Curcumin bioavailability: why most turmeric capsules do so little

A Japanese research group investigated in rats in what form curcuminoids appear in the blood after oral administration. They found almost exclusively glucuronides and mixed glucuronide sulphate conjugates, hardly any free curcumin. The enzymes needed for this sat in liver, kidney and intestinal mucosa.

For you that means: the conversion does not happen somewhere at the margins, but directly on the way from the capsule into the blood. That is not a product fault, that is physiology.

A Dutch team at Amsterdam UMC took a pragmatic approach. In 47 people taking their own, self bought curcumin products, it measured the actual plasma levels before and 1.5 hours after intake.

The values lay between 1.0 and 18.6 ng/ml. After enzymatic cleavage they rose to up to 25.4 ng/ml. The authors calculate that this is roughly a thousandfold below the concentration you would expect for a systemic effect from laboratory experiments. Products containing piperine did not perform better here.

What that means for you: the question is not whether your product is good. The question is whether enough arrives at all to make the question worth asking.

An Indian research group around Shoba gave healthy volunteers 2 g of curcumin, once alone and once together with 20 mg of piperine. Piperine inhibits glucuronidation, so exactly hurdle three.

Without piperine the serum levels were either undetectable or very low. With piperine they rose markedly in the first hour. From this the authors calculated an increase in bioavailability of 2000 percent. In rats it was 154 percent.

The important point: 2000 percent of almost nothing is still very little. The study showed a real effect, but it does not answer the question of whether therapeutically relevant levels are reached in the process.

In 2025 an independent Dutch crossover study in nine healthy men compared three formulations at around 570 mg, one of them additionally at a fourfold dose, each with and without piperine.

The unconjugated curcumin stayed below 2 nmol/l in most cases, including at the high dose and including with piperine. The micellar formulation briefly reached 6.7 to 38 nmol/l, but fell away quickly and was still around a hundredfold below the concentrations at which effects were shown in the laboratory. Adding piperine brought no advantage in this investigation.

A German comparison study in twelve adults, which set seven different preparations against each other, also found no statistically meaningful increase for the variant with adjuvants including piperine.

A team at the University of Hohenheim compared seven preparations with the same amount of curcumin, namely 207 mg, in a randomised double blind crossover study in twelve healthy adults: native, liposomes, with turmeric oils, with adjuvants including piperine, submicron particles, phytosomes, cyclodextrin complex and micelles.

In not a single participant was free curcumin detectable, only conjugated. Statistically meaningfully above native curcumin lay only the micelles at 57 times the value and the cyclodextrin complex at 30 times. From this the authors conclude: approaches that increase solubility after digestion are successful. Approaches meant to slow down breakdown appear to have little success.

In my eyes this single study explains more about the turmeric market than any advertising brochure.

In the Hohenheim study from 2014, 13 women and 10 men took 500 mg of curcuminoids in three forms. Micellar curcumin was 277 times above native curcumin in the women and 114 times in the men.

So in this investigation women absorbed the substance considerably more efficiently than men. Why that is has not been conclusively clarified. Differences in bile acids, body composition and enzyme activity are being discussed.

For you that means: blanket statements about absorption are not a fixed quantity even between two healthy people.

A Viennese research group gave 15 healthy volunteers either micellar or native curcumin for seven days, 105 mg daily in each case.

The micellar form reached roughly 39-fold higher peak concentrations and a roughly 14-fold greater total exposure. Even so, the inflammatory messengers interleukin-6 and TNF-alpha did not fall in the chosen test model. Striking was a roughly ten percent reduction in PCSK9, a regulator of cholesterol metabolism, which the authors class as worth observing.

Better absorption is therefore a necessary condition, but not a sufficient one. That is an important intermediate step which advertising likes to skip.

In the investigation of the lecithin formulation Meriva, not only a roughly 29-fold higher total uptake showed up. The profile also shifted.

The curcuminoid dominating in plasma after Meriva was not curcumin itself, but demethoxycurcumin, a related molecule from the turmeric root that appears more active in inflammation related terms in several laboratory experiments. The authors consider it possible that not only the amount but also this altered profile plays a role.

Here too the same applies: what was detected were phase 2 metabolites, and the plasma levels still lay below what would be needed for most laboratory targets.

A network meta-analysis from 2023 evaluated 23 randomised studies from seven countries with a total of 2,175 patients with knee osteoarthritis.

Compared with placebo, an improvement in the pain score and the overall WOMAC score showed up. Compared with anti inflammatory painkillers, fewer unwanted effects occurred under curcumin. An older meta-analysis with eleven studies and 1,009 participants came to a similar picture, but explicitly emphasised the overall low study quality and the small case numbers.

What that could mean for you: in knee osteoarthritis there are the comparatively best clinical signals. That is no substitute for treatment, but it is also not nothing.

A much discussed review in the Journal of Medicinal Chemistry classes curcumin as a so called PAINS substance, meaning a molecule that frequently produces unspecific signals in laboratory experiments without genuine target binding.

The authors describe curcumin as unstable, reactive and poorly bioavailable and therefore consider it an unlikely drug candidate. This paper has triggered a lively specialist debate in which it has also been contradicted.

I cite it because it belongs to honesty. Anyone recommending turmeric should know this criticism and not conceal it.

A Stuttgart research group showed in Caco-2 cells and human liver microsomes that piperine inhibits both the transport protein P-glycoprotein and the enzyme CYP3A4.

Both sit in gut and liver cells and are substantially involved in the breakdown of many orally taken medicines. From this the authors conclude that piperine from food could influence the blood levels of the corresponding active substances.

This is in vitro work, not proof of a clinical interaction in humans. But it is enough signal to ask when medicines are taken regularly, instead of simply combining.