Just published: Dose-response association between the intake of various subtypes of polyunsaturated fatty acids and depression 2025
Chinese paper but George Institute for Global Health, so I think it’s good.
The paper concludes:
Post-propensity score matching, multivariate logistic regression analysis revealed an inverse correlation between the intake of AA, ALA, EPA, and DHA and the risk of depression. Restricted cubic spline analysis demonstrated a linear inverse relationship between AA, ALA and EPA intake and depression. Conversely, DHA intake exhibited a nonlinear relationship with depression.
Different types of PUFAs appear to influence depression risks differently in adults, with increased intake, AA, ALA, EPA and DHA providing a protective effect against depression.
However, I bought the full paper I think it’s a bit more complex for DHA:
Daily intake of n-3 and n-6 PUFAs was assessed using two 24-h dietary recall interviews, with average intake values used for analysis.
Participants were categorized into depression (cases) and non-depression (controls) groups based on their PHQ-9 scores.
Model 1 adjusted for age, sex, and race, while Model 2 further adjusted for marital status, education, BMI, Poverty Income Ratio (PIR), alcohol consumption, smoking status, physical activity, diabetes, hypertension, kidney failure, heart failure, CHD, stroke, liver disease, and cancer.
PUFAs were initially analyzed as continuous variables and later categorized into quartiles for comparison of interquartile increases. A four-node restricted cubic spline (RCS) was used in Model 2 to assess the dose-response relationship between PUFAs intake and depression risk, with nodes at the 5th, 35th, 65th, and 95th percentiles. Statistical analyses were performed using R version 3.6.1, with significance defined as a two-tailed P-value of <0.05.
Daily intakes (after PSM):
- EPA (mg/day), median [IQR]: 8.00 [3.50;17.0]
- DHA (mg/day), median [IQR]: 26.0 [7.00;62.0]
They don’t give the values for the quartile.
As a continuous variable (“per 1 mg/day increment”), after adjustments (Model 2) and propensity score matching, there’s a tiny beneficial effect for both EPA and DHA:
If I interpret Fig. 2 and Fig. 3 correctly, for both EPA and DHA, the potential protective effect is stronger in young (<60), non-smoking, non-drinking, non-diabetic, non-hypertensive women.
For EPA, it seems that the more, the better (linear dose-response relationship), whereas for DHA, it’s non-linear and the confidence interval goes above 1 past some point, so we cannot exclude that high-dose DHA causes depression:
Our study addresses a gap in the literature by demonstrating that both EPA and DHA intake are negatively correlated with depressive symptoms. Specifically, DHA intake showed a nonlinear association with depression risk, whereas EPA intake exhibited a linear relationship. Despite both being n-3 PUFAs, EPA and DHA appear to have distinct impacts on depression. For example, some studies have shown that EPA-enriched supplements are more effective than DHA-enriched supplements in preventing and treating depression (Liao et al., 2019; Grosso et al., 2014; Su et al., 2014). Additionally, EPA was shown to be more effective than DHA in mitigating weight loss and depressive-like behavior in a mouse model (Peng et al., 2020). Moreover, a study has indicated that EPA is effective in improving depressive symptoms, whereas DHA does not demonstrate such efficacy (Hallahan et al., 2016). The results of our study indicate that consumption of DHA within a specific range may also be associated with a reduced risk of depression. These findings indicate a need for further research to clarify the comparative effectiveness of EPA and DHA in treating depression. Furthermore, the multivariate logistic regression analysis indicated a significant linear trend in the association between DHA intake and depression risk, suggesting that higher DHA intake is generally associated with a lower risk of depression. However, the RCS analysis revealed a significant nonlinearity, with a plateau observed at higher intake levels. This suggests that while higher DHA intake is beneficial, the protective effect may not continue to increase linearly beyond a certain threshold. Further research, particularly longitudinal studies, could help determine the optimal intake level for DHA’s protective effects. Our analysis also revealed a negative correlation between DPA intake and depression, but this association disappeared after PSM, potentially due to confounding factors. Further research is needed to explore the relationship between DPA intake and depression.
Based on the dosage-response charts, the optimal daily dose might be 500 mg EPA + 300 mg DHA. However, VITAL-DEP used a similar dose (465 mg EPA + 375 mg DHA) and found an increased risk of depression after 5 years in their RCT.
My own conclusion: I think this paper strengthens the case for EPA and against high-dose DHA supplementation (anything above 100 mg/day being high-dose, given that Q3 = 62 mg/day). For DHA, 62 mg/day = 434 mg/week. A single serving of salmon, anchovies, mackerel, sardines, or herring gives you more than that.