The short list — highest probability of mattering

What actually decides this outcome, in order of size

A short list is a claim about probability, not about potency. What puts something on it is that the deficiency is common, the correction is cheap, and the effect survives being measured in people who were not already sick. Ranked by how confident that chain is:

1. Creatine, because it is the one whose mechanism is fully described in mmol/kg. Skeletal muscle holds roughly 120 mmol/kg dry mass of total creatine and saturates around 150 to 160; six days at 20 g/day raises muscle total creatine by about 20%, and 2 to 3 g/day holds it there Hultman 1996. That is a storage compartment being filled, not a signal being pushed, which is why the response is predictable and why non-responders are usually people whose diet already filled it.
2. Protein, where the marginal supplement is small in kg of lean mass and the total intake is not. The meta-analysis and meta-regression of protein supplementation against resistance training found the added benefit bounded, and bounded specifically by how much protein was already being eaten Morton 2018. A shake is a delivery convenience for a target measured in g/kg; it is not a separate intervention from the target.
3. Vitamin D, whose value is entirely a function of the ng/mL you start from. The distance between a 25(OH)D of 12 ng/mL and 30 ng/mL is a different purchase from the distance between 40 and 60, and only the first one has ever behaved like a deficiency correction. This is the item whose effect size depends most completely on a number you have not measured, and 25(OH)D has a half-life of 2 to 3 weeks.
4. Magnesium, which is on the list for prevalence rather than for drama, and is a cofactor for more than 300 enzymes. It is a cofactor for over 300 enzymatic reactions, the counter-ion for intracellular ATP, and the mineral whose intake most reliably falls short on a diet built from refined grains. It is also the one whose serum measurement is close to useless, because under 1% of body magnesium circulates and the concentration is defended.
5. Omega-3, with the ceiling stated plainly at 1 g/day. In 15,480 adults with diabetes and no known vascular disease, 1 g/day of n-3 fatty acids produced serious vascular events in 8.9% against 9.2% on placebo, a rate ratio of 0.97 (95% CI 0.87 to 1.08, P=0.55) over a mean 7.4 years Bowman 2018. One gram a day is a nutritional dose and across 7.4 years it did not prevent events. That is not an argument against membrane adequacy; it is an argument against buying the dose that failed and expecting the outcome that was not tested.

The order to run these in, and what has to be true first

Two of these four are worth taking without measuring anything, and two are not. That split is the whole ordering principle, and it turns on whether a titration against a number is possible.

1. Creatine first, because it needs no test and no titration above 5 g. Monohydrate at 3 to 5 g/day is the form every trial used, and the position stand covering safety and efficacy is explicit that habitual intake at that level is well tolerated Kreider 2017. The loading protocol is optional: it saturates the pool in about six days instead of about four weeks and changes nothing about the end state Hultman 1996.
2. Whey Protein (RecoveryPro) second, and only as arithmetic against a g/kg target. Work out the daily total first, then decide whether a 25 g scoop is how the gap gets closed. Whey is a fast-digesting fraction with a high leucine content, which matters for the acute mTORC1 response, and matters far less than the daily total does for the fourteen-week outcome.
3. Vitamin D third, and after a number in ng/mL. This is the item on the list where guessing is most expensive in both directions: a person at 12 ng/mL under-doses at 1000 IU, and a person at 55 ng/mL is buying nothing. Vitamin D (25-Hydroxy) costs less than three months of the capsule.
4. Magnesium fourth, chosen by form rather than by dose, because solubility decides absorption. Oxide is roughly 60% elemental magnesium and poorly soluble, which is why it is the cheapest and the most laxative; glycinate, malate and citrate trade elemental percentage for solubility and tolerance. Magnesium, RBC is the measurement worth having, because Magnesium (Serum) reports the defended compartment rather than the stored one.
5. Omega-3 (Fish Oil) fifth, at a dose that matches the claim, which is 1 g or 4 g and not both. One gram a day is a membrane-adequacy dose. Triglyceride-lowering and the cardiovascular outcome literature run at two to four grams of EPA, which is a different product at a different price, and conflating them is how the category earns its reputation for disappointing people.
6. Foundation Bundle and Foundational Health Bundle are convenience, and they trade 4 titrations for 1 price. A bundle is cheaper per unit and impossible to titrate, so it belongs at the point where the doses have already been settled individually rather than at the start.

