(P23-023-20) Dose-Dependent Effects of EPA Supplementation on Plasma Specialized Pro-Resolving Mediators in Major Depressive Disorder Patients with Chronic Inflammation

Objectives: The immunomodulatory effects of eicosapentaenoic acid (EPA) support it as an effective treatment for major depressive disorder (MDD) associated with chronic inflammation. Increased production of specialized pro-resolving mediators (SPM) during EPA supplementation may help reduce inflammation. We compared the effects of different EPA doses on plasma fatty acids and SPM in MDD patients.

Methods: MDD patients with baseline inflammation (N=61, IDS-C score >25; BMI >25 kg/m2; serum CRP >3 mg/L) were enrolled in a 2-site, 12-wk randomized trial comparing EPA 1, 2 and 4 g/d to placebo (soybean oil). Plasma fatty acids (mol%) and SPM (pg/mL) were measured at baseline and 12 wks (n=8 [10], 12 [11], 11 and 10 in the placebo, EPA 1, 2 and 4 g/d arms, respectively). Dose effects were tested using the Kruskal-Wallis followed by pairwise Wilcoxon rank-sum tests with Benjamini-Hochberg correction.

Results: In the placebo arm, plasma EPA and EPA derivatives did not change at 12 wks, relative to baseline. EPA supplementation significantly increased plasma EPA dose-dependently (3-, 3- and 9-fold in the 1, 2 and 4 g/d arms). Similar significant dose-dependent increases in EPA derivatives were observed in the EPA arms, including 18-hydroxy-EPA (18-HEPE, 8-, 7- and 14-fold) and 15-HEPE (4-, 6- and 10-fold). The increase in 18-HEPE, a precursor of EPA-series resolvins (RvE), by 4 g/d EPA was significantly greater than with lower doses. While RvE1 remained undetected in all treatment arms, RvE2 was detected in some subjects at baseline and 12 wks with a dose-dependent median increase (+3, +6 and +8 pg/mL). RvE3, undetectable at baseline, became detectable in more subjects dose-dependently (n=1, 2, 3 and 6). The reduction in plasma arachidonic acid (AA) by EPA supplementation significantly differed from placebo only at 2 and 4 g/d but not between the two doses. AA derivatives were not affected, except for the pro-resolving lipoxin B4, which was rarely detected at baseline, but became detectable in most subjects after EPA supplementation (median: 46, 59 and 94 pg/mL).

Conclusions: Our results show a robust dose-effect of EPA in increasing plasma EPA and EPA-derived 18-HEPE, which may result in improved conversion to RvE2 and RvE3. Its associations with reduced inflammation and improved IDS-C scores will be further analyzed.

Funding Sources: NCCIH