Look at the mood shelf of any supplement store and you will find fish oil with a twist: "high EPA," "EPA-dominant," "3:1 EPA to DHA." The pitch is that the two main omega-3 fatty acids are not equal, and that EPA is the one that matters for mood. That idea did not come from marketing. It came from meta-analyses that sorted depression trials by their formulas and found the positive ones clustered on the EPA-heavy side.
But the same research base also contains large trials that found nothing, a prevention trial with more cases of depression on fish oil than on placebo, and a guideline group that recently moved omega-3 down its list. Here is what the trials tested, in whom, and what it means if you are reading the EPA number on a fish oil label.
The short version
- The EPA idea comes from comparing trials. Meta-analyses found that trials using formulas with about 60% or more EPA tended to show benefit in people with diagnosed depression, while DHA-heavy formulas did not. The few trials that tested EPA against DHA head to head disagree: the largest, in 196 adults taking omega-3 on its own, found neither beat placebo, while a smaller add-on trial found EPA beat both DHA and placebo.
- The overall effect is small. The 2021 Cochrane review found a small improvement in symptoms, about 2.5 points on a standard depression scale, below the 3 points often treated as clinically meaningful, on very low certainty evidence.
- It does not prevent depression. In VITAL-DEP, 18,353 older adults took 1 g of fish oil or placebo for about five years: there were slightly more cases of depression on fish oil.
- Several large EPA-heavy trials missed their main goal, including trials in adults and in young people.
- Guidelines have cooled. Canada's CANMAT group downgraded omega-3 to a third-line option for depression in 2023. An expert panel that favors it suggests 1 to 2 g of EPA a day, with a clinical diagnosis and monitoring.
- Depression needs professional care. Omega-3 is not a replacement for therapy or medication. Do not stop or change an antidepressant on your own, and if you have bipolar disorder, talk with your prescriber first. If you are thinking about suicide or harming yourself, call or text 988, the Suicide and Crisis Lifeline, in the US.
- There are tradeoffs. In heart trials, omega-3 supplements raised the risk of atrial fibrillation, more so above 1 g a day, and fish oil may add to bleeding risk with blood thinners.
Why EPA keeps coming up
Fish oil contains two main long-chain omega-3s. DHA (docosahexaenoic acid) is a building block of brain cell membranes. EPA (eicosapentaenoic acid) is present in the brain in much smaller amounts, but it competes with the omega-6 fat arachidonic acid in the body's inflammation pathways, and the signaling molecules made from EPA tend to be less inflammatory. Both are also raw material for compounds that help resolve inflammation.
That matters because of a long-running idea in psychiatry: that inflammation contributes to depression in some people. The observational support is real but indirect. A 2010 meta-analysis of 24 studies found higher levels of two inflammatory markers, TNF-alpha and IL-6, in people with major depression. Another 2010 meta-analysis of 14 studies found lower EPA, DHA and total omega-3 levels in the blood of people with depression than in people without it. Neither can show which way the arrow points: lower omega-3 could be a cause, a consequence, or a marker of diet and lifestyle.
Put those together and you get the hypothesis: if depression has an inflammatory component, EPA, the more anti-inflammatory omega-3, should be the one that helps. It is a plausible idea, not a demonstrated one. The proposed mechanism comes mostly from cell, animal and blood-marker studies, not from trials showing that lowering inflammation with EPA relieves depression.
