Research dossier
Clinical research on DHA
10 trials reviewed across 5 indications.
Strongest evidence
Pregnancy and fetal neurodevelopment
Mechanism
DHA accretes in fetal brain and retinal tissue most rapidly in the third trimester. Maternal diet and stores are the primary source; PEMT-driven endogenous DHA synthesis from ALA is too slow to meet fetal demand.
The pregnancy DHA story is real but more modest than supplement marketing implies. DOMInO (n=2,399, 800 mg/day) and ORIP (n=5,544, 800 mg/day) both came back null on their primary cognitive and preterm-birth endpoints. KUDOS (600 mg/day) showed modest gestational-length and birth-size improvements. The mechanism is unimpeachable; the RCT effect sizes in fish-replete populations are small.
Strongest case for supplementation is pregnant women with very low fish intake. ACOG and AAP both recommend ≥200 mg DHA/day during pregnancy and lactation, achievable through 1–2 servings of low-mercury fatty fish or an algal DHA supplement.
Trials cited
DOMInO — DHA in pregnancy
Null · RCT
Makrides et al., 2010, JAMAn=2399Largest pregnancy DHA RCT to date. 800 mg/day DHA from mid-pregnancy did not reduce maternal postnatal depression and did not improve infant Bayley cognitive scores at 18 months versus placebo. Later 4- and 7-year follow-ups also found no cognitive benefit. A secondary signal of reduced very-low-birthweight births and a possible language signal in girls.
Background Australian diet was relatively fish-replete. Effects of prenatal DHA may be larger in populations with very low baseline fish intake. The primary cognitive endpoint was null.
KUDOS / Kansas DHA Outcome Study
mixed · RCT
Carlson et al., 2013, American Journal of Clinical Nutritionn=350Industry-fundedAlgal DHA 600 mg/day modestly increased gestational length (~2.9 days) and infant birth weight, length, and head circumference at delivery. Cognitive endpoints in later follow-up were largely null, with isolated signals in attention measures.
Endpoints shifted across the publication series. Birth-outcome effects were the cleanest signal; cognitive effects were small and inconsistent.
ORIP — n-3 fatty acids to prevent preterm birth
Null · RCT
Makrides et al., 2019, NEJMn=55445,544 women across Australia. High-dose DHA-dominant n-3 supplementation did not reduce early preterm birth at the population level. A subgroup signal in women with low baseline n-3 status — but the headline endpoint was null. Tempered enthusiasm carried over from earlier smaller pregnancy n-3 trials.
Australian women generally have moderate fish intake. The 2018 Cochrane review (Middleton et al.) covering older pregnancy n-3 trials had suggested a preterm-birth benefit; ORIP, the largest single trial, did not replicate it cleanly.
Cognition in older adults
Mechanism
DHA constitutes roughly 97% of brain omega-3 fatty acids and is the dominant structural component of neuronal phospholipid membranes. Higher membrane DHA supports synaptic plasticity, neurotransmitter signaling, and BDNF expression in animal models.
The trial picture is mostly disappointing. MIDAS (900 mg DHA, 24 weeks) showed modest improvement on a specific learning task in adults with subjective cognitive decline — industry-funded and not strongly replicated. Quinn 2010 (2 g DHA in established Alzheimer's, 18 months) was null. OmegAD (DHA-dominant n-3 in mild-to-moderate AD) was null. Dementia prevention via DHA is not a settled claim — it's a failed-trial pattern with an open question about earlier intervention.
Reasonable for older adults with subjective decline who already have low fish intake. Do not expect DHA to slow established Alzheimer's. APOE4 non-carriers may respond better than carriers — hypothesis-generating, not confirmed.
MIDAS — DHA for age-related cognitive decline
positive · RCT
Yurko-Mauro et al., 2010, Alzheimer's & Dementian=485Industry-funded485 adults aged 55+ with subjective cognitive decline. 900 mg/day algal DHA over 24 weeks improved Paired Associate Learning errors by an effect roughly equivalent to having the learning ability of someone three years younger. Modest, but the cleanest positive trial in this population.
Industry-funded (Martek Biosciences, makers of life'sDHA). Subjective-decline population without confirmed cognitive impairment. Replication in independent labs has been limited.
