Society: AGA

Background: Diet is known to alter the risk of depression. Increasing data also demonstrate a causal role of the gut microbiome in mental illness, via the gut-brain axis. However, it remains unclear how diet and the microbiome mechanistically influence depression risk in humans. Leveraging dietary, metabolomics, microbiome, and depression data, we assessed how gut microbial species and their pathways may mediate the association between depression and citrus, a food group that possibly protects against risk of depression.

Methods: We conducted a prospective study in the Nurses’ Health Study II (NHSII) between 2003 and 2017 among 32,427 middle-aged women free of depression at baseline. Citrus intake was determined using validated food frequency questionnaires collected every 4 years. Depression was defined according to physician-diagnosis and antidepressant use. Between 2013-2014, 207 NHSII participants enrolled in a nested substudy, providing up to 4 stool samples (profiled by shotgun metagenomics) and a blood sample (profiled by LC-MS-based metabolomics). Cox proportional hazard models were used to relate citrus intake with depression risk. Linear mixed effects models were used to relate diet with gut microbial features, and microbial features with depression. We also associated microbial features with a depression-risk score, derived according to levels of circulating serotonin and GABA. All models were adjusted for multiple dietary, medication and lifestyle variables including age, BMI, calorie/alcohol intake, and diet quality. We validated our findings in a subcohort of 307 men in the Health Professionals Follow-up Study (HPFS).

Results: Total citrus intake was associated with a lower risk of incident depression (p_trend 0.001), with a multivariable relative risk of 0.80 (95% CI, 0.68-0.93), comparing extreme quintiles. Within the NHSII substudy, greater citrus intake was associated with increased abundance of Faecalibacterium prausnitzii (β 0.026, FDR 0.17). In turn, levels of F. prausnitzii were higher in non-depressed individuals compared to depressed participants (p 0.003). Greater abundance of F. prausnitzii was also associated with our metabolomics-based depression-risk score in the NHSII (p 0.03), and in the HPFS validation study (p 0.02). In an exploratory analysis of gut microbial pathways, S-Adenosyl-L-Methionine (SAM) cycle I, encoded by F. prausnitzii, was reduced in depressed participants.

Conclusion: Greater citrus intake was prospectively associated with lower risk of depression, and with greater abundance of F. prausnitzii. In turn, participants with depression had lower levels of F. prausnitzii and lower abundance of its genes capable of producing SAM, a compound known to have antidepressant properties. These data offer a potential mechanism by which diet influences the gut microbiome to reduce risk of depression.

Allison Chan

Appearances