Login

Guides

Inflammation: what the trials show

Four supplements with large pooled evidence for lowering inflammatory markers. What each one moved, at what dose, over how long, and the honest limit on what a lower blood marker actually buys you.

Last scientifically reviewed: November 22, 2026Editorial review: MDS Labs. External scientific review pending.11 minute read

In brief

  • Four supplements here have been pooled across dozens of randomized trials each. All four lower inflammatory markers in the blood. That is a real, repeatable finding.
  • Lowering a marker is not the same as feeling better. Keep that separation in mind for everything below; it is the single most important idea on this page.
  • Curcumin has the deepest evidence, and the honest version of it is that most measured outcomes did not move. The ones that did include C-reactive protein, rated high quality.
  • Boswellia is the outlier: it was tested on how people's knees actually felt and functioned, not just on blood tests. It needed at least four weeks.
  • Curcumin on its own absorbs poorly. If a label does not say what it uses to fix that, the trial it is pointing at probably used something the product does not contain.

What these trials actually measure

Almost every result below is a blood test, so it is worth thirty seconds on what the three recurring names mean.

CRP, or C-reactive protein, is made by the liver whenever inflammation is active anywhere in the body. It is cheap, standard, and the one your doctor is most likely to have already ordered. hs-CRP is the same thing measured with a more sensitive method, used when the levels are low.

IL-6 and TNF-alpha are signalling molecules. Cells release them to tell other cells that something is wrong, which is useful during an infection and corrosive when it never switches off. IL-1 beta is a third one in the same family, and it matters below because it is the marker that did not move.

None of these are diseases. They are messages. A supplement that quiets the messages has done something measurable, and whether that something is worth having depends on the next section.

Why inflammation is on this list at all

The reason this page exists is a pattern researchers call inflammaging: as people get older, these markers drift upward even without any illness to explain it, and that chronic low-grade background inflammation is itself a risk factor for illness and death in older adults. Most age-related diseases share an inflammatory thread 1.

That makes lowering the markers an attractive target. It does not make it a proven one. Nobody has shown that taking a supplement to push your CRP down adds years to your life, and the trials below were not built to test that. What they were built to test, they answer well.

Curcumin, from turmeric

This is the largest evidence base on the page by a distance. One analysis pooled 103 randomized trials covering 7,216 participants and looked at 42 different outcomes 2.

What the trial shows

Curcumin significantly improved 23 of those 42 outcomes. More useful than the count is the quality rating: using GRADE, the standard system for grading how much confidence a body of evidence deserves, the reviewers rated the credibility of the evidence high for C-reactive protein, fasting blood sugar, HDL cholesterol and body weight 2. "High" is not a word that appears often in supplement research.

A separate analysis of 66 trials put numbers on the size of the effect: CRP fell by 0.58 mg/L, TNF-alpha by 3.48 pg/mL and IL-6 by 1.31 pg/mL, with a dose-response relationship, meaning more produced more 3.

What it does not show

Nineteen of the 42 outcomes did not reach significance, and IL-1 beta specifically did not budge 23. Curcumin is not a general anti-inflammatory that dampens everything; it hits some signals and misses others.

And none of this is evidence about a specific illness. These are markers in mixed populations. A lower CRP is a plausible good sign, not a treated condition.

The absorption problem is the practical catch. Plain curcumin powder is poorly absorbed, and the trials above largely used formulations built to get around that. We cover which approaches have evidence behind them in the curcumin ingredient profile.

Omega-3 fatty acids

Omega-3 has been studied so many times that the useful summary is a review of reviews. An umbrella analysis pooled 32 separate meta-analyses across a range of health conditions 4.

What the trial shows

All three main markers moved in the same direction and all three were statistically significant: CRP, TNF-alpha and IL-6. The effect sizes were small to moderate, and the reviewers concluded omega-3 can be recommended as an adjuvant anti-inflammatory agent 4 — a supporting player alongside treatment, which is a fair description of what the numbers support.

What it does not show

Consistency across 32 meta-analyses is the strength here, not the size of the benefit. The effects are modest. And pooling across many different conditions means the average tells you less about your own situation than a single well-matched trial would.