DHA for mild-to-moderate Alzheimer's disease
Null · RCT
Quinn et al., 2010, JAMAn=402402 adults with mild-to-moderate Alzheimer's. 18 months of 2 g/day DHA did not slow cognitive or functional decline versus placebo. A possible exploratory signal in APOE4 non-carriers, but the primary endpoint was unambiguously null. Closes the door on DHA as a treatment once Alzheimer's is clinically established.
Earlier intervention — before clinical Alzheimer's onset — has not been ruled out, but is also not established by any positive trial.
OmegAD — n-3 in mild-to-moderate Alzheimer's
Null · RCT
Freund-Levi et al., 2006, Archives of Neurologyn=204204 adults with mild-to-moderate Alzheimer's. DHA-dominant omega-3 did not slow cognitive decline on MMSE or ADAS-cog over 6 months versus placebo. An exploratory signal in patients with very mild AD (MMSE >27) hinted at delayed decline, but the overall result was null.
Very-mild-AD subgroup signal generated the hypothesis that DHA needs to be deployed earlier in the disease course — a hypothesis no subsequent trial has cleanly confirmed.
Retinal structure, infant visual acuity, dry eye
Mechanism
DHA makes up roughly 50% of the fatty acids in rod-photoreceptor outer-segment membranes — the most DHA-rich tissue in the body. Membrane DHA supports rhodopsin packing density and photoreceptor membrane fluidity.
Two separate clinical questions, very different answers. Infant visual acuity: DHA-fortified formula vs unfortified shows real acuity benefit in early life (Birch). Adult vision: AREDS2 specifically tested DHA + EPA for AMD progression and found nothing. Dry eye: DREAM (n=535, 3 g/day) was null in moderate-to-severe disease; smaller positive trials (Bhargava) were in low-omega-3-baseline computer users. Don't extrapolate the infant case to adult AMD prevention or dry-eye treatment.
Infant DHA needs are real (formula fortification reflects this). Adult AMD prevention and dry-eye treatment are not supported by the strongest available trials.
AREDS2 — DHA + EPA in age-related macular degeneration
Null · RCT
AREDS2 Research Group, 2013, JAMAn=42034,203 adults at high risk for advanced AMD. Adding 1 g/day combined DHA + EPA to the AREDS antioxidant + zinc base did not reduce progression to advanced AMD. The trial that answered the long-running 'should I take fish oil for my eyes?' question with a clean no.
Background AREDS micronutrient regimen may have masked an effect, but the dedicated DHA + EPA arm specifically failed to add benefit. AREDS-type formulations no longer include DHA + EPA.
DREAM — n-3 fatty acids for dry-eye disease
Null · RCT
Asbell et al., 2018, NEJMn=535535 adults with moderate-to-severe dry eye. 3 g/day combined EPA + DHA did not improve OSDI symptom scores, conjunctival staining, tear break-up time, or Schirmer testing over 12 months versus olive-oil placebo. The trial that broke the small-trial enthusiasm for omega-3 in dry eye.
Both arms improved substantially, suggesting strong placebo or regression-to-the-mean effects. Olive-oil control has been debated as not truly inert. Smaller earlier trials in subjects with low baseline omega-3 intake were more positive — DREAM did not select for that population.
Omega-3 for computer-vision-syndrome dry eye
positive · RCT
Bhargava et al., 2013, Cornean=478478 symptomatic computer users in India. A modest 300 mg DHA + 360 mg EPA per day improved subjective dry-eye symptoms, raised tear break-up time, and increased Schirmer values over 3 months versus placebo. Population had a low baseline-fish-intake profile that DREAM did not.
Single-center, low-baseline-omega-3 population, milder dry eye than DREAM enrolled. The two trials are not contradictions — they tested different populations, and the larger, more rigorous DREAM did not replicate the symptom benefit.
Infant DHA-fortified formula and visual acuity
positive · RCT
Birch et al., 2010, American Journal of Clinical Nutrition (and earlier 2002 trial)n=244Industry-fundedTerm infants fed DHA + ARA-fortified formula showed better sweep visual evoked potential acuity through the first year and into early childhood versus unfortified formula. The structural-mechanism evidence (DHA is the dominant fatty acid in rod outer-segment membranes) finally lined up with a functional vision outcome.
