what workslast verified 2026-09-24

does changing what you eat treat depression?

nutritional psychiatry has a founding trial, a small one, with a big effect and a wide confidence interval, and a decade of headlines built on it. it also has a 1,025-person trial that found nothing, a 45,826-person meta-analysis that found something small, and a cardiology trial whose depression result missed significance. all of it is here. the honest answer is that a better diet probably helps a little, the trials that found a lot were small and unblinded, and nobody has shown it works as well as the things it gets compared to.

five questions, answered from the trials

has anyone randomised people with depression to a better diet?[1][3][4]

three trials, all Australian, all Mediterranean-style. SMILES, 2017: 67 adults with moderate to severe depression, seven sessions with a dietician against a social-support control matched for time; remission 32.3 percent against 8.0 percent at twelve weeks, number needed to treat 4.1. AMMEND, 2022: 72 young men, 18 to 25, diet against befriending; a 14.4-point advantage on the Beck inventory. HELFIMED, 2019: 152 adults with self-reported depression, food hampers and cooking classes plus fish oil against social groups; a larger fall in depression at three months.

how big is the effect when you pool the trials?[5]

small. the 2019 meta-analysis of 16 randomised trials, 45,826 people, put the pooled effect on depressive symptoms at g 0.16, with a confidence interval from 0.06 to 0.27. fifteen of the sixteen trials were in people without clinical depression, so most of that number is about mood in the general population, not treatment of a disorder. for anxiety there was no effect at all. women benefited more than men.

what did the big trials find?[7][8]

PREDIMED, the Spanish cardiovascular trial, tracked 224 new cases of depression over at least three years in older adults at high heart risk: the Mediterranean-with-nuts arm had a hazard ratio of 0.78 with an interval crossing 1, “not significant”; only a subgroup with type 2 diabetes reached significance. MooDFOOD randomised 1,025 overweight adults with low-grade symptoms to a nutrient supplement, food-focused behavioural therapy, both or neither, for a year: 10 percent developed depression, and neither intervention changed that. the authors: “these findings do not support the use of these interventions for prevention.”

why do the observational studies look so much better?[5][6]

because people who eat well differ from people who do not in a hundred other ways. across four longitudinal cohorts the top Mediterranean-adherence group had a relative risk of 0.67 for new depression against the bottom group, and a less inflammatory diet 0.76. those are associations. the randomised trials, which are the only design that can separate the diet from the person, give a much smaller number, and that gap is the usual shape of nutrition science.

what is the strongest case for diet as treatment?[1][2]

SMILES: a randomised trial in clinically depressed adults with an active control, a large effect, and a remission difference that survived sensitivity analysis. the same trial is also the case against overreading: 67 people enrolled, 56 with complete data at twelve weeks, single-blind (the participants knew which arm they were in), 55 of 67 already on therapy or medication or both, and a correction two years later noting the recruitment methods had not all been disclosed. an NNT of 4.1 with an interval from 2.3 to 27.8 means the true number could be anywhere from excellent to nearly useless.

what the trials were, and what they were not

every one of the positive trials compared eating better against talking to someone: social support in SMILES, befriending in AMMEND, social groups in HELFIMED.[1][3][4] that is a fair control for attention, and it means the result is “diet beat company,” not “diet beat treatment.” no trial on this page compared a diet against an antidepressant or against a course of therapy, and the people in SMILES were mostly already on one or both.[1] none of them was blinded in the way a drug trial is; you know whether you were given a food plan. the effect sizes in the small trials are the size you see in unblinded trials with attentive interventions, and the pooled effect across sixteen trials is the size you see once the large, duller studies are counted.[5]

the other direction

the case for diet is not nothing, and the null trials do not erase it. MooDFOOD tested prevention in people who did not have depression yet, with a supplement and a behavioural program rather than a whole-diet change, so it does not test what SMILES tested.[8] PREDIMED was a heart trial in people over 55, with depression as a secondary outcome, and its point estimate still leaned the right way.[7] the meta-analysis found a real if small effect that held up in the higher-quality trials and against active controls.[5] and the observational signal is consistent across cohorts and dietary indices in the same direction.[6] a small effect from something with no known harm is not a small thing. it is only a small thing if it is sold as a cure.

what to ask

questions people ask

Can a Mediterranean diet treat depression?