The evidence at a glance
| The claim | What was tested | Best evidence | Our read |
|---|---|---|---|
| Omega-3 eases symptoms in diagnosed depression | Placebo-controlled trials in adults with depression, often alongside antidepressants | Cochrane 2021: 33 trials, 1,848 people, a small effect (SMD -0.40), about 2.5 points on the Hamilton scale, very low certainty; remission and response did not differ | A small average effect that may not be clinically meaningful, and possibly biased toward a positive finding |
| EPA-dominant formulas work and DHA does not | Meta-analyses that grouped different trials by formula | Trials with 60% or more EPA showed an effect (ES 0.53), those below did not. Head-to-head trials disagree: the largest (196 adults, omega-3 on its own) found neither beat placebo; a smaller add-on trial (62 analyzed) favored EPA | A real pattern across trials, not a proven rule. Several large EPA-heavy trials missed their main goal |
| It works as an add-on to antidepressants | Omega-3, mostly EPA, added to ongoing antidepressant treatment | Small positive trials (20 to 70 people) and supportive meta-analyses, but a meta-analysis of sertraline add-on trials found no clear benefit | Possible modest effect in trials, low certainty. Something to raise with the prescriber, never a substitute |
| It prevents depression | Long-term trials in people without depression | VITAL-DEP, 18,353 adults over 5.3 years: HR 1.13 for depression on fish oil. A review of 31 trials in 41,470 people found little or no effect | Not supported |
| It helps in pregnancy and after birth | Trials in pregnant and postpartum women | A 2020 meta-analysis found a small overall effect, negligible in pregnancy and larger but not significant after birth; a 2,399-woman DHA trial found no significant difference | Not supported in pregnancy; unproven after birth. Perinatal depression needs clinical care |
| It helps teens and young adults | Trials in young people with depression | Two trials of 233 and 257 young people, both using EPA-leaning formulas with therapy, found no significant difference | Not supported by the largest trials |
| The right dose is 1 to 2 g of EPA, at more than 2:1 | Dose comparisons across and within trials | An expert panel recommends this; trials point in different directions, and a 2:1 trial in adolescents found no benefit | An expert opinion, not an established dose |
These summaries describe research in people with diagnosed depression or in prevention trials. They are not claims about what a supplement does, and omega-3 is not a treatment for depression. SMD and ES are standardized effect sizes; HR is a hazard ratio, where 1.13 means 13% higher risk. Depression needs care from a health professional. Do not stop or change an antidepressant on your own, and see the safety section on bipolar disorder, blood thinners and heart rhythm. In a crisis, call or text 988 in the US.
Where the 60% EPA rule comes from
Early trials of omega-3 for depression gave wildly different results. In 2009 and 2011, researchers looked for a reason and found one in the formulas.
The 2009 analysis. A meta-analysis of 28 trials found an overall benefit (SMD -0.29), but when the trials were split by formula, those using mostly EPA showed an effect (SMD -0.45) while pure DHA and DHA-heavy formulas showed none. The same analysis also found that benefit shrank as study quality, size and length went up, and it detected publication bias.
The 2011 analysis. A meta-analysis of 15 double-blind trials with 916 participants produced the number that ended up on labels. Supplements in which EPA made up 60% or more of the EPA plus DHA had an effect size of 0.53 (95% CI 0.28 to 0.73). Supplements below 60% had an effect of -0.03, which is no effect at all. An exploratory model suggested benefit depended on how much EPA exceeded DHA, in a range of 200 to 2,200 mg a day.
Later analyses kept finding the pattern. A 2016 analysis of 35 trials found EPA-predominant formulas helped in diagnosed depression and DHA-predominant ones did not. A 2019 analysis of 26 trials with 2,160 people found benefit with EPA-pure or EPA-major formulas at doses of 1 g of EPA a day or less, and none with DHA-pure or DHA-major ones. A 2023 analysis found that formulas with at least 60% EPA helped (SMD -0.36), and that doses of 1 to under 2 g of EPA did, while 2 g or more did not reach significance.
Why this is a pattern, not a rule. Every one of these splits compares different trials, run in different people, by different teams. The 60% line is an analysts' cutoff, not something a trial tested. The direct test is to give one group EPA, another DHA and a third placebo, in the same trial, and the few trials that did so disagree. The largest, in 2015, gave 196 adults with major depression about 1 g of EPA-enriched oil, 1 g of DHA-enriched oil, or placebo, taken on their own rather than with an antidepressant, for 8 weeks. Neither omega-3 separated from placebo, and response rates were 40 to 50% in all three groups. A smaller 2013 trial, in 81 adults with mild to moderate depression (62 analyzed) taking 1 g a day added to their medication for 12 weeks, found the opposite: EPA beat both DHA and placebo. A 2014 trial in 162 people starting interferon therapy for hepatitis C, a treatment that often triggers depression, found new depression in 10% on EPA, 28% on DHA and 30% on placebo.