Industry-funded (Mead Johnson era). Formula DHA fortification is now near-universal; the comparator group no longer exists in routine practice. The signal is in infant visual acuity, not adult vision or AMD prevention.
Membrane structure and fluidity
Mechanism
DHA's six double bonds in a 22-carbon chain make it uniquely flexible, enabling tight membrane packing in neuronal and photoreceptor cells. Membrane DHA influences lipid raft composition, GPCR signaling, and ion-channel kinetics.
The most rigorous mechanistic story in the omega-3 family. DHA's role as a structural lipid in neuronal, retinal, and cardiac membranes is well established at the biophysical level. Translating membrane biophysics to specific clinical outcomes in adults has been the recurring difficulty — repletion benefits depleted tissue; supraphysiological dosing in already-adequate adults rarely shows measurable change.
Mechanism is real and important. Don't extrapolate from 'DHA is structurally critical' to specific supplemental-dose claims without an RCT in your population of interest.
Infant DHA-fortified formula and visual acuity
positive · RCT
Birch et al., 2010, American Journal of Clinical Nutrition (and earlier 2002 trial)n=244Industry-fundedTerm infants fed DHA + ARA-fortified formula showed better sweep visual evoked potential acuity through the first year and into early childhood versus unfortified formula. The structural-mechanism evidence (DHA is the dominant fatty acid in rod outer-segment membranes) finally lined up with a functional vision outcome.
Industry-funded (Mead Johnson era). Formula DHA fortification is now near-universal; the comparator group no longer exists in routine practice. The signal is in infant visual acuity, not adult vision or AMD prevention.
Cardiovascular: present but secondary to EPA
Mechanism
DHA lowers fasting triglycerides (less potently than EPA), modulates platelet activity, and incorporates into cardiac membranes. Unlike EPA, DHA tends to raise LDL particle size and LDL-C modestly in dose-response trials.
DHA does lower triglycerides, but the dominant cardiovascular omega-3 story belongs to EPA. The major positive cardiovascular trials in the omega-3 era — REDUCE-IT (pure EPA), GISSI-Prevenzione (mixed EPA + DHA in post-MI patients) — do not establish a primary DHA cardiovascular benefit. See the EPA and omega-3 dossiers for the cardiovascular case. The DHA-specific quirk is a small LDL-C rise at high doses, relevant for adults already managing hyperlipidemia.
Pick DHA for cognition / pregnancy / vision. Pick EPA-dominant formulas or icosapent ethyl for cardiovascular indications. Do not pick high-dose DHA-only if your LDL-C is already a concern.
AREDS2 — DHA + EPA in age-related macular degeneration
Null · RCT
AREDS2 Research Group, 2013, JAMAn=42034,203 adults at high risk for advanced AMD. Adding 1 g/day combined DHA + EPA to the AREDS antioxidant + zinc base did not reduce progression to advanced AMD. The trial that answered the long-running 'should I take fish oil for my eyes?' question with a clean no.
Background AREDS micronutrient regimen may have masked an effect, but the dedicated DHA + EPA arm specifically failed to add benefit. AREDS-type formulations no longer include DHA + EPA.
6 forms of DHA compared
Triglyceride form (algal or fish-derived)
Reference-standard absorption; ~50% better than ethyl ester in head-to-head trials
Best forPregnancy, cognition, vision — every documented DHA use caseThe form DHA occurs in naturally in fish and algae. Algal DHA (Schizochytrium-derived, branded life'sDHA among others) is the preferred pregnancy form — it provides DHA at trial-level doses without the methylmercury exposure profile of fish-derived oils.
Re-esterified triglyceride (rTG)
Equivalent to natural triglyceride form
Best forHigh-concentration DHA capsules without bumping into the ethyl-ester absorption ceilingConcentrate ethyl esters first to raise the DHA fraction per gram, then convert back to triglycerides. Combines high concentration with good absorption. Premium fish-oil suppliers (Nordic Naturals' Pro line, EPAX-based brands) commonly use rTG.
