Association of Dietary Flavonoid Intake with Incident Depression Risk and Brain Structural Changes: A Prospective Study in the UK Biobank

ObjectivesTo prospectively evaluate the association of dietary flavonoid intake with incident depression risk and brain structural changes. DesignProspective cohort study. SettingUK Biobank. Participants114 848 non-depressed individuals (with [&ge;] 2 eligible 24-h dietary recalls) for depression risk and a subgroup of 2120 for brain structural changes. ExposuresTotal flavonoid, the Flavodiet Score (reflecting servings of top 10 flavonoid-rich foods), and specific flavonoid subclasses and individual compounds. Main outcome measuresIncident depression (ascertained via inpatient and death register data) and alterations in cortical volumes and white matter integrity. ResultsDuring a median follow-up of 10.5 years, 2965 incident depression cases were identified. After adjusting for demographic, major lifestyle, and clinical factors, total flavonoid intake and the Flavodiet Score were both linearly associated with a reduced depression risk, with hazard ratios (HRs) and 95% confidence intervals (CIs) for quintile 5 vs. corresponding quintile 1 being 0.84 (0.75 to 0.95) and 0.77 (0.68 to 0.87). No significant interactions were found between total flavonoid intake and potential effect modifiers, such as demographic, lifestyle, and chronic disease factors. Subclass analysis showed that intakes of anthocyanins and flavanones both were linearly associated with a lower depression risk, with HRs (95% CIs) for quintile 5 vs. corresponding quintile being 0.63 (0.55 to 0.71) and 0.84 (0.95 to 0.94). Beyond specific compounds from flavanones and anthocyanins subclasses, proanthocyanidin dimers, catechin, and epicatechin as a cluster derived by the k-means method were also associated with a reduced depression risk. Path diagrams suggested a sequential chain of mediation via sarcopenia biomarkers (creatinine to cystatin C ratio and muscle quality index) and subsequent diabetes status in the association of total flavonoid intake with depression risk. Regarding brain structural changes, higher flavonoid intake was associated with less atrophy in specific cortical regions (left Caudal anterior cingulate, left Supramarginal, right Caudal middle frontal, and right Precentral) and a smaller increase in mean diffusivity in the left Anterior Thalamic Radiation tract (all P < 0.05). ConclusionsOur findings suggest that dietary flavonoids, notably anthocyanins and flavanones, may contribute favourably to depression risk reduction and may offer protection against adverse brain structural changes. Further mechanistic studies and randomized controlled trials are warranted to validate these effects and translate them into public health policy. Research summaryO_ST_ABSWHAT IS ALREADY KNOWN ON THIS TOPICC_ST_ABSFlavonoids are known to exert beneficial effects through key mechanisms highly relevant to depression pathophysiology, as shown in preliminary research. Population-based research between flavonoid intake and risk of depression is generally scarce and largely supports the positive impact of total flavonoid intake on depression risk. Nevertheless, the association of flavonoid with depression risk is inconsistent in the subclass level and exploration analysis in the individual compound level is conducted ignoring the correlations in between. WHAT THIS STUDY ADDSPartial monomers (i.e., catechin and epicatechin) and dimers of proanthocyanidins as a cluster from the k-means method were associated with a reduced depression risk, alongside specific compounds from flavanones and anthocyanins. The effect of total flavonoid intake on depression risk was mediated by a sequential chain of intermediate factors, specifically sarcopenia biomarkers and subsequent diabetes status. The total flavonoid intake and sum of servings for top 10 flavonoids-rich food items presented protection against depression-related brain structural changes.