A follow-up analysis of that trial found that people with signs of high inflammation improved more on EPA than on placebo or DHA. The authors called it proof of concept. A 2022 trial in 61 unmedicated, overweight adults with depression and high inflammation tested 1, 2 and 4 g of EPA: response was 64% at 4 g against 40% on placebo, but none of the comparisons was statistically significant. The inflammation idea is interesting and unconfirmed.
What trials in diagnosed depression found
The pooled picture. The 2021 Cochrane review is the most careful single summary. Across 33 trials with 1,848 adults, omega-3 reduced depressive symptoms slightly more than placebo (SMD -0.40, 95% CI -0.64 to -0.16). That works out to about 2.5 points on the Hamilton depression scale, below the 3-point difference often used as the smallest change patients notice. Remission and response rates did not differ significantly, and the reviewers rated the evidence very low certainty, with signs the estimate may be biased toward benefit. The confidence interval, 1.0 to 4.0 points, does include a clinically important effect. An earlier meta-analysis that adjusted for publication bias found the effect fell to about zero.
Other meta-analyses are more positive. A 2016 analysis limited to interview-confirmed major depression found a modest effect after correcting for publication bias (SMD 0.40, where positive here means improvement), larger with higher EPA doses and when more participants were on antidepressants, though the effect also shrank in more recent studies. A 2024 dose-response analysis of 67 trials found a U-shaped pattern in people with depression, with the largest improvement around 1.5 g a day of total omega-3, and more remissions on omega-3 (low certainty evidence). A 2025 meta-analysis of 36 studies found a small symptom effect but no difference in response or remission.
The large EPA-heavy trial. The biggest test of an EPA-dominant formula enrolled 432 adults with a major depressive episode and gave 1,050 mg EPA plus 150 mg DHA (about 88% EPA) or placebo for 8 weeks. It missed its primary outcome (P = 0.088). A benefit appeared only in the subgroup without an anxiety disorder.
Add-on trials. Most of the encouraging trials added EPA to an antidepressant. In a 2002 dose-finding trial in 70 people still depressed despite medication, 1 g of ethyl-EPA beat placebo, but 2 g showed little evidence of benefit and 4 g only trends. In a 2008 trial of 60 outpatients, EPA alone, fluoxetine alone and the two combined were compared; the combination did best, but there was no placebo group. A 2021 meta-analysis of four trials adding omega-3 to sertraline found no clear benefit, on very low certainty evidence, while a 2016 meta-analysis of add-on nutraceuticals reported a moderate to strong effect for adjunctive omega-3, with signs of publication bias. A 2026 network meta-analysis found effects for EPA so large that its own authors flagged them as unrealistic, with low to very low confidence.
The fair summary: in trials in people with diagnosed depression, omega-3, especially EPA-heavy formulas, may have produced a small improvement, more likely as an add-on than alone, with low to very low certainty. Large trials with strong designs have mostly not confirmed it.
Prevention: the large trials
If omega-3 corrected a cause of depression, taking it should lower the chance of developing it. In the general population, the large trials say it does not.
VITAL-DEP. In this trial, 18,353 US adults aged 50 and older without current depression took 1 g a day of fish oil (465 mg EPA plus 375 mg DHA) or placebo for a median of 5.3 years. There were 651 cases of depression or clinically relevant symptoms on fish oil and 583 on placebo, a hazard ratio of 1.13 (95% CI 1.01 to 1.26). Mood scores did not differ. The authors concluded the results do not support omega-3 supplements to prevent depression. Supporters of the EPA idea point out that this formula was about 55% EPA, just under the 60% line, which is fair; it is also the composition of a standard prescription omega-3 (Lovaza).
Other prevention evidence. A European trial in 1,025 overweight adults with mild symptoms found that a daily supplement containing 1,412 mg of omega-3 plus other nutrients did not prevent major depression over a year. A 2021 review of 31 trials lasting at least six months, in 41,470 people, concluded omega-3 probably has little or no effect on preventing depression (risk ratio 1.01). A 2016 analysis also found EPA did not prevent depressive symptoms in people without a diagnosis.
