A supplement label tells you what is in the capsule and how much of it is there. It tells you far less about how much of that dose will actually reach your bloodstream, how quickly it will arrive, or what happens when a meal, another supplement, or a medicine enters the equation. Those details can matter enormously. For some compounds, eating first raises exposure several times over. For others, food sharply lowers it. Sometimes the formulation matters more than the meal, and sometimes the way a food is processed decides whether the active compound is even produced. This is why “take one capsule daily” is only part of the pharmacology.
The dose you take is not the dose your body receives
Before an oral compound can do anything, it has to dissolve, survive the gut, cross the intestinal wall, and in many cases run a gauntlet of metabolism in the gut and liver. The fraction that finally reaches the circulation is its bioavailability, and it is nowhere on the panel.
Absorption and bioavailability are not the same thing, which is a distinction worth keeping. A compound can cross the gut wall efficiently and still arrive in the blood in tiny amounts, because so much of it is metabolized on the way. Resveratrol is the textbook case. Its oral absorption has been estimated at roughly 75%, yet the bioavailability of the unchanged compound is under 1%, because metabolism in the gut and on the first pass through the liver is so extensive.[1] The useful question, then, is not how many milligrams you took. It is what exposure those milligrams actually produced.
More absorption is not automatically better
It is tempting to treat maximum absorption as the goal. However, pharmacology is rarely that tidy, and two examples pull in opposite directions.
A high-fat meal has been reported to increase CBD exposure roughly fourfold, and its peak concentration by more, in a human food-effect study.[2] That is a genuine, substantial effect. It does not follow that every person taking CBD should deliberately maximize it.
Green tea poses the reverse problem. Its catechins reach much higher blood levels fasted, and in one trial the peak EGCG concentration was reported to be more than three times higher after an overnight fast than with breakfast.[3] However, sustained high-dose green-tea extract has also produced clinically significant liver-enzyme abnormalities in a randomized trial.[4] That study dosed the extract with food, so it does not show that fasting caused the liver signal. It does show why “higher exposure” and “better use” have to be treated as separate questions. The real objective is appropriate exposure, with acceptable safety, tolerability, and interaction risk.
Food can move exposure in either direction
The fed-or-fasted state is one of the better-supported variables in this literature. Even so, there is no universal “take supplements with food” rule, because the direction of the effect depends on the compound.
Boswellic acids sit at one extreme. In human studies, frankincense preparations have been reported to reach several-fold higher blood levels when taken with a fat-containing meal.[5,6] Omega-3 ethyl esters are similarly meal-sensitive. A high-fat meal has been reported to roughly triple the absorption of EPA and DHA from ethyl esters.[7] Vitamin D3 also benefits from dietary fat, although far more modestly, by about 32% in one study.[8] EGCG, as above, moves the other way. Therefore “with food” is not a property of supplements as a class. It is a compound-specific, and often formulation-specific, question.
Formulation can outweigh the clock
The product itself can matter more than any schedule: two supplements with the same ingredient name can behave quite differently. Silybin shows this cleanly. In a human crossover trial, a silybin-phosphatidylcholine complex was reported to produce substantially higher plasma levels than conventional silymarin tablets.[9]
Omega-3 makes the same point. Re-esterified triglyceride and ethyl-ester preparations do not necessarily give the same exposure, and some self-emulsifying formulations were designed specifically to improve EPA and DHA absorption under fasting conditions.[10] In other words, the “take it with fat” rule does not even apply to every fish-oil product. Curcumin is often discussed as though one rule fits all of it, when plain curcumin, phospholipid complexes, emulsified products, and piperine-enhanced products are genuinely different pharmacokinetic propositions. Before asking when to take a compound, it is often necessary to ask which version of it you have.
What the studies actually say, compound by compound
The table below distils the best-supported observation for each compound in this research set, alongside the qualification that keeps it honest. It also makes a subtler point clear. Some compounds carry a real, practical timing signal. Others mainly carry formulation, mechanistic, or interaction information, and those categories should not be read as if they were the same thing.
