms. Diet was measured in 2006 to 2007, and food addiction was measured in 2008 to 2009 with the Modified Yale Food Addiction Scale. Prevalence was 5.4%, and high intake of key ultra-processed foods showed much higher odds: burgers OR 4.08 (95% CI 2.66 to 6.25), fries OR 2.37 (1.59 to 3.51), and pizza OR 2.49 (1.67 to 3.69).

This systematic review and meta-analysis pooled Yale Food Addiction Scale data across many countries and sample types. From 6,425 screened abstracts, 272 studies were included in the analysis, and prevalence was estimated with random-effects models. The pooled diagnosis prevalence was 20% (95% CI 18% to 21%), and prevalence rose to 55% (95% CI 34% to 75%) in participants with binge-eating diagnoses.

This Obesity Reviews meta-analysis focused on children and teens and used the Yale scale designed for younger populations. The dataset included 22 cross-sectional studies with 6,996 participants. Estimated prevalence was 15% overall (95% CI 11% to 19%), 12% in community samples (8% to 17%), and 19% in overweight/obese samples (14% to 26%); weight status was linked to symptom severity (p = 0.002).

This Appetite study created a tool to measure withdrawal-like symptoms when children reduce highly processed foods. Parents of 304 children aged 3 to 11 reported behavior changes after recent attempts to cut highly processed foods. The scale had high internal consistency (alpha = 0.94), correlated with food-addiction symptoms (r = 0.55, p < 0.001), and was linked to lower success in reducing processed-food intake (p = 0.001).

This BMJ analysis reviews clinical and policy evidence on whether ultra-processed foods meet addiction-style criteria. It combines evidence on craving, loss of control, withdrawal-like symptoms, and repeat overconsumption patterns seen in highly processed products. The paper highlights prevalence around 14% in adults and 12% in children, supporting the view that addictive-like response to ultra-processed foods is measurable at population scale.