Two small trials in people already recovered from depression (39 and 60 people) found fewer relapses on omega-3, and the interferon trial above found less depression on EPA. Those are narrower questions than preventing a first episode in the general population.
Pregnancy, young people and bipolar disorder
Pregnancy and after birth. A 2020 meta-analysis of 18 trials with 4,052 participants found an effect in pregnancy that was negligible, and its authors advised against prescribing omega-3 for depression during pregnancy. The same analysis found a larger, though not statistically significant, effect after birth, and called omega-3 a possible add-on for postpartum depression that needs more study. A trial in 2,399 pregnant women taking 800 mg of DHA a day found no significant difference in postpartum depressive symptoms. Canada's CANMAT 2024 perinatal guideline lists omega-3 under insufficient or mixed evidence. Depression during or after pregnancy is common and treatable, and belongs with your obstetric or mental health team.
Young people. A 2024 Cochrane review of omega-3 in children and adolescents with depression found very low certainty evidence and no clear benefit. Two larger trials since then point the same way. In one, 233 young people aged 15 to 25 received 1.4 g of omega-3 (60% EPA) with cognitive behavioral case management: symptom change was nearly identical to placebo. In the other, 257 adolescents with moderate to severe depression took 1 g EPA plus 0.5 g DHA, exactly 2:1, with psychotherapy for 36 weeks: there was no significant difference, and response and remission were numerically lower on omega-3.
Bipolar disorder. A small 1999 trial found longer periods of stability with omega-3, and a 2006 trial in 75 people found ethyl-EPA improved bipolar depression scores. But a 2006 trial of 6 g a day found no differences, and a 2012 trial in 45 people found no benefit. A 2022 clinician guideline gave omega-3 weak support in bipolar depression. There are also isolated case reports of hypomania associated with omega-3. Anyone with bipolar disorder should involve their prescriber before adding it.
What guidelines say
- CANMAT (Canada), 2023. In 2016 this group listed omega-3 as a second-line option for milder depression. In its 2023 update, citing the Cochrane review, it wrote that omega-3 fatty acids are "now downgraded to a third-line treatment."
- International Society for Nutritional Psychiatry Research, 2019. An expert consensus panel considered pure EPA or an EPA to DHA ratio "higher than 2" effective, at 1 to 2 g of net EPA a day, and advised clinicians to make a proper diagnosis and monitor side effects. In our reading, several of its positions, such as use in pregnancy and youth, have not been supported by trials published since.
- WFSBP and CANMAT taskforce, 2022. A clinician guideline on nutraceuticals rated adjunctive omega-3 as supported in depression, and omega-3 on its own as not supported.
- NIH (NCCIH). The National Center for Complementary and Integrative Health says it is uncertain whether omega-3 supplements help depression, and notes that other reviews suggest that, if there is an effect, EPA may be more helpful than DHA and omega-3 may work best alongside antidepressant medication rather than in place of it.
The common thread: where guidelines give omega-3 a place, it is a low-priority or add-on one, third-line in CANMAT 2023 and supported as an add-on but not on its own by the WFSBP and CANMAT taskforce, chosen with a clinician. US law does not allow a supplement to claim to treat depression.
How much EPA is in your fish oil
If you read the back of a fish oil bottle, the numbers that matter are the EPA and DHA per serving, not the total "fish oil" on the front. According to the NIH Office of Dietary Supplements, a typical 1,000 mg fish oil softgel contains about 180 mg of EPA and 120 mg of DHA, or 300 mg of omega-3 in total.
