Most supplement decisions are made by reading a label. But in one controlled study the same iron dose was absorbed roughly four times better depending purely on what it was taken with — a difference of about twenty milligrams of actual absorbed iron, from one tablet, changed by a glass of orange juice instead of a coffee.
Part one argued that biochemistry drifts long before symptoms arrive, and that the drift is measurable. This is the harder half: what you actually do about it, and why most of the variables that decide whether an intervention works are invisible on the bottle.
I am going to use iron as the worked example throughout, because it is the nutrient where this has been studied most rigorously — with stable isotopes, in randomised crossover designs, measuring what was genuinely absorbed rather than what was swallowed.
Thirty-four iron-depleted women, 204 labelled doses of 100 mg iron, given under six different conditions and measured by isotope incorporation into red cells.
80 mg of vitamin C increased absorption by 30%. Raising it to 500 mg added nothing further.
Coffee reduced absorption by 54%. Coffee with breakfast reduced it by 66% — despite the meal containing around 90 mg of vitamin C from orange juice. The inhibitors won.
Afternoon dosing was 37% lower than morning, tracking a circadian rise in hepcidin.
Net effect: taking that dose with orange juice alone rather than with coffee or breakfast produced roughly a four-fold difference in absorption, and about twenty milligrams more absorbed iron per dose. Same tablet. Same number on the label.
This is why I find "just take a supplement" an unsatisfying sentence. Two people can follow identical advice and receive materially different amounts of the nutrient.
The iron story here is more interesting than the version circulating, and it repeats a mistake worth naming.
Stoffel and colleagues showed in The Lancet Haematology that giving iron on alternate days produced higher cumulative fractional absorption than consecutive days — 21.8% against 16.3% — because each dose raises hepcidin for about 24 hours and blunts the next one. In women with iron deficiency anaemia, absorption on alternate days ran 40–50% higher.
The functional medicine world adopted this quickly, and I did too. Then the outcome trials arrived.
A double-blind randomised controlled trial of 150 women found that at equal total doses, alternate-day dosing did not produce higher ferritin than consecutive-day dosing. It did reduce gastrointestinal side effects meaningfully, and iron deficiency at six months was lower (3.0% against 11.4%).
Meanwhile a 2025 trial found daily 60 mg outperformed 120 mg on alternate days on red cell indices over fourteen days. And in pregnancy, two separate trials found no advantage to alternate-day dosing at all, with the more recent suggesting daily dosing better meets pregnancy demand.
So: better absorption, better tolerability, and an unclear advantage on the outcome that matters. Fractional absorption is a surrogate. Ferritin and haemoglobin are the outcome. This is the same error I made with folate, and it is the commonest way a mechanism-led field gets things wrong.
What survives is still useful: morning, single dose, away from coffee, with something containing vitamin C — and alternate-day dosing as a legitimate option for someone whose gut cannot tolerate daily, rather than as a universally superior schedule.
Form matters, but selectively, and this is where labels do the most misleading.
Sometimes it genuinely matters. Methylcobalamin and cyanocobalamin are not the same molecule and are not handled identically. Magnesium oxide is poorly absorbed compared with citrate or glycinate. Curcumin is so poorly absorbed that trial formulations almost always include a bioavailability enhancer — which is why culinary turmeric does not replicate the trial evidence.
Sometimes it does not. N-acetylcysteine is N-acetylcysteine. Ascorbic acid from a bottle behaves much like ascorbic acid from a pepper at equivalent doses. And "natural" or "food-state" on a label is a marketing category, not a pharmacological one.
And sometimes the interesting variable is not the form at all. A genetic variant in TMPRSS6 was associated with fractional iron absorption of 18.5% against 26.6% in women without it — a substantial individual difference that no amount of choosing the right salt addresses.
Two things get confused here constantly.
Fractional absorption falls as dose rises. In the iron work, a sixfold increase in dose from 40 to 240 mg produced only a threefold increase in iron actually absorbed. More went in; proportionally less arrived. Larger doses also drive the hepcidin response that blunts the next one, and cause most of the gastrointestinal misery that ends adherence.
And unabsorbed material is not inert. Excess enteral iron irritates the gut, alters the microbiome, and reduces compliance — which is the mechanism by which a higher dose can produce a worse outcome.
Which is the argument for titration: start lower than the target, move up only as tolerated, and judge by a measurement rather than by how you feel three days in. A dose someone takes for six months beats a better dose they abandon in three weeks, and adherence is the variable most protocols quietly assume rather than plan for.
The idea that nutrients should be sequenced — support digestion before repletion, correct a cofactor before pushing the pathway it gates — is intuitive, mechanistically reasonable, and used constantly in my field.
Well supported: iron timing and co-ingestion effects; dose-dependent fractional absorption; specific form differences where the pharmacology differs; that adherence determines outcomes.
Reasonable but unproven: most sequencing protocols. There is very little trial evidence comparing "fix absorption first, then replete" against simply repleting. It is plausible. It has largely not been tested.
And where I would be sceptical: elaborate multi-phase protocols with a dozen products and a specific order. The complexity is rarely evidence-derived, it multiplies cost, and it makes adherence — the thing that actually predicts outcome — substantially harder.
The variables that decide whether a supplement does anything are mostly not on the label. When to take it. What to take it with. What to keep away from it. How much, and whether that amount is one you can actually sustain.
For iron specifically, and it is worth knowing because iron deficiency is common and badly treated: morning, single dose, with something containing vitamin C, at least an hour away from coffee, tea or a meal. Retest rather than guessing. If it wrecks your stomach, alternate days is a legitimate option with fewer side effects.
And the honest caveat that ought to accompany every protocol: absorption is not the outcome, and the fact that something reaches the bloodstream is not evidence that it changed anything.
Educational content, not medical advice. Do not start iron supplementation without knowing your iron status — iron overload is a real condition and supplementing on the assumption of deficiency is a genuinely bad idea. Ferritin is inexpensive and available from your GP.
Including whether the thing you are taking is arriving. That is a retest, not a hope.
See testing options Book a discovery call