A randomised controlled trial of dietary improvement for adults with major depression (the ‘SMILES’ trial)

12-week, parallel-group, single-blind randomized controlled trial conducted at two Australian sites (Geelong and Melbourne). Participants: adults ≥18 years with major depressive disorder diagnosed by DSM-IV-TR criteria; total randomized 67 (dietary intervention n=33; social support control n=34). 55 participants were receiving some form of therapy (21 psychotherapy+pharmacotherapy; 9 psychotherapy alone; 25 pharmacotherapy alone). Baseline characteristics: mean age 40.3 (SD 13.1); 71.6% female; 71.2% had comorbid disorders; BMI 29.5 (SD 8.0).

ModiMedDiet dietary program for 12 weeks: seven 60-minute sessions of personalized dietary support delivered by an Accredited Practising Dietitian; sessions: first four weekly, last three every two weeks; diet emphasizes a ModiMedDiet (modified Mediterranean) pattern focusing on 12 key food groups (whole grains 5-8 servings/day; vegetables 6/day; fruit 3/day; legumes 3-4/week; low-fat/unsweetened dairy 2-3/day; raw unsalted nuts 1/day; fish at least 2/week; lean red meats 3-4/week; chicken 2-3/week; eggs up to 6/week; olive oil 3 tbsp/day) with limited extras (sweets, refined cereals, fried foods, fast foods, processed meats, sugary drinks) to <=3 per week; alcohol limited to wine with meals (up to ~2 standard drinks/day; more than 2 counted as extras); diet was ad libitum with no weight-loss focus; participants received a food hamper, recipes and 7-day food diaries; motivational interviewing used to support adherence.

Primary outcome: MADRS score improved more with the dietary intervention than control at 12 weeks (mean change from baseline to 12 weeks: DS −11.3; SS −4.2; between-group difference 7.1 points; t(60.7)=4.38; p<0.001). Remission (MADRS <10) reached by 32.3% in DS vs 8.0% in SS; number needed to treat = 4.1 (95% CI 2.3-27.8). Sensitivity analyses showed results robust to missing data. Secondary outcomes: greater improvements in HADS-D and HADS-A and CGI-I in DS; ModiMedDiet score rose more in DS, indicating better adherence; BMI and overall well-being (WHO-5) did not differ; some biomarkers (polyunsaturated fatty acids) shifted modestly, but changes did not correlate with MADRS improvements. Conclusion: Dietary improvement may provide an efficacious and accessible adjunct treatment for major depressive disorder, with potential benefits for co-morbidities; findings support integrating dietetic dietary support into depression treatment.

Small sample size and underpower relative to planned sample; short 12-week duration; differential dropout between groups (more completers in dietary group); lack of complete blinding and potential expectation biases; possible selection bias and limited generalisability to broader populations; concomitant therapies varied across participants and may confound results; not all results were powered for secondary outcomes; some findings rely on self-reported dietary adherence.