| Product type | EPA and DHA | EPA share | Softgels to match the 1 g of EPA used in some trials |
|---|---|---|---|
| Typical fish oil (1,000 mg softgel) | About 180 mg EPA, 120 mg DHA | 60% (a 1.5 to 1 ratio) | About 6 softgels; doses this high are for a clinician to direct |
| EPA-dominant fish oil | Varies; EPA listed well above DHA | Higher than 60%; check the label | Fewer softgels; check the label |
| Prescription omega-3 ethyl esters (Lovaza) | About 465 mg EPA, 375 mg DHA per 1 g capsule | About 55% | A prescription drug for very high triglycerides, not for mood |
| Prescription icosapent ethyl (Vascepa) | Purified EPA | 100% | A prescription drug for triglycerides and heart risk, not for mood |
| Algal oil | Often mostly DHA; some products add EPA | Varies | Check the label; many vegan oils are DHA-heavy |
First, an ordinary fish oil already sits right at the 60% line that the meta-analyses drew, but its ratio is about 1.5 to 1, below the "more than 2 to 1" an expert panel suggested. Second, the 1 to 2 g of EPA that panel recommended takes about 6 to 11 ordinary softgels, which is why EPA-concentrated products exist. Our arithmetic here is based on the NIH figures; always check the actual EPA number on your label.
Fish oil comes as natural triglycerides, re-esterified triglycerides, ethyl esters and other forms. NIH notes that triglyceride forms are absorbed somewhat better than ethyl esters, but that all forms raise blood EPA and DHA. Freshness and purity testing matter more for most buyers. Our fish oil buyer's guide ranks products on EPA and DHA per serving, form and third-party testing, and our krill oil guide covers the phospholipid alternative; check its EPA per serving too.
Safety and who should be careful
- Depression needs professional care. Omega-3 is not a treatment for depression and should not replace therapy or medication, or delay seeing a clinician. If you are having thoughts of suicide or self-harm, call or text 988 (the Suicide and Crisis Lifeline) in the US, or your local emergency number.
- Do not stop or change an antidepressant on your own. If you want to add omega-3, tell your prescriber.
- Bipolar disorder. There are case reports of hypomania with omega-3. Talk with your prescriber first.
- Blood thinners. Fish oil might raise the INR in people taking warfarin, though most research at 3 to 6 g a day found no significant effect, and the label for a 4 g prescription EPA reports more bleeding with aspirin, clopidogrel or warfarin. Your clinician may want to monitor.
- Heart rhythm. In heart trials, omega-3 raised the risk of atrial fibrillation, more so above 1 g a day. If you have atrial fibrillation or heart disease, ask your doctor.
- Pregnancy. Talk with your obstetric team; trials found no meaningful mood benefit in pregnancy.
- Fish or shellfish allergy. Check the source oil and ask your clinician.
Heart rhythm, in numbers. A 2021 meta-analysis of seven heart outcome trials in 81,210 people, mostly older adults with heart disease or risk factors, found omega-3 supplements raised the risk of atrial fibrillation by 25% overall (HR 1.25), by 49% in trials using more than 1 g a day and by 12% in trials using 1 g or less, although the largest single trial at about 1 g a day, VITAL Rhythm in 25,119 adults, found no significant difference (3.7% vs 3.4%). In the REDUCE-IT trial of 4 g a day of purified EPA, hospitalization for atrial fibrillation or flutter was 3.1% against 2.1% on placebo. These were not depression trials, and the absolute risks were small, but the EPA doses discussed for mood overlap with them. Our fish oil and atrial fibrillation review covers this in detail.
Upper amounts. In 2019 FDA concluded that dietary supplements providing no more than 5 g a day of EPA and DHA are safe when used as labeled, based on bleeding risk. The European Food Safety Authority found no safety concerns for adults at supplemental intakes up to 5 g a day of EPA and DHA combined, or up to 1.8 g a day of EPA alone. The 1 to 2 g of EPA discussed for mood reaches that EPA-alone figure at the upper end and exceeds it at 2 g.
Everyday side effects. NIH lists an unpleasant taste, bad breath, heartburn, nausea, diarrhea and headache, usually mild. Purified fish oils have not been found to contain meaningful methylmercury, according to NIH.
For the wider evidence on supplements and mood, see our low mood supplement hub, our reviews of saffron and vitamin D, and our guide to supplement and drug interactions.
Frequently asked questions
Does omega-3 help with depression?
In trials in people with diagnosed depression, the effect was small at most, and the evidence is weak. The 2021 Cochrane review of 33 trials found a small improvement in symptoms, about 2.5 points on a standard depression scale, below the 3 points usually treated as clinically meaningful, with no significant difference in remission or response, and rated the evidence very low certainty. Canada's CANMAT guideline downgraded omega-3 to a third-line option in 2023. Omega-3 is not a treatment for depression and should not replace therapy or medication. If you are in crisis, call or text 988, the Suicide and Crisis Lifeline, in the US.
