“Does curcumin reduce inflammation”

The strongest evidence is a 2023 review pooling 66 randomised trials in adults, meaning trials where people were assigned by chance to curcumin or to a comparison group [1]. Three markers of inflammation fell: C-reactive protein by about 0.58 milligrams per litre, interleukin-6 by 1.31 picograms per millilitre and tumour necrosis factor alpha by 3.48 picograms per millilitre [1]. Interleukin-6 and tumour necrosis factor alpha are chemical messengers the immune system uses to switch inflammation on and keep it running.

A fourth messenger, interleukin-1 beta, did not move [1]. That review measured blood markers only, so it says nothing about symptoms or illness [1].

Knee arthritis is the one place curcumin has been tested against something a person actually notices. Sixteen trials from a 2021 review, covering 1,810 adults with knee osteoarthritis, the wear-related joint disease, ran for up to 16 weeks [2]. Turmeric extract beat placebo on both pain and physical function by a large margin and matched the anti-inflammatory drugs it was compared against, with 12% fewer side effects than those drugs and no more than placebo [2].

Then the awkward part. Those same trials found no difference in inflammation markers and no difference on joint imaging [2]. Whatever helped those knees, the blood markers did not track it. Benefit also shrank as body mass index rose, a measure of weight scaled to height, so heavier participants improved less [2].

Absorption is the other problem. In a 2021 crossover study, twelve healthy adults each swallowed a single 207 milligram dose of curcumin in eight different formulations, and no free curcumin turned up in anyone's blood - only forms the body had already tagged for disposal [3].

The product decided almost everything. A micellar formulation, which wraps curcumin in tiny soap-like spheres so it stays dissolved, put 57 times more curcumin into the blood than plain curcumin powder, and a gamma-cyclodextrin version, another way of keeping it dissolved, 30 times more [3]. Formulations built around piperine, the black-pepper compound sold as an absorption booster, were no better than plain powder [3].

Two trials of "curcumin" can therefore deliver very different amounts to the body, which is part of why pooled results scatter [1][3].

None of this is a reason to start or stop a prescribed medicine. That is a decision for a doctor.

Where the evidence runs out:

• It shows markers moving, not disease changing. The pooled trials measured blood chemistry, not illness [1].
• The knee findings cover up to 16 weeks, and the researchers rated the trials as carrying a moderate risk of bias, meaning design flaws that could flatter the result [2].
• The absorption work used single doses in twelve healthy adults, so it says nothing about weeks or months of use [3].
• Nobody has shown that a 0.58 milligram per litre fall in C-reactive protein makes a measurable difference to anyone's health [1][2].