Three small randomised trials in Australia found that a Mediterranean-style diet reduced depressive symptoms more than social support or befriending over 12 weeks: SMILES (67 adults, remission 32.3 percent vs 8.0 percent), AMMEND (72 young men, a 14.4-point BDI-II difference) and HELFIMED (152 adults, with fish oil). Pooled across 16 trials and 45,826 people, mostly without clinical depression, the effect is small (g 0.16). Nothing here is a recommendation to change any treatment.

What did the SMILES trial actually find?

In 67 adults with moderate to severe depression, seven sessions of dietary counselling over 12 weeks produced a larger improvement on the MADRS than a matched social-support control (Cohen’s d 1.16), with remission in 32.3 percent vs 8.0 percent and a number needed to treat of 4.1 (95% CI 2.3 to 27.8). The trial was single-blind, 56 of 67 had complete data at 12 weeks, and 55 of the 67 were already receiving psychotherapy, medication or both.

Is there a trial that found diet did not help?

Yes. MooDFOOD (JAMA 2019) randomised 1,025 overweight adults with low-grade depressive symptoms to a multinutrient supplement, food-related behavioural activation therapy, both or placebo for one year. About 10 percent developed major depression and none of the interventions changed that (supplement OR 1.06, therapy OR 0.93). In PREDIMED, the Mediterranean-diet arm’s hazard ratio for new depression was 0.78 with a confidence interval from 0.55 to 1.10, which the authors called not significant.

Does this page say anything about medication or supplements?

No. It is about whole-diet trials only. The supplement evidence, including omega-3, is on the supplements pages, and nothing on this page is a reason to start, stop or change any treatment on your own.