What gets bought for this that cannot move it

A short list fails when the 4 items become a floor rather than a ceiling. The purpose of 4 items is that the fifth through the fortieth have worse odds, and the commonest failure on this page is somebody buying the four and then buying eleven more anyway. Every addition dilutes the attribution: with 15 products started in 1 week, a good response is uninterpretable and a bad one is unassignable.

Creatine's reputation for kidney harm is the most durable wrong belief in this category, and it comes from an assay artifact rather than from a filtration change. Creatine is non-enzymatically converted to creatinine at a roughly constant fractional rate, so a larger creatine pool produces more creatinine, and an estimated glomerular filtration rate calculated from creatinine falls without any change in kidney function. The narrative review of that literature is explicit that the failure signal has not materialized Longobardi 2023. Cystatin C with eGFR is not produced by muscle and is the way out of the confusion.

Buying the 1 g dose that failed and expecting the 4 g outcome is the specific error this list invites. One gram a day in 15,480 people did not reduce vascular events Bowman 2018. The products with cardiovascular outcome data are high-dose purified EPA, and a reader who wants that outcome is on ApoB & LDL particle reduction deciding about a prescription, not on this page deciding about a fish oil.

And if the reason you are here is that you feel bad, this list is the wrong tool by design. A short list is designed for somebody with no symptoms who wants the best odds per dollar over 12 months. A symptom is a question, and questions get answered by Test before you guess before they get answered by a shelf. The nutrient-specific version of this page is Micronutrient insurance, which exists because a deficiency and a hedge are not the same purchase.

How you would know it was working, on a real read-out and a real timescale

Three of the four have a number that moves; one does not, and saying which is which is what stops this page from being a promise. The prediction: creatine and protein show up in performance within weeks while the blood work stays still, and vitamin D shows up in blood work within weeks while nothing else changes at all.

• Vitamin D (25-Hydroxy) at 8 to 12 weeks. Roughly 1 ng/mL per 100 IU/day in an adult of average mass is the working expectation, with a wide spread driven by body fat, because 25(OH)D distributes into adipose. Somebody who does not move at all on 2000 IU/day has an absorption or a volume-of-distribution problem, and that is a real finding.
• Magnesium, RBC at 12 weeks rather than Magnesium (Serum) at any time. The red cell value integrates over the erythrocyte lifespan and reports the stored compartment; the serum value is held within a narrow band by bone exchange and renal handling and can be normal in frank depletion.
• Body mass at 7 days for creatine, which is the fastest honest signal on this page. Creatine is osmotically active and intracellular, so saturation pulls water into muscle: 0.5 to 1.5 kg in the first week on a loading protocol is the expected and entirely intracellular change. If nothing moves and strength does not change over 6 weeks, this is a non-responder with a full pool, which is a result rather than a failure.
• Cystatin C with eGFR alongside the Comprehensive Metabolic Panel (CMP) if kidney function is ever questioned on creatine. Cystatin C is produced by all nucleated cells at a steady rate and is independent of muscle mass, so a normal cystatin C with a raised creatinine is the pattern that resolves the artifact rather than arguing about it Longobardi 2023.
• ApoB (Apolipoprotein B) and hs-CRP (High-Sensitivity C-Reactive Protein) at 12 weeks if omega-3 is being asked to do a cardiovascular job. A nutritional dose should move neither much, and that is the point: the null result tells you the dose matches the membrane claim rather than the outcome claim.

What will fool you. The first four weeks of any new routine include the placebo, the novelty and the fact that people who start supplements usually also start sleeping and training differently. Change 1 thing at a time, or accept that you have bought a bundle whose components you will never separate. And weight gained in week one on creatine is water in muscle, not fat, which is worth knowing before the scale talks somebody out of the best-evidenced item on the list.

Sources read for these sections

• Hultman E, et al. Muscle creatine loading in men. Journal of Applied Physiology, 1996 · PMID 8828669
• Kreider RB, et al. International Society of Sports Nutrition position stand: safety and efficacy of creatine supplementation in exercise, sport, and medicine. Journal of the International Society of Sports Nutrition, 2017 · PMID 28615996
• Morton RW. A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. British Journal of Sports Medicine 2018;52(6):376-384 · PMID 28698222
• Bowman L. Effects of n-3 Fatty Acid Supplements in Diabetes Mellitus. New England Journal of Medicine 2018;379(16):1540-1550 · PMID 30146932
• Longobardi I, et al. Is It Time for a Requiem for Creatine Supplementation-Induced Kidney Failure? A Narrative Review. Nutrients 2023 · PMID 36986197