Is EPA or DHA better for mood?
Meta-analyses that grouped trials by formula found that trials using mostly EPA, about 60% or more of the total, tended to show benefit in diagnosed depression, while DHA-heavy formulas did not. That pattern comes from comparing different trials. The largest trial that tested EPA against DHA directly, in 196 adults taking omega-3 on its own, found neither beat placebo, while a smaller add-on trial found EPA beat both DHA and placebo; the largest EPA-dominant trial, in 432 adults, missed its main goal.
How much EPA is used in mood research?
An expert panel, the International Society for Nutritional Psychiatry Research, suggested 1 to 2 g of EPA a day, as pure EPA or with EPA more than twice the DHA, after a clinical diagnosis and with monitoring. Trials do not pin down a dose: some found benefit at 1 g or less and not at higher doses, others the reverse. A typical 1,000 mg fish oil softgel contains about 180 mg of EPA. Doses this high should be discussed with your clinician, especially if you take blood thinners or have a heart rhythm problem.
Can omega-3 prevent depression?
The large trials say no. In VITAL-DEP, 18,353 adults aged 50 and older took 1 g of fish oil a day or placebo for about five years, and there were slightly more cases of depression on fish oil (hazard ratio 1.13). A review of 31 long-term trials in more than 41,000 people concluded omega-3 probably has little or no effect on preventing depression.
Can I take fish oil with antidepressants?
Many of the trials that found a benefit added omega-3 to an antidepressant, and adverse events were similar to placebo in the Cochrane review, though the estimate was imprecise and could not rule out a modest increase. Still, tell your prescriber before adding it, especially if you also take a blood thinner or have bipolar disorder, and never stop or change an antidepressant on your own. NIH notes that some reviews suggest omega-3 may work best in addition to medication rather than in place of it.
How long does omega-3 take to work for mood?
Most treatment trials lasted 8 to 12 weeks, so any effect, if it occurs, would be expected over weeks rather than days. Many of those trials found no significant difference from placebo, so it should not be counted on. If low mood lasts more than two weeks or affects daily life, see a health professional rather than waiting on a supplement.
Is omega-3 helpful for depression during pregnancy?
The evidence does not support it. A 2020 meta-analysis found a negligible effect in pregnancy and its authors advised against prescribing omega-3 for depression then (they saw a larger but not significant effect after birth), and a trial in 2,399 pregnant women taking DHA found no significant difference in postpartum symptoms. Depression during or after pregnancy is common and treatable: talk with your obstetric or mental health team.
Does the type of fish oil matter?
For mood research, the main difference is how much EPA a product contains. Check the EPA number per serving on the label rather than the total fish oil. Triglyceride forms are absorbed somewhat better than ethyl esters, but NIH notes that all forms raise blood EPA and DHA. Third-party testing for purity and freshness matters for any fish oil.
The bottom line
EPA-heavy formulas keep coming up for mood because, when researchers sorted depression trials by formula, the positive results clustered on the EPA side. That pattern is real, and so is the plausible link to inflammation. But it comes from comparing different trials, the largest direct EPA against DHA trial found neither beat placebo, and the overall effect in diagnosed depression is small and uncertain. Large trials of EPA-leaning formulas in adults and young people mostly missed their goals, prevention trials found nothing, and VITAL-DEP found slightly more depression on fish oil.
So the most the trials support is a possible small effect, on low certainty evidence, when omega-3 was added to treatment in people with diagnosed depression. Omega-3 is not a treatment for depression, and the large trials do not show it prevents depression. If you want to raise it with the clinician treating you, mention any blood thinner, heart rhythm problem or bipolar disorder, and do not stop or change an antidepressant on your own. If you take fish oil for other reasons, reading the EPA number on the label is still worth it: an ordinary softgel has about 180 mg. If your mood has been low for weeks, the most effective step is to talk with a health professional, and if you are in crisis, call or text 988, the Suicide and Crisis Lifeline, in the US.