Triglyceride DHA (generic)
Same reference as rTG when labeled as triglyceride form
Best forGeneral-purpose DHAIf a label says 'triglyceride form' without specifying rTG, it is likely natural triglyceride at lower concentration. Both absorb well.
Ethyl ester DHA (EE)
Roughly 70% of triglyceride-form absorption when taken without a fatty meal (Dyerberg 2010)
Best forMost cheaper retail DHA capsules use this form because it is easier to concentrateFunctional, but take with a fat-containing meal to narrow the absorption gap. The form used in icosapent ethyl (EPA-only pharmaceutical) and many grocery-store fish-oil capsules. Worse oxidative stability than triglyceride form.
Flaxseed oil (ALA — not a real DHA source)
ALA → DHA conversion is under 1% in most adults; under 0.5% in men
Best forCooking oil; not a substitute for direct DHAFlaxseed contains ALA, the short-chain plant omega-3. The body's conversion to DHA is too inefficient to count toward DHA intake. Vegans needing DHA should use algal DHA, not flax.
Alpha-linolenic acid (ALA, plant omega-3)
Same conversion problem as flaxseed — under 1% to DHA
Best forBackground plant omega-3 from chia, walnuts, hemp, flaxUseful in the diet, useless as a substitute for direct DHA supplementation. For DHA needs, algal DHA is the only practical plant-based route.
Are you deficient? Symptoms, risk groups, lab tests
Average US DHA intake from non-fish-eating diets sits around 60–80 mg/day, well below the 200–300 mg/day floor that AAP and ACOG suggest for pregnancy and lactation, and far below the 500 mg/day combined EPA + DHA suggested by major dietary guidelines for adults.
Common symptoms
- Low Omega-3 Index (<4% RBC EPA + DHA, with DHA typically the larger fraction)
- Dry skin and reduced skin barrier function in low-fish-intake adults
- Reduced visual acuity in DHA-deplete infants (formula-era data)
- Cognitive complaints in older adults with very low fish intake — mechanistically plausible, not RCT-validated
- Faster maternal red-blood-cell DHA depletion through pregnancy and lactation in low-intake mothers
Who is at risk
Pregnant and lactating women
Fetal and infant brain DHA accretion draws heavily on maternal stores in the third trimester and during the first year of life. Mothers with intakes below 200 mg/day have measurably lower breast-milk DHA. AAP and ACOG both recommend ≥200 mg DHA/day in pregnancy and lactation.
Vegans and vegetarians
Zero direct DHA intake from diet. ALA-to-DHA conversion is typically under 1%, often under 0.5% in men. Algal DHA is the only practical supplementation route.
Adults eating little or no fatty fish
Most direct DHA in the typical US diet comes from salmon, sardines, herring, mackerel, and trout. Adults consuming less than 1 serving of fatty fish per week sit consistently low on the Omega-3 Index.
Premature infants
Most third-trimester DHA accretion happens late in pregnancy. Preterm infants miss part of that window and have lower DHA status at birth. Modern neonatal nutrition addresses this through DHA-fortified preterm formula.
Older adults with cognitive complaints and low fish intake
Population-level epidemiology associates higher fish intake with slower cognitive decline. Supplemental DHA RCTs have been mostly null in this population, but baseline diet matters — the highest-risk-deficiency subgroup is the population most likely to benefit if a benefit exists.
Adults with PEMT or FADS gene-cluster variants
Common variants in FADS1 / FADS2 reduce the already-modest endogenous conversion of ALA to EPA and DHA. Combined with low fish intake, these adults are functionally DHA-deplete even on omega-3-adequate plant diets.
Lab markers
Omega-3 Index (RBC EPA + DHA as % of total fatty acids)
The most validated long-chain n-3 status marker. DHA is typically the larger contributor to the index. More informative than serum DHA, which fluctuates with the last fatty meal.
- High risk (very low long-chain n-3 status)
- <4%
- Intermediate
- 4–8%
- Cardioprotective target
- >8%
RBC DHA fraction (% of total fatty acids)
Specific to DHA rather than EPA + DHA combined. Useful in pregnancy research and infant-feeding studies; less commonly available in routine clinical labs.
