Consumption of Ultraprocessed Food and Risk of Depression

Increasing evidence suggests that diet may influence risk of depression.^1-3 Despite extensive data linking ultraprocessed foods (UPF; ie, energy-dense, palatable, and ready-to-eat items) with human disease,⁴ evidence examining the association between UPF consumption and depression is scant. Prior studies have been hampered by short-term dietary data^1,2 and a limited ability to account for potential confounders.^1,3 Additionally, no study has identified which UPF foods and/or ingredients that may be associated with risk of depression or how the timing of UPF consumption may be associated. Therefore, we investigated the prospective association between UPF and its components with incident depression.

This cohort study was approved by the institutional review board (IRB) at the Brigham and Women’s Hospital and the Harvard T.H. Chan School of Public Health. The return of a completed questionnaire was accepted by the IRB as implied informed consent. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

We conducted a prospective study in the Nurses’ Health Study II between 2003 and 2017 among middle-aged females free of depression at baseline. Diet was assessed using validated food frequency questionnaires (FFQs) every 4 years. We estimated UPF intake using the NOVA classification,² which groups foods according to the degree of their processing. In secondary analyses, we classified UPF into their components, including ultraprocessed grain foods, sweet snacks, ready-to-eat meals, fats and sauces, ultraprocessed dairy products, savory snacks, processed meat, beverages, and artificial sweeteners.⁴ We used 2 definitions for depression: (1) a strict definition requiring self-reported clinician–diagnosed depression and regular antidepressant use and (2) a broad definition requiring clinical diagnosis and/or antidepressant use.

We estimated hazard ratios (HRs) and 95% CIs for depression according to quintiles of UPF intake using Cox proportional hazards models, with adjustment for known and suspected risk factors for depression, including age, total caloric intake, body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), physical activity, smoking status, menopausal hormone therapy, total energy intake, alcohol, comorbidities (eg, diabetes, hypertension, dyslipidemia), median family income, social network levels, marital status, sleep duration, and pain. In an exploratory analysis, we examined the association between changes in UPF consumption updated every 4 years with incident depression. All analyses were performed using 2-sided tests from SAS (version 9.4). Data were analyzed from September 2022 to January 2023.

Our cohort included 31 712 females, aged 42 to 62 years at baseline (mean [SD] age, 52 [4.7] years; 30 190 [95.2%] non-Hispanic White females). Participants with high UPF intake had greater BMI, higher smoking rates, and increased prevalence of comorbidities like diabetes, hypertension, and dyslipidemia and were less likely to exercise regularly. We identified 2122 incident cases of depression using the strict definition and 4840 incident cases using the broad definition. Compared with those in the lowest quintile of UPF consumption, those in the highest quintile had an increased risk of depression, noted for both strict definition (HR, 1.49; 95% CI, 1.26-1.76; P < .001) and broad definition (HR, 1.34; 95% CI, 1.20-1.50; P < .001) (Table). Models were not materially altered after inclusion of potential confounders. We did not observe differential associations in subgroups defined by age, BMI, physical activity, or smoking. In a 4-year lag analysis, associations were not materially altered (strict definition: HR, 1.32; 95% CI, 1.13-1.54; P < .001), arguing against reverse causation.

Next, we examined the association of specific UPF components with risk of depression. Comparing extreme quintiles, only artificially sweetened beverages (HR, 1.37; 95% CI, 1.19-1.57; P < .001) and artificial sweeteners (HR, 1.26; 95% CI, 1.10-1.43; P < .001) were associated with greater risk of depression and after multivariable regression (Figure). In an exploratory analysis, those who reduced UPF intake by at least 3 servings per day were at lower risk of depression (strict definition: HR, 0.84; 95% CI, 0.71-0.99) compared with those with relatively stable intake in each 4-year period.

These findings suggest that greater UPF intake, particularly artificial sweeteners and artificially sweetened beverages, is associated with increased risk of depression. Although the mechanism associating UPF to depression is unknown, recent experimental data suggests that artificial sweeteners elicit purinergic transmission in the brain,⁵ which may be involved in the etiopathogenesis of depression.⁶ Strengths of our study include the large sample, prospective design, high follow-up rate, ability to adjust for multiple confounders, and extensively validated dietary assessment tools. This study had limitations. The cohort primarily included non-Hispanic White females. Additionally, without structured clinical interviews, misclassification of the outcome may be considered.

Accepted for Publication: August 15, 2023.

Published: September 20, 2023. doi:10.1001/jamanetworkopen.2023.34770

Correction: This article was corrected on October 18, 2023, to fix transcription errors in the Table.

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2023 Samuthpongtorn C et al. JAMA Network Open.

Corresponding Authors: Raaj S. Mehta, MD, MPH (rmehta2@mgh.harvard.edu), and Andrew T. Chan, MD, MPH (achan@mgh.harvard.edu), Clinical and Translational Epidemiology Unit, Massachusetts General Hospital and Harvard Medical School, 100 Cambridge St, Ste 1580 Boston, MA 02114.

Author Contributions: Drs Samuthpongtorn and Mehta had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Samuthpongtorn, Chan, Mehta.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Samuthpongtorn, Chan, Mehta.

Critical review of the manuscript for important intellectual content: All authors.

Statistical analysis: Samuthpongtorn, Chan, Mehta.

Obtained funding: Chan.

Administrative, technical, or material support: Samuthpongtorn, Okereke, Song, Chan, Mehta.

Supervision: Chan, Mehta.

Conflict of Interest Disclosures: Dr Okereke reported receiving grants from the National Institutes of Health and royalties from Springer Publishing outside the submitted work. Dr Chan reported receiving grants from Bayer Pharma AG and Zoe and personal fees from Boehringer Ingelheim, Pfizer, and Freenome outside the submitted work. No other disclosures were reported.

Funding/Support: The Nurses’ Health Study II was funded by grant U01 CA176726 from the National Cancer Institute, National Institutes of Health.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Data Sharing Statement: See the Supplement.

Additional Contributions: We thank the participants and staff of the Nurses’ Health Study II for their valuable contributions. They received no compensation for their contributions.