| Compound | Best-supported absorption / co-dosing observation | Key qualification |
|---|---|---|
| EGCG / green tea | Exposure is substantially higher fasted; peak EGCG was reported above 3.5× the fed value. | Higher exposure is not automatically preferable. High-dose extract carries a human liver-safety signal. |
| Sulforaphane | Enzyme (myrosinase) availability is decisive. Fresh broccoli beat frozen; added brown mustard raised availability >4×; split 12-hour dosing held later levels better. | Unusually strong human preparation and dosing data. |
| Boswellia | A fat-containing meal has been reported to raise boswellic-acid exposure several-fold. | One of the clearest human food effects in the set. |
| CBD | A high-fat meal raised exposure roughly 4×, and peak concentration by more. | Greater exposure can also raise effects and interaction potential. |
| CBG | Oral exposure appears poor; animal work points to formulation and lymphatic transport. | No robust human food-timing rule yet. |
| Omega-3 (EPA/DHA) | Ethyl-ester absorption improves markedly with a fatty meal; triglyceride and engineered forms differ. | The triglyceride-versus-ethyl-ester gap is not a fixed number. |
| Curcumin | Piperine produced a large increase in one small human study; lipid delivery improves it. | No specific dietary-fat multiplier is established. Do not read formulation data as “eat it with fat.” |
| Silymarin / silybin | A phosphatidylcholine complex substantially raises silybin exposure versus tablets. | Primarily a formulation issue, not a clock-time one. |
| CoQ10 | Absorption is lipid- and formulation-sensitive; elimination is long (about a 92-hour half-life reported for one formulation). | The often-quoted “5× food effect” actually belongs to idebenone, not CoQ10. |
| Ursolic acid | Phospholipid complexing greatly raised exposure in animal studies. | Human timing data are limited. Poor exposure should not be blamed on intestinal glucuronidation. |
| Thymoquinone | Lipid formulations raise exposure in animal models. | No quantified human fed-versus-fasted effect. Its redox behaviour is context-dependent. |
| Phosphatidylcholine | Handled through lipid digestion (pancreatic phospholipase, lyso-PC) before absorption. | Tied to fat digestion, but “must be taken with fat” overstates it. |
| Artichoke extract | Intact major polyphenols are scarce in plasma; metabolites dominate. | Its choleretic (bile-stimulating) effect does not prove bile-dependent absorption. |
| Quercetin | Raised exposure to the drug fexofenadine in humans by inhibiting an efflux pump. | Supports interaction awareness, not a rule to space it from other polyphenols. |
| Berberine | Oral bioavailability is poor; efflux, metabolism, and microbial conversion shape exposure. | A berberine/PPI or gastric-pH timing rule is unsubstantiated. |
| Dihydroberberine | Animal work suggests better intestinal absorption than berberine, then re-oxidation to it. | Human data remain thin. |
| Resveratrol | Piperine dramatically raised exposure in mice. | A human randomized trial did not reproduce the increase. |
| Luteolin | Oral bioavailability appears poor. | No good evidence that dietary fat reduces its absorption; that claim seems to come from confusion with the carotenoid lutein. |
| Taurine | Interacts with β-alanine at the TauT transporter in cell systems. | A long human β-alanine trial did not lower muscle taurine, so this should not become a spacing rule. |
| Tributyrin | Human dosing produced measurable but transient plasma butyrate, mostly gone by about five hours. | Suggests frequency may matter; does not prove an interval, and fasted dosing still worked. |
Co-dosing can raise or lower exposure
Food is only one variable. The intestinal wall is studded with transport proteins that move compounds into and out of cells, and nutraceuticals can interfere with them. The direction of the effect depends on which transporter is involved, and the two examples below use the very same probe drug.
Green-tea extract has been reported to cut exposure to the drug fexofenadine by roughly 70%, apparently by inhibiting an intestinal uptake pathway (an OATP transporter) that the drug relies on.[11] Quercetin did the opposite with the same drug. In healthy volunteers, it was reported to raise fexofenadine exposure by about 55%, consistent with blocking a P-glycoprotein efflux pump that normally pushes the drug back out.[12] Same drug, opposite results. This is also why the popular instruction to “space polyphenols apart so they do not compete” is not well supported. Almost all of the good human evidence pairs a nutraceutical with a pharmaceutical probe, not one supplement with another. There is not enough direct supplement-versus-supplement data to build a universal spacing timetable.
The black-pepper lesson: a real effect that does not generalize
Piperine, the pungent compound in black pepper, is the best-known nutraceutical “absorption enhancer.” In a small human study, it was reported to increase curcumin exposure dramatically.[13] The mistake is assuming the trick transfers. Piperine greatly increased resveratrol exposure in mice, yet a human randomized trial found no significant improvement.[14] “Enhances bioavailability” is not a portable property that can be pasted from one compound onto the next.
Grapefruit tells the same cautionary story from the drug side. Its furanocoumarins can inhibit an intestinal enzyme (CYP3A), which may raise exposure to susceptible medicines, and this is the basis of the familiar pharmacy warning. However, grapefruit can also inhibit intestinal OATP uptake and thereby reduce exposure to other substrates. Once again, direction is the whole point.
Sulforaphane: four levers, and none of them a clock
No compound makes the point better than sulforaphane, because four separate levers change how much you absorb, and not one of them is the hour you take it. Broccoli mostly contains a precursor, glucoraphanin, and turning it into sulforaphane depends on the enzyme myrosinase. Commercial blanching inactivates that enzyme, which helps explain why fresh broccoli was reported to produce far greater sulforaphane exposure than commercial frozen broccoli in a human study.[15]
The missing enzyme can also be supplied from elsewhere. Adding just 1 g of raw brown mustard powder to cooked broccoli was reported to raise sulforaphane availability more than fourfold in a crossover study.[16] Frequency shifted the picture too. In another human study, dividing the dose into two portions 12 hours apart sustained later blood levels better than a single daily dose.[17] For a single compound, then, processing, enzyme availability, co-dosing, and frequency all change exposure materially. Choosing 8 a.m. over 8 p.m. barely registers by comparison.