sources

  1. Jacka FN, O’Neil A, Opie R, et al. A randomised controlled trial of dietary improvement for adults with major depression (the ‘SMILES’ trial). BMC Medicine 2017;15:23. PMID 28137247. “12-week, parallel-group, single blind”; “seven individual nutritional consulting sessions delivered by a clinical dietician”; control “a social support protocol to the same visit schedule and length.” “We assessed 166 individuals for eligibility, of whom 67 were enrolled (diet intervention, n = 33; control, n = 34). Of these, 55 were utilising some form of therapy”; “31 in the diet support group and 25 in the social support control group who had complete data at 12 weeks”; MADRS “Cohen’s d = -1.16”; “Remission, defined as a MADRS score <10, was achieved for 32.3% (n = 10) and 8.0% (n = 2)”; “number needed to treat (NNT) based on remission scores was 4.1 (95% CI of NNT 2.3-27.8).” https://doi.org/10.1186/s12916-017-0791-y
  2. Jacka FN, O’Neil A, Opie R, et al. Correction to: A randomised controlled trial of dietary improvement for adults with major depression (the ‘SMILES’ trial). BMC Medicine 2018;16:236. PMID 30591046. “The original version of this paper did not specify that a website was used in the final year of recruitment, in addition to the other stated recruitment methods.” https://doi.org/10.1186/s12916-018-1220-6
  3. Parletta N, Zarnowiecki D, Cho J, et al. A Mediterranean-style dietary intervention supplemented with fish oil improves diet quality and mental health in people with depression: a randomized controlled trial (HELFIMED). Nutritional Neuroscience 2019;22(7):474-487. PMID 29215971. “Adults with self-reported depression were randomized to receive fortnightly food hampers and MedDiet cooking workshops for 3 months and fish oil supplements for 6 months, or attend social groups”; “n = 152 eligible adults aged 18-65 were recruited (n = 95 completed 3-month and n = 85 completed 6-month assessments)”; “The MedDiet group had greater reduction in depression (t = -2.24, P = 0.03)… at 3 months.” https://doi.org/10.1080/1028415X.2017.1411320
  4. Bayes J, Schloss J, Sibbritt D. The effect of a Mediterranean diet on the symptoms of depression in young males (the “AMMEND” study): a randomized controlled trial. American Journal of Clinical Nutrition 2022;116(2):572-580. PMID 35441666. “A 12-wk, parallel-group, open-label, randomized controlled trial… in young males (18-25 y). Befriending therapy was chosen for the control group”; “A total of 72 participants completed the study”; BDI-II “mean difference: 14.4; 95% CI: 11.41, 17.39; P < 0.001.” https://doi.org/10.1093/ajcn/nqac106
  5. Firth J, Marx W, Dash S, et al. The effects of dietary improvement on symptoms of depression and anxiety: a meta-analysis of randomized controlled trials. Psychosomatic Medicine 2019;81(3):265-280. PMID 30720698. “Sixteen eligible randomized controlled trials… with outcome data for 45,826 participants were included; the majority of which examined samples with nonclinical depression (n = 15 studies)”; “dietary interventions significantly reduced depressive symptoms (g = 0.162, 95% CI = 0.055 to 0.269, p = 0.003)”; “No effect of dietary interventions was observed for anxiety (k = 11, n = 2270, g = 0.085, 95% C.I. = -0.031 to 0.202, p=0.151)”; “Studies with female samples observed significantly greater benefits.” https://doi.org/10.1097/PSY.0000000000000673
  6. Lassale C, Batty GD, Baghdadli A, et al. Healthy dietary indices and risk of depressive outcomes: a systematic review and meta-analysis of observational studies. Molecular Psychiatry 2019;24(7):965-986. PMID 30254236. “A total of 20 longitudinal and 21 cross-sectional studies were included”; “The most compelling evidence was found for the Mediterranean diet and incident depression, with a combined relative risk estimate of highest vs. lowest adherence category from four longitudinal studies of 0.67 (95% CI 0.55-0.82)”; lower Dietary Inflammatory Index “relative risk 0.76; 95% CI: 0.63-0.92.” Two corrections were published (Molecular Psychiatry, July 2019, PMID 30464329, and July 2021, PMID 33664476). https://doi.org/10.1038/s41380-018-0237-8
  7. Sánchez-Villegas A, Martínez-González MA, Estruch R, et al. Mediterranean dietary pattern and depression: the PREDIMED randomized trial. BMC Medicine 2013;11:208. PMID 24229349. Men 55-80 and women 60-80 at high cardiovascular risk, “51% of them had type 2 diabetes”; “We identified 224 new cases of depression during follow-up. There was an inverse association with depression for participants assigned to a Mediterranean diet supplemented with nuts (multivariate hazard ratio (HR) 0.78; 95% confidence interval (CI) 0.55 to 1.10) compared with participants assigned to the control group, although this was not significant. However, when the analysis was restricted to participants with DM2… (multivariate HR = 0.59; 95% CI 0.36 to 0.98).” https://doi.org/10.1186/1741-7015-11-208
  8. Bot M, Brouwer IA, Roca M, et al. Effect of multinutrient supplementation and food-related behavioral activation therapy on prevention of major depressive disorder among overweight or obese adults with subsyndromal depressive symptoms: the MooDFOOD randomized clinical trial. JAMA 2019;321(9):858-868. PMID 30835307. “A total of 1025 adults were randomized… and followed up for 1 year”; supplements were “1412-mg omega-3 fatty acids, 30-μg selenium, 400-μg folic acid, and 20-μg vitamin D3 plus 100-mg calcium”; “During the 12-month follow-up, 105 (10%) developed MDD”; “None of the treatment strategies affected MDD onset. The odds ratio (OR) for supplements was 1.06 (95% CI, 0.87-1.29); for therapy, 0.93 (95% CI, 0.76-1.13)”; “These findings do not support the use of these interventions for prevention of major depressive disorder.” https://doi.org/10.1001/jama.2019.0556

related: what works that costs nothing · do any supplements actually work? · omega-3

adam