SupplementsUpdated November 1, 2026

Best supplements for biohacking and recovery

Q: What supplements are actually worth taking?

Very few carry outcome evidence, and the two on this page carry it conditionally. Omega-3 has a strong triglyceride endpoint at 3 to 4 g/day (PMID 31422671) and a null cardiovascular prevention result at 1 g/day (PMID 30415637). Vitamin D is suggested for four specific groups and suggested against above the dietary reference intake for healthy adults under 75 (PMID 38828931). Everything else sold in the same aisle needs its own evidence, and most of it does not have one.

Q: How much omega-3 do I need for triglycerides?

At least 3 to 4 g/day of combined EPA and DHA for very high triglycerides, which is a pharmacological dose rather than a dietary one, and it can raise LDL at the same time (PMID 31422671). That is at least four capsules of a typical retail fish oil, which is why the serving size on the back is not the dose in the trial.

Q: Is fish oil good for your heart or not?

For preventing a first heart attack or stroke, 1 g/day did not work: major cardiovascular events came in at HR 0.92 (0.80 to 1.06), p=0.24 across 25,871 adults (PMID 30415637). One secondary endpoint, total myocardial infarction, was lower at HR 0.72 (0.59 to 0.90). The prescription drug result people quote is icosapent ethyl, not fish oil (PMID 30415628).

Q: Does fish oil raise the risk of atrial fibrillation?

The pooled data say yes, and the risk scales with dose. Across 7 trials and 81,210 patients, HR was 1.25 (1.07 to 1.46), rising to 1.49 (1.04 to 2.15) above 1 g/day against 1.12 (1.03 to 1.22) at or below it (PMID 34612056). It is graded Moderate because AFib was often an adverse event rather than an adjudicated endpoint.

Q: Do I need to test my vitamin D level?

Not routinely, if you are otherwise healthy. The 2024 guideline suggests against routine 25(OH)D testing in all the populations it considered, because no trial evidence supported screening and no target level for disease prevention was established (PMID 38828931). Testing still makes sense where a clinician has a specific reason for it.

Q: Should I take 5000 IU of vitamin D daily?

For a healthy adult under 75, the guideline suggests against it. Empiric vitamin D above the current dietary reference intake is not recommended to lower disease risk in that group, and the guideline states that the optimal dose across the populations it did consider remains unclear (PMID 38828931). The 4000 IU arm of D2d did not reduce diabetes either, HR 0.88 (0.75 to 1.04), p=0.12 (PMID 31173679).

Q: Is vitamin D3 better than D2?

For raising serum 25(OH)D, yes. A systematic review and meta-analysis found D3 more efficacious than D2 (PMID 22552031). That is a blood level rather than a health outcome, so treat it as a reason to prefer D3 in a product you were already going to take, not as a reason to take one.

Q: Do multivitamins do anything?

Not for mortality, in the largest long-term data available. In 390,124 US adults followed up to 27 years, daily use was not associated with lower all-cause mortality, HR 1.04 (1.02 to 1.07) in the first half of follow-up (PMID 38922615). It is observational, which is why the grade is Moderate rather than Strong.

Q: Is it worth taking NMN or collagen?

Neither has outcome data that survives this page standard. NMN and NR lack human outcome data at consumer doses, and collagen has no head-to-head trial against adequate dietary protein. Protein itself has a clear ceiling of about 1.62 g/kg/day (PMID 28698222), so the supplement is a convenience for hitting a target you should be hitting with food.

Q: Should I take creatine and magnesium too?

Those are different pages, because both have real evidence and both have buying decisions this one deliberately does not make. Creatine and magnesium each have a guide on this site with the dose and the verified price. This page covers only what the shelf sells hardest and explains least, which is omega-3 and vitamin D.

The supplement shelf is mostly a marketing surface. The evidence sits at doses that a retail bottle makes awkward and no label advertises, and for most healthy adults under 75 the finding for vitamin D points the other way. Here is what the trials support, and what they do not.

Written by

David Westbrooks

Published November 1, 2026

Not medical advice. Talk to a qualified clinician before starting any new health protocol, especially if you take medication or have a medical condition.

