Greater with the extension of duration when duration was shorter than 12 months, is consistent with the study by Katan et al. and is plausible. A significant negative linear relationship between age and the effect size of marine-derived n-3 PUFAs supplementation on fasting blood level of IL-6 was observed in subjects with chronic non-autoimmune disease, indicating that the lowering effect was stronger in older subjects. Subgroup analysis also found a consistent result. One previous study by Rodrigo et al. also found that older subjects showed a better antioxidant response to n-3 PUFAs supplementation. As suggested in the second paragraph of this discussion, one Alisol-A-24-acetate possible mechanism for the anti-inflammatory effect of n-3 PUFAs might be attributed to its antioxidant effect. Therefore, our result concerning the effect of age on the association between marine-derived n-3 PUFAs supplementation and IL-6 was consistent with the study by Rodrigo et al.. One study by Rees et al. found that older subjects incorporated long-chain n-3 PUFAs into plasma and mononuclear cells more readily than younger ones. Similar results were also found in the study by Meydani et al.. These results were consistent with one cross-sectional study which found that long-chain n-3 PUFAs in adipose tissue increased with age. This may be one reason why the lowering effect of marine-derived n-3 PUFAs supplementation on IL-6 was more effective in older subjects. In addition, one study by Cartier et al. found that CRP, IL-6 and TNF-a level increased with age in men. Our present study showed that higher baseline level was associated with a greater lowering effect. This may also help explain the difference in effect size caused by age. We also observed a significant negative linear relationship between daily dose of EPA and the effect size on fasting blood level of CRP in healthy subjects, indicating that higher dose of EPA might lead to a greater lowering effect on CRP. But similar relationship was not observed for daily dose of total n-3 PUFAs or DHA. This provided some evidence that EPA might be at least more effective than DHA on lowering fasting blood level CRP. We also observed a significant positive linear relationship between daily dose of DHA and the effect size on fasting blood level of IL-6 in healthy subjects. This result was consistent with the negative relationship between daily dose of EPA and the effect size on fasting blood level of CRP if we take into consideration that marine-derived n-3 PUFA supplementation is mainly comprised of EPA and DHA, and that higher dose of DHA might lead to a lower 10-Gingerol proportion of EPA. One previous studies in mice showed that the lowering effect on cytokines production by thioglycollate-induced macrophages were only significant in the group fed by diet with the highest EPA/DHA ratio. These results came to a consistence with those of our present metaanalysis. The mechanism behind this phenomenon may involve the competing role of n-3 PUFAs with arachidonic acid in eicosanoids metabolism which has been described in detail in the fifth paragraph of discussion. Briefly, EPA reduces the production of arachidonic acid-derived eicosanoids by inhibiting the oxidation of arachidonic acid by COX, acting as competing substrates of COX and 5-LOX, and incorporating into cell membrane phospholipids at the expense of arachidonic acid, and thus reduces the production of cytokines and other inflammatory factors. Although previous studies have shown that DHA can be retroconverted to EPA, the retroconversion ratios were low. In addition, one previous study in mice showed that diet high in EPA rather than diet high in DHA significantly increased IL-10 production by macrophages and lymphocytes.