Can a Mediterranean Diet Prevent Depression? The Evidence

A Mediterranean-style diet may help some people with depressive symptoms, but it has not proven that it can prevent depression in healthy adults. The largest review in this field pooled more than 700,000 participants across 92 prospective cohorts and 21 randomized trials. It found a modest association between higher Mediterranean diet adherence and lower self-reported depressive symptoms: odds ratio (OR) 0.91, 95% confidence interval (CI) 0.88 to 0.96. Yet the same review found strong evidence of no prevention effect in randomized trials and no association when studies used clinical depression diagnoses instead of symptom questionnaires.

That split is the story. A very large dataset can estimate a small association precisely. It cannot, by itself, prove that olive oil, beans, fish, and vegetables caused the difference.

What "diet quality" means here

The Mediterranean pattern emphasizes vegetables, fruit, legumes, whole grains, nuts, olive oil, and fish, with less highly processed food and red meat. The related MIND diet combines Mediterranean and DASH-style principles. In a 2025 prospective multi-cohort study that compared 10 dietary indices across UK Biobank and two independent cohorts, the MIND diet performed best for depression risk among the indices assessed. That is useful comparative evidence, but it remains observational: people who sustain a high-quality diet also tend to differ in income, sleep, physical activity, social connection, illness burden, and early mood symptoms.

Food-frequency questionnaires and 24-hour recalls add another problem. They depend on memory and reporting. A person whose mood, appetite, or routine is already changing may report and eat differently long before receiving a depression diagnosis. That can make worsening mood look like a dietary cause when it is partly a dietary consequence.

The headline association is real, but small

The 2025 Molero et al. systematic review and meta-analysis found that greater Mediterranean diet adherence was statistically associated with lower odds of self-reported depressive symptoms (OR 0.91, 95% CI 0.88 to 0.96). An odds ratio below 1 points toward lower odds; it does not tell us how much depression a diet prevents in an individual person.

For scale only, applying an OR of 0.91 to an illustrative 7% annual baseline incidence would imply about a 0.6-percentage-point absolute difference and an approximate number needed to treat of 167. That calculation is not a result from one trial, and odds ratios do not translate perfectly into risk differences. It does show why a statistically significant result can still have modest clinical consequences for an individual.

The review's Bayesian assessment rated the Mediterranean diet signal only "anecdotal-to-moderate." It also found no dose-response relationship between greater Mediterranean diet adherence and depression risk. If a dietary exposure causes an outcome, more exposure does not have to produce a perfectly linear benefit, but the absence of any dose-response pattern weakens the causal case.

Randomized trials answer a narrower, more important question

Randomized trials do not erase every bias in nutrition research, but they do better than cohorts at testing whether an intervention changes an outcome. Here, their results depend on the population and question.

In people already experiencing depressive symptoms, the Molero review pooled treatment trials at standardized mean difference d = −0.80 (95% CI −1.46 to −0.14). A 2024 meta-analysis of randomized Mediterranean diet trials also found symptom improvement in adults with depressive symptoms. The direction is encouraging, but the confidence interval in the larger synthesis is wide, spanning nearly 1.3 standardized units. Results varied substantially across trials, so the pooled number does not predict what any one patient should expect.

Diet trials have a stubborn design problem: participants know whether they are getting intensive dietary counseling. More contact with a clinician, clearer routines, expectation of benefit, and the general effect of being supported can improve symptoms. Those effects matter to patients, but they do not establish that diet composition alone produced the change.

The PREDIDEP randomized trial, with 168 recovered depressed patients, illustrates the more targeted end of this research. It tested a Mediterranean diet-based intervention for relapse prevention and subsyndromal symptom management in a clinically high-risk group, not healthy volunteers. Its relevance is as an adjunctive strategy for people with a depression history; it does not answer whether dietary change prevents first-onset depression in the general population.

On that prevention question, the Molero review found strong evidence of no association in prevention randomized controlled trials among non-clinical populations. This null result should carry more weight than the observational headline for claims that a healthy diet insures someone against depression. It does not make diet irrelevant to health. It means the evidence has not shown that changing diet prevents depression.

Why the observational and randomized results may diverge

The largest review found diet associations only when studies used self-reported depression scales, not clinical diagnoses. Several explanations fit: diet might influence subthreshold symptoms without preventing diagnosable depression; dietary recall and mood reporting might share measurement bias; or social desirability could affect both. The data cannot distinguish these explanations cleanly.

Other prospective evidence also cautions against an easy causal story. A 2025 UK MS Register study found cross-sectional associations between Mediterranean diet and mental-health outcomes in people with multiple sclerosis, but no prospective associations. That pattern is consistent with reverse causation: poorer mental health can precede dietary deterioration. It does not prove reverse causation in every cohort, but it shows how a cross-sectional signal can evaporate over time.

Similarly, the 2024 REGICOR population-based cohort study noted that several recent prospective studies failed to reproduce inverse diet-depression associations after adequate confounder adjustment. A 2021 review in middle-aged and older adults also documented inconsistent results. Larger samples do not fix incomplete adjustment for factors that travel with both diet and mental health.

Two practical signals deserve separate treatment

Postpartum fish and omega-3 intake

The strongest specific finding in the Molero synthesis appeared in postpartum populations. Fish consumption had a dose-response association with lower depression risk (OR 0.84, 95% CI 0.78 to 0.91), as did DHA/EPA intake (OR 0.84, 95% CI 0.77 to 0.92). Pregnancy and lactation shift maternal omega-3 stores toward fetal and infant development, which gives the finding biological plausibility. A 2024 Alberta pregnancy cohort also examined pre-pregnancy diet and prenatal depressive symptoms in 1,141 participants; it reported lower risk in white participants in the highest Mediterranean diet tertile versus the lowest (OR 0.56, 95% CI 0.33 to 0.95), while its overall tertile analyses found no association.

These results do not establish a universal omega-3 antidepressant effect. Fish was null in non-postpartum cohorts in the Molero review. The postpartum signal is more credible because it showed dose-response, but it remains population-specific.

Ultra-processed food

A 2023 Mediterranean cohort study found ultra-processed food intake independently associated with depressive symptoms after adjustment for overall diet-quality scores. This makes reducing ultra-processed foods a practical target, though the cross-sectional design cannot determine direction: depressive symptoms can also make convenient, highly processed food more appealing or accessible.

Mechanistic work offers hypotheses, not proof. A 2026 systematic review of butyrate and depression describes how fiber fermentation products may affect HPA-axis signaling and serotonin-related gut epithelial signaling. That biology fits a fiber-rich dietary pattern. It cannot rescue a prevention claim that randomized trials have not supported.

What to do with this evidence

• If you have depressive symptoms, a Mediterranean-style dietary intervention is reasonable alongside standard treatment. Randomized evidence supports symptom improvement, but the intervention should not replace psychotherapy, medication when indicated, or clinical care.
• If you are pregnant or postpartum, prioritize adequate fish and DHA/EPA intake within your clinician's dietary and food-safety guidance. This is the only finding here with a reported dose-response relationship.
• Reducing ultra-processed foods is a sensible health goal with an independent observational signal for depressive symptoms, although causation remains unproven.
• If you are otherwise healthy, do not treat dietary improvement as depression insurance. Prevention trials were null.

Diet quality belongs in a depression care plan because it affects overall health and may improve symptoms for some people. The evidence does not support selling it as a proven way to prevent depression.

If depression includes suicidal thoughts, inability to function, or rapidly worsening symptoms, seek urgent professional help. A grocery list is not crisis care.