Read time

20 minutes

Evidence review

16 peer-reviewed sources

Products sold here

None

The gap between the bottle and the trial

The gap that matters most here is not what a pill does. It is how much of it you have to take before the trial result applies to you. People buy a bottle, take the suggested serving, and assume they are following the science. They usually are not, because the science mostly ran on doses that would mean swallowing a handful of capsules a day, and the serving size on the back is set by how many capsules feel reasonable, not by what was tested.

This page covers the two compounds no other buying guide on this site owns: omega-3 and vitamin D. There is no product pick, no price, and no ranking anywhere below. If you came for a specific bottle, this is not that page, and the honest reason is that no head-to-head trial of retail brands exists to rank them with.

Creatine, magnesium and nootropics each have their own evidence guide, because each has its own dose math and its own risks. Those links are at the end. Refusing to stack all five into one shopping list is not an apology for having nothing to sell. It is the only version of this page the evidence allows.

The one strong endpoint for omega-3

Omega-3 is not a general health tonic. It has one very specific, very strong endpoint, and it is triglycerides. In people with very high triglycerides, at or above 500 mg/dL, 4 g/day of EPA and DHA cuts them by at least 30% (PMID 31422671).

The trade is the part most people get wrong. At that dose LDL cholesterol rises alongside the triglyceride drop, and the EPA-only agent in the same advisory did not raise LDL (PMID 31422671). So the combined product is the one that works on triglycerides and the one that makes your LDL look worse, which is the opposite of the tidy story the shelf tells.

Then the arithmetic. A dose above 3 g/day is pharmacological rather than dietary, by the advisory's own framing, and reaching 3 to 4 g/day takes at least four capsules of a typical retail fish oil (PMID 31422671). That is the number a roundup leaves out, because telling you to take one capsule is easier than telling you the effective dose is bulky.

The honest read on heart health

Most people take fish oil hoping to avoid a heart attack or a stroke, and that is the endpoint where the evidence is weakest. VITAL randomised 25,871 adults to 1 g/day of marine n-3 or placebo for a median 5.3 years (PMID 30415637).

Major cardiovascular events came in at a hazard ratio of 0.92 (95% CI 0.80 to 1.06), p=0.24. That is a null result, written plainly: 1 g/day did not prevent heart attacks or strokes. Invasive cancer came in at HR 1.03, which is also nothing (PMID 30415637).

There is one movement in the data worth knowing. Total myocardial infarction was lower as a secondary endpoint, HR 0.72 (0.59 to 0.90) (PMID 30415637). People point at that 0.72 and say fish oil saves hearts. A secondary endpoint is a signal, not a proven benefit, and the honest sentence is that the primary outcome did not move while one subtype of event did.

The drug result is not the supplement result

REDUCE-IT is the study quoted to prove fish oil is a miracle, and it did not use a supplement. It used icosapent ethyl, a purified EPA prescription drug, in patients already on a statin with elevated triglycerides, and it found significantly fewer ischemic events including cardiovascular death (PMID 30415628).

That is a strong result for a drug. It is not evidence for a bottle off a shelf, and moving one onto the other is the most common error in this product class. The dose differs, the form differs, the population was already sick in a specific way, and one is regulated as a medicine (PMID 30415628). A drug that works in a selected population under supervision tells you almost nothing about what a combined EPA and DHA capsule does for a healthy forty-year-old.

The risk nobody prints on the label

There is a signal in this literature that appears on essentially no supplement label. Higher-dose omega-3 is associated with more atrial fibrillation. A 2021 meta-analysis pooled 7 cardiovascular outcome trials covering 81,210 patients and found HR 1.25 (1.07 to 1.46) for AFib (PMID 34612056).

It is dose-related, which is the part that should change what you do. Above 1 g/day the hazard ratio was 1.49 (1.04 to 2.15), and at or below 1 g/day it was 1.12 (1.03 to 1.22), with a significant interaction (p<0.001) and a dose-response of 1.11 per extra gram (PMID 34612056). The grade is Moderate rather than Strong because AFib was often recorded as an adverse event, so how hard each trial looked for it varies.

That combination is the argument of this page in one line. Taking 4 g/day for vague heart health, with no diagnosis and no clinician watching, buys you a null primary outcome and a raised risk of a rhythm disorder.

