Turn the bottle around and read the per-serving EPA and DHA numbers. Suppose a serving lists 650 mg EPA and 350 mg DHA. Divide 650 by the combined 1,000 mg and you get 65% EPA, which clears the 60% line the mood trials use to separate EPA-enriched formulas from the rest.
Turn the bottle around and read the per-serving EPA and DHA numbers. Suppose a serving lists 650 mg EPA and 350 mg DHA. Divide 650 by the combined 1,000 mg and you get 65% EPA, which clears the 60% line the mood trials use to separate EPA-enriched formulas from the rest.
That line comes from a 2023 meta-analysis of 10 RCTs covering 1,426 adults, which found that depression severity dropped significantly only with EPA-enriched interventions, defined as EPA making up 60% or more of total EPA+DHA (SMD -0.36; 95% CI -0.68 to -0.05; p=0.02) (PMID 37028202). Dose also mattered. EPA at 1 g/day or more but under 2 g/day was associated with reduced severity (SMD -0.43; 95% CI -0.79 to -0.07).
So the next check is how many servings you need to reach 1 g of EPA. For the product above the answer is two servings, which gives you 1.3 g EPA and keeps you inside the 1 to 2 g band. A standard 1 g fish oil softgel with 180 mg EPA and 120 mg DHA comes out at 60% EPA, but you would need six capsules to reach about 1 g of EPA. That is the main reason generic "take 1 g of fish oil" advice falls short when mood is the goal.
The dose finding is less settled than that band suggests. In the same 2023 analysis, EPA at 2 g/day or more showed no significant effect (SMD -0.20; 95% CI -0.48 to 0.07; p=0.14). A 2016 meta-analysis of 13 RCTs in 1,233 adults with clinically diagnosed major depression found an overall benefit after adjusting for publication bias (SMD 0.398; 95% CI 0.114 to 0.682), and its meta-regression linked higher EPA dose to better outcomes (P=0.009) (PMID 26978738). The two analyses disagree on whether more EPA keeps helping. Both also carry warning signs. Heterogeneity in the 2023 analysis ran at an I² of 86 to 88%, and its funnel plot asymmetry suggests publication bias. The 2016 regression found that earlier publication year predicted larger effects, which hints that newer trials are showing smaller benefits. Anxiety could not be meta-analyzed because only one trial reported a reduction, at 2.1 g/day EPA making up 85.6% of EPA+DHA (PMID 37028202). Every one of these trials studied people with depression. None of them supports a mood claim for a healthy person adding EPA on top of a stable baseline.
For cognition and joints, the supplied evidence does not give you a ratio. In 90 Australian children aged 7 to 12 with ADHD symptoms, four months of EPA-rich oil, DHA-rich oil, or safflower oil produced no significant group differences on the primary outcomes (PMID 22541055). In a pilot of 40 people with gout, 6.2 g/day of fish oil for 24 weeks did not change serum urate, weight, or BMI compared with control, and four participants on fish oil reported GI side effects (PMID 35672866). [Evidence for DHA-dominant use cases in adults (e.g., cognition, pregnancy); the supplied DHA evidence is pediatric ADHD correlations only]. Until that evidence exists, a DHA-heavy product chosen "for the brain" is a guess you are making, and you should treat it as one.
Those null trials still contain something useful. In the ADHD study, children whose erythrocyte DHA rose more showed better word reading (r=0.394) and lower oppositional behavior ratings (r=0.392), and the correlations were stronger in a subgroup of 17 children with learning difficulties (PMID 22541055). In the gout pilot, red cell omega-3 levels, measured monthly by blood spot, correlated with fewer flares between weeks 12 and 24 (total r=-0.75, EPA r=-0.75, DHA r=-0.76) (PMID 35672866). In both trials the randomized arms did not separate, yet measured tissue levels tracked the outcome. These findings are correlational. They come from small studies, one of them open-label, and they do not show that fish oil reduces gout flares or improves reading. What they do show is that the dose printed on the label and the amount that ends up in a person's membranes can differ enough to matter, and the label dose was a worse predictor of outcome than the blood measurement.
The omega-3 index is the number that captures this: the percentage of EPA plus DHA in red blood cell membranes (PMID 21063431). In young Canadian adults who started with a low index of about 3.5%, DHA-only supplementation raised it by 4.90 percentage points, while EPA-only supplementation at 3.0 g/day raised it by 2.99 (PMID 37730162). DHA appears to be the more efficient way to move that number, even though the mood trials favor EPA-dominant formulas. The ratio with the best mood evidence and the ratio that moves the index furthest are not the same ratio, which is one more reason to measure instead of assuming.
Form is another reason the same label dose can produce different results. In a 6-month double-blind RCT of 150 volunteers, everyone took 1.01 g EPA plus 0.67 g DHA per day. Participants on re-esterified triglycerides saw their omega-3 index rise 186% at 3 months, compared with 161% for those on ethyl esters. At 6 months the figures were 197% and 171% (PMID 21063431). Both forms raised the index significantly from baseline; the triglyceride form simply raised it further. In a crossover trial of 24 healthy volunteers, krill oil and fish oil each delivering 600 mg of n-3 PUFA for 4 weeks were compared. Krill raised RBC EPA+DHA more than fish oil (p=0.0143) and also lowered the n-6:n-3 ratio (PMID 24304605). Total and LDL cholesterol also rose after krill, while triglycerides and HDL did not change with any treatment. The trial was small, short, and used a low dose, so the lipid finding is far from conclusive. If you switch to krill, check your next lipid panel for that change.
Here is a retest protocol that can separate those variables:
- Get a baseline omega-3 index before you change anything. Without a baseline, the retest number has nothing to be compared against.
- Change one variable at a time. That variable can be ratio, form (for example, ethyl ester to triglyceride), or dose. If you switch to a 70% EPA triglyceride product at a higher dose all at once, you will not know which change did the work.
- Retest at about 3 months. The rTG vs EE trial measured at 3 and 6 months and saw most of the rise by the first timepoint (PMID 21063431). [A citation for red blood cell lifespan and turnover to justify a ~3-4 month retest window; P4's 3- and 6-month timepoints are the only supplied evidence].
- Log daily intake and a weekly check-in on whatever you are trying to change, such as mood, joint stiffness, or focus. Without that log, the index result has no outcome to be compared against.
When you read the result, a flat index means you should suspect dose, form, or missed capsules before you conclude anything about the ratio. A rising index paired with check-ins that did not move is a different problem, and it is worth taking seriously.
[Omega-3 supplement guide link: https://microstracker.com/learn/supplements/
Here is a question to write down before your retest. If your index climbs two or three points and your weekly check-ins stay flat, will you conclude that the dose failed, or that your hypothesis about what omega-3s would do for you was wrong? Decide which result would change your protocol before the number arrives, then track the stack and the retest in Micros Tracker.