Night owls are linked to how they eat. Not just when.

Meta-analysis of observational studies. 50,741 people. 51 studies (2001-2026) included in the review, 34 of them meta-analysed; search to February 2026.

Many studies, added up. People were watched, not told what to do. A clue, not proof.

Who: General (non-clinical) populations: mean ages 13.3-52.6 years and mean BMI 19.83-40.60 across studies; 23 of 51 studies in university students, 14 in the general population, 6 in people with overweight/obesity; 17 of 51 studies from Türkiye, 5 from the US. People diagnosed with eating disorders and pregnant/lactating women excluded

• Night eatinghigher11 studies
• Binge eatinghigheronly 3 studies
• Food cravingshighersmall link

More to know, from our read of the paper:

• Observational and almost all cross-sectional (one point in time): cannot show that being a night owl causes any eating pattern; the authors call for longitudinal studies and randomized trials
• Everything is questionnaire-based and self-reported: chronotype mostly as a morning/evening preference (MEQ in 34 of 51 studies), eating behaviours by retrospective questionnaires
• Effect sizes are correlations, mostly small: r = 0.06 (emotional eating) to 0.21 (night eating); only binge eating reached r = 0.39. Never translate r into '% of people' or 'X times more likely'
• Binge eating (r = 0.39) rests on only 3 studies (1,122 people), with very high heterogeneity (I2 = 94%) and significant publication bias (Egger p < 0.001; trim-and-fill imputed nothing). Present it as the strongest but thinnest result
• High heterogeneity for most outcomes (I2 70-94%); authors say results are 'general trends rather than uniform risks across individuals'; no meta-regression or subgroup analysis possible

This meta-analysis reviewed 51 observational studies with 50,741 people from the general population (average ages 13 to 53, many of them university students). People with a later body clock (night owls) reported more night eating, binge eating, food cravings and loss of control over eating, and less deliberate holding back on food. The links were mostly small, the binge-eating result came from only 3 studies, and nearly all studies were one-time questionnaires: linked, not proven.

We read the abstract and the full text of this paper and checked every number on this page against it (16 facts, checked on 2026-10-29). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.