The 2024 guideline undercuts the shelf

The standard routine for years has been to test your 25(OH)D, then dose to a target, often 5000 IU. The 2024 Endocrine Society guideline does not support that routine for healthy adults (PMID 38828931).

It suggests empiric vitamin D for four groups: children and adolescents aged 1 to 18, adults 75 and older, people who are pregnant, and adults with high-risk prediabetes. It suggests against routine 25(OH)D testing in all the populations it considered, because no trial evidence supported screening and no target level was established (PMID 38828931).

For healthy adults younger than 75 it goes further and suggests against empiric vitamin D above the current dietary reference intake to lower disease risk (PMID 38828931). There is no number to quote for that group, and the guideline says why: trial doses varied too widely and participants were allowed to keep their own supplements, so the optimal dose remains unclear. What it does say is that for nonpregnant people over 50 who do need vitamin D, daily dosing beats intermittent high doses.

Why more vitamin D is not better

The idea that vitamin D is cheap insurance against everything does not survive the large trials. VITAL gave 2000 IU/day for a median 5.3 years and found no reduction in invasive cancer or cardiovascular events (PMID 30415629).

The fracture data went the same way. LeBoff found no effect on total fractures (HR 0.98, 0.89 to 1.08), nonvertebral fractures (0.97) or hip fractures (1.01), and baseline 25(OH)D did not change that (PMID 35939577). D2d pushed the dose to 4000 IU/day in people at high risk of diabetes and got HR 0.88 (0.75 to 1.04), p=0.12, which is not significant (PMID 31173679).

One result survives, and it needs describing exactly as it is. A secondary analysis of VITAL found advanced, metastatic or fatal cancer at HR 0.83 (0.69 to 0.99), strongest in participants with a normal BMI at HR 0.62 (0.45 to 0.86), with no effect at a BMI of 30 or above (PMID 33206192). It is graded Limited. It is a secondary analysis of a trial whose primary cancer endpoint was null, and the BMI split is a subgroup. Mention it, do not sell it.

What this page will not claim

That any brand beats another. No head-to-head outcome trials of consumer supplements exist, so there is nothing to grade and nothing to rank.
That third-party certification improves health. NSF, USP and IFOS are a proxy for label accuracy, and label accuracy is not a clinical endpoint.
A dose for anyone on anticoagulants, with kidney disease, or who is pregnant beyond what the guidelines themselves state. Those are clinical decisions.
That a supplement substitutes for the things with the strongest evidence on this site, which are training, sleep and not smoking.

Protein is a food decision, not a product

Protein powder is a convenience. The marginal benefit of adding more of it falls off fast. Across 49 trials, protein supplementation added about 0.30 kg of fat-free mass, with no further benefit above 1.62 g/kg/day (PMID 28698222).

The distribution advice matters more than the format: roughly 0.4 g/kg per meal spread across at least four meals (PMID 29497353), and 1.0 to 1.2 g/kg/day for adults over 65 (PMID 23867520). The before-sleep study that circulates widely used 27.5 g in 44 young men over 12 weeks and is graded Limited for that reason (PMID 25926415).

The kidney fear does not hold up in healthy adults, where higher-protein diets did not change kidney function (PMID 30383278). The practical conclusion needs no guide: under your daily target, the cheapest protein per gram wins, and over it, nothing helps. That is why protein products are not on this page. They moved to their own slot, waiting on a price pass.

The skip list

The multivitamin is the clearest example of a shelf product with no outcome behind it. In 390,124 US adults followed up to 27 years, daily multivitamin use was not associated with lower all-cause mortality, HR 1.04 (1.02 to 1.07) in the first half of follow-up (PMID 38922615). It is observational, and it is large, long and adjusted for the healthy-user problem, which is why it grades Moderate rather than weak.

Also on the skip list: NMN and NR, which have no human outcome data at consumer doses; collagen, with no head-to-head trial against adequate dietary protein; ashwagandha, where the trials are small and the liver injury case reports are real; and testosterone boosters, which are marketing with a milligram figure attached to something else.

One rule catches most of the rest. If a product is sold on a subscription, calls itself clinical strength, and will not state the milligrams, the dose is a marketing number rather than a dose.

The evidence, graded

Every study below was resolved against the live NCBI record and its abstract read on November 1, 2026. Grades describe how well the research supports the claim as written. Where a result rests on a secondary endpoint, a subgroup or a single population, the row says so.