Some things really should be kept apart
The evidence for “avoid” rules is thinner than the evidence for useful pairings, but a few are legitimate. The clearest concerns iron. Galloyl-rich tea polyphenols have been reported to substantially reduce absorption of non-heme (plant-source) iron, and in a classic human experiment the effect grew strongly with the amount of tannic acid present.[18] Concentrated green tea or an EGCG supplement and an iron supplement are therefore a sensible pair to separate. That finding should not be inflated into “polyphenols block minerals,” however. It concerns iron and particular galloylated phenolics, not every mineral or every flavonoid.
It is just as useful to name the plausible “avoid” rules the evidence does not support. A berberine and proton-pump-inhibitor timing interaction has not been established. β-Alanine competes with taurine at the TauT transporter in cell systems, yet long-term human β-alanine supplementation did not significantly lower muscle taurine. The idea that dietary fat suppresses luteolin absorption appears to be a lutein-versus-luteolin mix-up rather than a demonstrated effect. Negative findings matter, because they stop a plausible mechanism from hardening into an unnecessary schedule.
The redox question: biology first, stopwatch later
The idea of separating antioxidants from certain compounds does have a mechanistic basis. In laboratory cancer models, sulforaphane and thymoquinone generate oxidative stress as part of how they kill cells, and antioxidants such as N-acetylcysteine can suppress that effect in the dish.[19] That is a real mechanism, in cells. However, no human trial shows that a person taking oral sulforaphane or thymoquinone should separate it from vitamin C or NAC by two hours, or four, or six, or any other invented interval. A cellular mechanism can justify further study. By itself, it cannot generate a human dosing schedule. Reading a petri-dish result as a clock is one of the easier ways to manufacture false precision, which is also the theme of what “preclinical” actually means.
Clock time is often the least-studied kind of timing
Circadian biology clearly affects how the body handles drugs, and human aspirin studies have reported differences in platelet effects between morning and bedtime dosing.[20] For most nutraceuticals in this article, however, the evidence is far stronger for fed-versus-fasted, formulation, co-dosing, and dose frequency than for morning versus evening. Melatonin is the obvious exception, because clock timing is part of its biology. For the rest, there is simply not enough evidence to claim that 8 a.m. is intrinsically better than 8 p.m. That absence is itself informative. It marks where precision is earned and where it is invented, and consistency of routine usually matters more than the exact hour.
Medicines change the stakes
The same mechanisms that make supplement absorption interesting can make drug interactions clinically important, which is the part that matters most. A CBD-dominant extract has been reported to raise exposure to CYP2C19 and CYP3A probe drugs in healthy adults.[21] In addition, laboratory work suggests CBD can induce CYP3A after prolonged exposure, so single-dose and repeated use may not behave identically.
Berberine warrants similar care alongside glucose-lowering treatment. It has been reported to lower glucose on its own, and one add-on trial reported further reductions on top of metformin.[22] That add-on product contained several ingredients, however, so it does not prove a clean berberine-alone effect. Statins offer a different lesson. They have been reported to lower circulating CoQ10.[23] Whether that reflects meaningful depletion in muscle is far less certain, and a randomized trial found that CoQ10 supplementation raised serum levels markedly but did not relieve confirmed statin muscle pain.[24] The point is not to self-manage any of this by rearranging a supplement schedule. It is that “natural” compounds move through the same enzymes and transporters as medicines, which is a conversation for your care team.
So, when should you take it?
There is no single timing rule, but there is a better sequence of questions. First, identify the formulation, because two versions of the same compound can behave differently. Next, ask whether there is direct human evidence for a fed-or-fasted effect. Boswellia, CBD, EGCG, and omega-3 ethyl esters all have useful evidence here. Then ask whether the compound depends on conversion or preparation, as sulforaphane so clearly does. Also weigh co-dosing: piperine can transform curcumin exposure, tea polyphenols can blunt iron absorption, and transporter or enzyme effects can change how a prescription drug behaves. Finally, separate a demonstrated effect from a mechanistic theory. A cell study can explain why something might matter. It does not, by itself, tell you what time to take it.
The most useful advice, in the end, is not “take everything with breakfast” or “space supplements two hours apart.” It is more specific than that. Know the compound. Know the formulation. Know the meal effect. Know what else is on board. And make the timing rule only as precise as the evidence allows. None of this is a cancer-treatment claim, and none of it is a dosing instruction. It is the reason to hand your care team the whole list, not just the what, but the when and the with-what-else.