Better:Plasma phospholipid DHA, Omega-3 Index (RBC EPA + DHA)
Side effects and drug interactions
Side effects
Fishy burps and reflux
Common
The most common complaint with fish-derived DHA capsules. Worse with oxidized capsules and ethyl-ester forms. Less of an issue with algal DHA, which is largely odorless.
Worse with:ethyl ester
Gentler:Triglyceride form (rTG), Algal DHA
GI upset and loose stools at high dose
Common · Most pronounced above 3 g/day
Combined doses above ~3 g/day n-3 can cause loose stools and stomach discomfort. Splitting the dose across meals helps.
LDL-C rise (DHA-specific)
Uncommon · Most relevant above 2 g/day pure DHA
Dose-response trials show DHA tends to raise LDL-C modestly more than EPA does, particularly above ~2 g/day. The magnitude is small but relevant for adults already on lipid-lowering therapy or with elevated baseline LDL-C.
Increased bleeding tendency
Uncommon · Most relevant at ≥3 g/day combined EPA + DHA
Mild antiplatelet effect at high combined n-3 doses. Clinically meaningful mostly when stacked with anticoagulants or in the lead-up to surgery.
Methylmercury exposure (fish-derived DHA)
Uncommon
Crude or under-refined fish oils can carry methylmercury. Molecularly distilled supplements minimize this. The simpler workaround in pregnancy is algal DHA, which is mercury-free by source.
Worse with:triglyceride form, ethyl ester
Gentler:Algal DHA (Schizochytrium-derived)
Rancidity and oxidation byproducts
Uncommon
DHA's six double bonds make it particularly vulnerable to oxidation. Spoiled capsules carry oxidation products that may negate benefit. Smell-test the contents; refrigerate after opening; prefer capsules with vitamin E or rosemary extract as antioxidants.
Atrial fibrillation (high-dose n-3)
Uncommon · ≥3 g/day combined n-3
Pooled cardiovascular trials at the 3–4 g/day n-3 dose level (REDUCE-IT, STRENGTH) showed small but real increases in atrial fibrillation. Most relevant for high-dose users with cardiac history; not an issue at pregnancy/cognition doses.
Drug interactions
Additive effect
warfarinapixabanrivaroxabanclopidogrelaspirinDHA contributes to the mild antiplatelet effect of long-chain omega-3s. Additive with anticoagulants and antiplatelets.
Routine 200–500 mg/day DHA is generally fine on these meds. Discuss higher doses (≥2 g/day) with your prescriber. Pause supplementation 1–2 weeks before elective surgery.
Other
antihypertensivesHigh-dose combined n-3 produces a small (~2–4 mmHg) blood-pressure reduction. Stacking is usually beneficial; watch for over-correction in well-controlled hypertensives.
Monitor BP if you're starting high-dose DHA on top of antihypertensive therapy.
Combined-effect risk
statinsfibratesDHA modestly raises LDL-C while statins lower it. The interaction is usually clinically minor at typical DHA doses but worth recognizing in adults with brittle LDL targets.
Continue statin therapy; recheck lipid panel ~8–12 weeks after starting high-dose DHA. EPA-dominant formulas are a cleaner pairing for adults focused on cardiovascular endpoints.
Other critical caveats
- AREDS2 specifically tested DHA + EPA for age-related macular degeneration progression and found nothing. Supplement marketing that implies fish oil protects aging eyes is not supported by the dedicated RCT — adult AMD prevention via DHA is a failed trial, not an open question.
- DHA in established Alzheimer's disease has failed in two well-designed trials (Quinn 2010 in mild-to-moderate AD; OmegAD in mild-to-moderate AD). Modest cognitive signals in age-related-decline populations (MIDAS) are industry-funded and not strongly replicated. Do not treat DHA as a dementia preventive or treatment.
- DOMInO — the largest pregnancy DHA RCT — was null on its primary infant cognitive endpoint at 18 months and at 4- and 7-year follow-up. Pregnancy DHA matters most in mothers with very low baseline fish intake; the 200–300 mg/day pregnancy recommendation from AAP and ACOG is a reasonable floor, not a guarantee of cognitive benefit.
