A genetic signal linked to emotional eating
A seven-cohort study found one chromosome 5 region associated with emotional-eating scores, but it did not identify a causal gene or a clinical test.
The 60-second version
A seven-cohort GWAS linked one chromosome 5 region to emotional-eating scores, without identifying a causal gene or clinical test.
Key points
- The analysis covered 11,250 participants of European ancestry and three questionnaire-based eating traits.
- Two tightly linked variants represented one signal; the effect was small and consistent across the seven cohorts.
- The association no longer met genome-wide significance without BMI adjustment, and independent replication is still absent.
- The nearby processed pseudogene is only a candidate location; functional evidence did not establish a mechanism.
- Genetic influence can coexist with psychological, social, and environmental causes and does not make behavior inevitable.
Verdict. The result is useful evidence against blame-based explanations, but it is not yet a basis for prediction, diagnosis, or treatment.
A meta-analysis of seven cohorts found one chromosome 5 region associated with emotional-eating scores. The result supports a possible biological contribution, but it does not reveal a “stress-eating gene,” predict an individual's behavior, or justify a clinical genetic test.
Study designWhat the researchers measured
The team combined genetic and questionnaire data from seven studies of people of European ancestry. Versions of the Three-Factor Eating Questionnaire measured three related traits, while cohort-level genetic models adjusted for at least age, sex, BMI, and ancestry-related principal components.
| Emotional eating | Eating more during difficult moods such as loneliness, anxiety, or low mood. |
|---|---|
| Cognitive restraint | Deliberately limiting food intake to influence body weight. |
| Uncontrolled eating | A tendency to eat more with a sense of lost control and hunger. |
Roughly 10.85 million variants passed quality control. The primary analysis used the usual genome-wide threshold of P < 5×10^-8, although the authors did not further tighten that threshold for testing three eating traits.
Main resultOne signal, represented by two variants
The variants rs6877636 and rs6897090 were each associated with slightly higher emotional-eating scores. Their effects pointed in the same direction in every cohort, and leave-one-study-out analyses did not identify a single study driving the result.
No region reached genome-wide significance for cognitive restraint or uncontrolled eating. Several other results, including an FTO association with cognitive restraint, were only suggestive and require larger studies.
InterpretationThe causal biology is still unknown
The strongest marker sits within RP11-24P24.1, a processed pseudogene related to ODC1. That location is a lead, not a mechanism: the 99% credible set contained 32 variants, the top variant had only a 9% posterior probability of being causal, and the pseudogene's biological relevance remains unclear.
| With BMI adjustment | The strongest variant met genome-wide significance at P = 3.2×10^-8. |
|---|---|
| Without BMI adjustment | The effect stayed positive but fell short of genome-wide significance at P = 2.13×10^-6. |
| Gene-expression follow-up | No tested gene association survived correction for multiple comparisons. |
ODC1 encodes an enzyme in polyamine synthesis, but a known function of the parent gene does not establish that this chromosome 5 signal changes human eating behavior through ODC1. Functional experiments are still needed.
Shared influencesWhat the BMI overlap does and does not mean
Known BMI-associated variants were enriched among nominal emotional-eating associations, and additional polygenic-score and twin analyses supported some shared genetic contribution. These are population-level correlations. They do not show that emotional eating inevitably causes obesity or that inherited variants dictate food choices.
A genetic contribution can reduce blame without turning behavior into destiny.
Evidence boundaryWhy this finding needs replication
- Ancestry: all analyzed participants were of European ancestry, limiting transfer to other populations.
- Measurement: cohorts used different questionnaire versions and harmonized related but non-identical scales.
- Power: 11,250 participants is modest for a GWAS, especially across three behavioral traits.
- Mechanism: neither the causal variant nor the causal gene has been established.
- Replication: the seven-cohort consistency is reassuring, but no independent external sample confirmed the signal.
Practical meaningLess stigma, not a genetic label
The study supports moving away from the idea that emotional eating simply reflects weak willpower. It does not make psychological, social, or environmental influences less important. For someone distressed by the behavior, the relevant options remain nonjudgmental support, emotion-regulation strategies, dietary counseling, and help distinguishing physical hunger from eating triggered by emotion.
The next evidence to watch is independent replication in larger and more diverse cohorts, using consistent measures, followed by experiments that identify the functional variant and pathway. Until then, this is an informative association, not a diagnostic result.