Ginger

Sixteen randomized trials in 1,010 participants were pooled 5.

What the trial shows

Ginger significantly reduced CRP, hs-CRP and TNF-alpha 5. For a culinary spice with no safety concerns at food-like doses, that is a reasonable result.

What it does not show

IL-6 did not change, and neither did sICAM, a marker of blood-vessel-lining activation 5. The same selective pattern as curcumin.

The bigger caveat is that the trials disagreed with each other to an unusual degree. Statisticians measure this as heterogeneity, and here it was very high, which means the pooled average is a summary of studies that found quite different things rather than a consistent signal repeated many times. Treat the ginger numbers as weaker than their sample size suggests.

Boswellia, or Indian frankincense

Seven randomized trials in 545 people with osteoarthritis 6. This is the smallest evidence base on the page and, in one specific way, the most useful.

What the trial shows

Boswellia was measured on how people's joints actually were, not on their blood. Pain scores fell, stiffness fell, physical function improved, and the Lequesne index — a combined measure of how much knee or hip trouble interferes with daily life — improved too 6.

The reviewers also did something unusually practical: they identified a minimum useful treatment duration of four weeks 6. Shorter than that and you should not expect to be able to tell.

What it does not show

This is osteoarthritis, a specific joint condition, in people who had it. It is not evidence for general inflammation, for a healthy person's background markers, or for other kinds of joint pain. The pool is seven trials, which is thin enough that one or two more could change the picture.

The doses and durations that were actually studied

This is the table to take away. The right-hand column is the one most labels will not tell you.

WhatWhat improvedHow long it took
Curcumin, absorption-enhancedCRP (rated high quality), blood sugar, HDL, weightVaries across 103 trials; dose-related
Omega-3CRP, IL-6, TNF-alpha — all modestlyVaries across 32 meta-analyses
GingerCRP, hs-CRP, TNF-alpha — but trials disagreedVaries across 16 trials
BoswelliaJoint pain, stiffness, function in osteoarthritisAt least 4 weeks

How to read a result like "CRP fell by 0.58"

Three habits make these numbers readable.

Check the units against a normal range. CRP under about 3 mg/L is usually considered unremarkable, and 10 and above typically means something active is going on. A drop of 0.58 mg/L is meaningful if you started at 5 and close to irrelevant if you started at 1. Averages across trials hide where people started.

Check whether the outcome is a marker or a symptom. Three of the four sections above report markers. Boswellia reports symptoms. Symptoms are what you actually notice, and they are much harder to improve, which is why there is less of that evidence everywhere in this field.

Check how much the trials agreed. Reviews report this as I-squared, a percentage. Low means the studies broadly found the same thing; above about 75% means they did not, and the pooled number deserves less weight. This is what separates the ginger result from the curcumin result even though both are "significant".

What this adds up to

If the question is "can a supplement lower inflammatory markers", the answer is a clear yes, for all four of these, across thousands of participants. That is about as settled as supplement research gets, and curcumin's CRP result carries a formal high-quality rating.

If the question is "will that make me healthier", the honest answer is that these trials did not test it. The chain from lower CRP to better outcomes is plausible — the inflammaging literature is built on that link — but plausible is not demonstrated.

So the practical reading is narrow and useful. If you have a specific inflammatory problem with a specific measure attached to it, these have evidence worth taking seriously, and Boswellia for osteoarthritis is the clearest case because it was tested on how knees felt rather than on a blood draw. If you are generally well and want to lower your background inflammation as insurance, you are extrapolating past the data, and the things with much better evidence for that are sleep, exercise, not smoking and dental care.

Keep reading

For the pain side of this, which overlaps heavily, see The Poppy Family Without the Poppy, our deep dive into what botanical painkillers reach at each tier. For why one inflammatory signal became a serious longevity target in animal work, see IL-11 and Lifespan, and for why combinations behave differently from single compounds, Disease Is Multifactorial.

Persistently raised inflammatory markers have causes worth identifying, several of them treatable, and that is a conversation with a doctor rather than a reason to start a supplement. Omega-3 at higher doses affects bleeding time and matters if you take a blood thinner or have surgery coming up.

Back to Top
Added to cart
Your Cart
0
Close

No products in the cart.