- DHA in pregnancy should generally come from algal sources, not fish oil. Algal DHA (Schizochytrium-derived) avoids the methylmercury exposure profile that gets flagged for pregnant women.
- DHA raises LDL-C modestly more than EPA does at higher doses. Adults focused on cardiovascular endpoints should reach for EPA-dominant formulas (or pharmaceutical-grade icosapent ethyl) — see the EPA and omega-3 dossiers. High-dose DHA is the wrong tool for cardiovascular risk reduction.
Frequently asked
Should I take DHA or EPA?
Depends on the goal. For pregnancy, infant feeding, cognition in older adults with low fish intake, or general brain/retinal structural support, DHA is the relevant fraction. For mood and depression as add-on therapy, cardiovascular event reduction in high-risk statin-treated patients with high triglycerides, and triglyceride-lowering, EPA is the active fraction — DHA-dominant blends and pure DHA do not show the same effect on those endpoints. Most general-purpose fish-oil capsules deliver both.Is algal DHA as good as fish-derived DHA?
Yes, and arguably better for pregnancy. Algal DHA (Schizochytrium-derived, branded as life'sDHA among others) has the same triglyceride-form bioavailability as fish-derived DHA at the same mg of DHA. It avoids the methylmercury exposure profile and is the only practical DHA source for vegans. Higher per-capsule cost is the main trade-off.How much DHA should I take in pregnancy?
AAP and ACOG recommend at least 200 mg/day DHA during pregnancy and lactation, achievable through 1–2 servings of low-mercury fatty fish per week or an algal DHA supplement. The largest controlled trials (DOMInO at 800 mg/day, ORIP at 800 mg/day) used higher doses and did not show population-level cognitive benefit — but the floor of 200 mg/day is well-supported as a deficiency-avoidance target. Pregnant women with very low fish intake have the strongest case for supplementation.Will DHA make my child smarter?
The honest read of the trial literature is: probably not at the population level, but maternal-deficiency avoidance still matters. DOMInO (n=2,399), the largest prenatal DHA RCT, was null on infant Bayley cognitive scores at 18 months and stayed null at 4- and 7-year follow-up. Supplement marketing that implies clear cognitive uplift from any maternal DHA supplementation is selling more than the controlled trials deliver.Does DHA help dry eye?
Mixed and probably no in moderate-to-severe disease. The DREAM trial (n=535, 3 g/day combined EPA + DHA for 12 months) was null on every dry-eye endpoint versus olive-oil placebo. Smaller earlier trials in low-omega-3-baseline computer users with mild symptoms (Bhargava 2013) were positive. If you have severe dry eye, treat the underlying ocular-surface disease — DHA isn't the answer the larger trial supports.Does DHA prevent Alzheimer's or slow cognitive decline?
Prevention: unknown but unproven. Treatment of established Alzheimer's: no — Quinn 2010 (2 g/day DHA, 18 months) and OmegAD were both null. The one positive signal (MIDAS, 900 mg DHA, 24 weeks in adults with subjective decline) was industry-funded, modest, and not strongly replicated. Don't reach for DHA expecting it to slow dementia.Will DHA raise my cholesterol?
Modestly, at higher doses. Dose-response trials show DHA tends to raise LDL-C more than EPA does, particularly above ~2 g/day. The rise is small and the LDL-particle size shift is generally favorable, but adults already managing hyperlipidemia should know about it. EPA-dominant formulas are a cleaner pick when lipid management is the goal.
References
- 01NIH Office of Dietary Supplements — Omega-3 Fatty Acids Health Professional Fact Sheet
- 02ACOG Committee Opinion — Nutrition During Pregnancy
- 03AAP — Fish, Shellfish, and Children's Health (committee guidance on omega-3 / DHA)
- 04Makrides et al., 2010, JAMA — DOMInO trial of prenatal DHA
- 05Quinn et al., 2010, JAMA — DHA in mild-to-moderate Alzheimer's
- 06AREDS2 Research Group, 2013, JAMA — Lutein, zeaxanthin, DHA + EPA for AMD
- 07Asbell et al., 2018, NEJM — DREAM dry-eye trial
Last reviewed2026-05-22