Claim	Source	Grade	What it means
EPA+DHA at 4 g/day lowers triglycerides by at least 30% in very high triglycerides	AHA science advisory 2019 PMID 31422671	Strong	At or above 500 mg/dL. The same advisory notes a concurrent rise in LDL at that dose, while the EPA-only agent did not raise LDL.
Doses above 3 g/day of EPA+DHA are pharmacological rather than dietary	AHA science advisory 2019 PMID 31422671	Moderate	A definitional statement from an advisory rather than a trial outcome, which is why it is graded down. It is the basis for the capsule arithmetic.
1 g/day of marine n-3 did not reduce major cardiovascular events	VITAL, Manson 2019, NEJM PMID 30415637	Strong	25,871 adults, median 5.3 years, HR 0.92 (0.80 to 1.06), p=0.24. Invasive cancer HR 1.03. The primary result is null.
Total myocardial infarction was lower as a secondary endpoint	VITAL, Manson 2019, NEJM PMID 30415637	Strong	HR 0.72 (0.59 to 0.90). A prespecified secondary endpoint, so it is a signal from a trial whose primary outcome did not move, not a benefit.
Icosapent ethyl reduced ischemic events in statin-treated patients with elevated triglycerides	REDUCE-IT, Bhatt 2019, NEJM PMID 30415628	Strong	Strong for the prescription drug in that population. Dose, form, population and regulatory status all differ from retail fish oil.
Omega-3 supplementation is associated with higher atrial fibrillation risk	Gencer 2021, Circulation PMID 34612056	Moderate	7 trials, 81,210 patients, HR 1.25 (1.07 to 1.46). AFib was often an adverse event rather than an adjudicated endpoint, so ascertainment varies.
The atrial fibrillation risk is larger above 1 g/day	Gencer 2021, Circulation PMID 34612056	Moderate	HR 1.49 (1.04 to 2.15) above 1 g/day against 1.12 (1.03 to 1.22) at or below it, interaction p<0.001, dose-response 1.11 per extra gram.
The 2024 guideline suggests empiric vitamin D for ages 1 to 18, adults 75 and older, pregnancy and high-risk prediabetes	Demay 2024, Endocrine Society guideline PMID 38828931	Strong	GRADE-graded guideline built on systematic reviews. It also states that optimal doses across these populations remain unclear because trial doses varied and participants kept their own supplements.
The guideline suggests against routine 25(OH)D testing without established indications	Demay 2024, Endocrine Society guideline PMID 38828931	Strong	No trial evidence supported routine screening, and no target 25(OH)D level for disease prevention was established.
The guideline suggests against empiric vitamin D above the DRI for healthy adults younger than 75	Demay 2024, Endocrine Society guideline PMID 38828931	Strong	The single most useful sentence on this page for anyone taking 5000 IU as insurance. It is a suggestion in a graded guideline, not a trial outcome.
2000 IU/day of vitamin D3 did not reduce invasive cancer or cardiovascular events	VITAL, Manson 2019, NEJM PMID 30415629	Strong	Median 5.3 years in 25,871 adults. The large primary-prevention trial the supplement shelf was waiting for.
Vitamin D3 did not reduce total, nonvertebral or hip fractures	LeBoff 2022, NEJM PMID 35939577	Strong	Total fracture HR 0.98 (0.89 to 1.08), nonvertebral 0.97, hip 1.01, with no modification by baseline 25(OH)D.
4000 IU/day of vitamin D3 did not significantly reduce diabetes in high-risk adults	D2d, Pittas 2019, NEJM PMID 31173679	Strong	HR 0.88 (0.75 to 1.04), p=0.12. The cleanest test of the more-is-better premise, and it failed.
Vitamin D was associated with less advanced cancer, strongest at normal BMI	Chandler 2020, JAMA PMID 33206192	Limited	HR 0.83 (0.69 to 0.99) overall, 0.62 (0.45 to 0.86) at normal BMI, no effect at BMI 30 or above. Secondary analysis of a trial with a null primary cancer endpoint, so the subgroup is the weakest part.
Daily multivitamin use was not associated with lower all-cause mortality	Loftfield 2024, JAMA Network Open PMID 38922615	Moderate	390,124 US adults, up to 27 years, HR 1.04 (1.02 to 1.07) in the first half of follow-up. Observational, but large, long and adjusted for healthy-user bias.
Protein supplementation adds 0.30 kg of fat-free mass and nothing beyond 1.62 g/kg/day	Morton 2018, British Journal of Sports Medicine PMID 28698222	Strong	49 trials, meta-regression. The plateau is the finding, and it is why protein products do not need a guide of their own on this page.
The per-meal protein target is about 0.4 g/kg across at least four meals	Schoenfeld and Aragon 2018, Journal of the International Society of Sports Nutrition PMID 29497353	Moderate	A review synthesising the per-meal dose literature rather than a single trial. The four-meal figure is the practical floor.
Adults over 65 should target 1.0 to 1.2 g protein per kg per day	PROT-AGE, Bauer 2013, JAMDA PMID 23867520	Strong	A position paper from an expert group, higher than the general adult requirement and specifically for older adults.
27.5 g of protein before sleep increased mass and strength gains over 12 weeks	Snijders 2015, Journal of Nutrition PMID 25926415	Limited	44 young men. Small, single population, short. It is the study behind the pre-sleep protein advice and it does not carry the weight that advice gets.
Higher-protein diets did not change kidney function in healthy adults	Devries 2018, Journal of Nutrition PMID 30383278	Strong	Addresses the kidney concern for people without existing kidney disease. It says nothing about people who have it.
Vitamin D3 raises serum 25(OH)D more than vitamin D2	Tripkovic 2012, American Journal of Clinical Nutrition PMID 22552031	Moderate	Systematic review and meta-analysis. It settles the D2 versus D3 question on the surrogate endpoint, which is a blood level, not a health outcome.

Sources

Skulas-Ray et al. 2019. Omega-3 fatty acids for the management of hypertriglyceridemia: AHA science advisory. Arterioscler Thromb Vasc Biol. (PMID 31422671)
Manson et al. 2019. Marine n-3 fatty acids and prevention of cardiovascular disease and cancer (VITAL). N Engl J Med. (PMID 30415637)
Bhatt et al. 2019. Cardiovascular risk reduction with icosapent ethyl for hypertriglyceridemia (REDUCE-IT). N Engl J Med. (PMID 30415628)
Gencer et al. 2021. Long-term effect of omega-3 fatty acid supplementation on atrial fibrillation: meta-analysis. Circulation. (PMID 34612056)
Demay et al. 2024. Vitamin D for the prevention of disease: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. (PMID 38828931)
Manson et al. 2019. Vitamin D supplements and prevention of cancer and cardiovascular disease (VITAL). N Engl J Med. (PMID 30415629)
LeBoff et al. 2022. Supplemental vitamin D and incident fractures in midlife and older adults. N Engl J Med. (PMID 35939577)
Pittas et al. 2019. Vitamin D supplementation and prevention of type 2 diabetes (D2d). N Engl J Med. (PMID 31173679)
Chandler et al. 2020. Effect of vitamin D3 supplements on development of advanced cancer. JAMA Netw Open. (PMID 33206192)
Loftfield et al. 2024. Multivitamin use and mortality risk in 3 prospective US cohorts. JAMA Netw Open. (PMID 38922615)
Morton et al. 2018. A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength. Br J Sports Med. (PMID 28698222)
Schoenfeld and Aragon 2018. How much protein can the body use in a single meal for muscle-building? J Int Soc Sports Nutr. (PMID 29497353)
Bauer et al. 2013. Evidence-based recommendations for optimal dietary protein intake in older people (PROT-AGE). J Am Med Dir Assoc. (PMID 23867520)
Snijders et al. 2015. Protein ingestion before sleep increases muscle mass and strength gains during prolonged resistance-type exercise training in healthy young men. J Nutr. (PMID 25926415)
Devries et al. 2018. Changes in kidney function do not differ between healthy adults consuming higher- compared with lower- or normal-protein diets. J Nutr. (PMID 30383278)
Tripkovic et al. 2012. Comparison of vitamin D2 and vitamin D3 supplementation in raising serum 25-hydroxyvitamin D status: systematic review and meta-analysis. Am J Clin Nutr. (PMID 22552031)

Frequently asked questions

What supplements are actually worth taking?