GLP-1 Drugs Lower an Inflammation Marker Fast. What That Means for Your Body Is Less Clear

A clue that showed up early

SELECT was a big trial. It enrolled 17,604 people who had heart disease and overweight or obesity, but no diabetes. Half got a weekly shot of semaglutide. Half got a placebo.2

Researchers also tracked a blood test called hs-CRP. CRP (C-reactive protein) is a marker of inflammation, and hs-CRP is a more sensitive version of the test.1 People who started the trial with higher hs-CRP had more heart attacks, strokes and deaths later on.1

Semaglutide lowered hs-CRP by 37.8% at 104 weeks.1 Then came the part that caught people's eye. The drop showed up by 4 and 8 weeks, before the big weight loss. It also showed up in people who did not lose weight.1

That is the clue behind the "anti-inflammatory wonder drug" idea. If CRP falls before the scale moves, maybe the drug acts on inflammation directly. The clue is real. What it proves is the open question.

The CRP drop holds up across trials

SELECT is not a one-off. A 2025 meta-analysis, which pools the results of many studies, combined 52 randomized trials in 4,734 people with type 2 diabetes. Compared with placebo or other diabetes drugs, GLP-1 drugs lowered CRP. They also lowered IL-6 (interleukin-6, another inflammation signal) and TNF-alpha.5

The CRP drop showed up by 4 and 8 weeks, before the big weight loss.From a 2026 randomized trial in Circulation

A 2026 meta-analysis of 41 trials was more careful. CRP came down across 27 studies with 1,991 people. But the changes in IL-6 and TNF-alpha were too small to rule out chance, and results varied a lot from trial to trial.6

Tirzepatide acts on GLP-1 and a second gut hormone called GIP. Across six studies, hs-CRP fell 32.9 percentage points more than with placebo, and IL-6 fell 17.8 points more.7 The 2026 review, which looked only at people with diabetes, found a clear IL-6 drop in two tirzepatide trials but only a trend for CRP.6

In a 52-week trial of 529 people with obesity and a type of heart failure called HFpEF, CRP fell 43.5% on semaglutide and 7.3% on placebo.8

How much of it is weight loss?

This is where the story gets harder. In SELECT, people who lost more weight on semaglutide also had bigger CRP drops.1 So weight loss is part of it. But the early drop, and the drop in people who lost no weight, suggest it is not the whole story.1

The heart results point the same way. A 2025 look at SELECT found fewer heart events on semaglutide at every starting weight. In the drug group, the amount of weight lost by week 20 did not track with who did better later.4 The authors estimated that 33% of the benefit came through a smaller waist.4

A 2026 study tried to work out where the benefit came from. On paper, the hs-CRP drop accounted for 42.1% of it. But the range of likely values ran from 17.9% to 110.5%, far too wide to pin down.3 The authors said the benefit could not be tied to known risk factors "with any certainty."3

Hearts: the strongest evidence

Whatever the reason, the heart result itself is solid. In SELECT, a heart attack, stroke or heart death happened to 569 of 8,803 people on semaglutide (6.5%). On placebo, it was 701 of 8,801 (8.0%). People were followed for 39.8 months on average.2 That is a 20% lower risk.3

A 2025 meta-analysis of 11 large trials found that GLP-1 drugs lowered major heart events by 13% in people with type 2 diabetes.12

There is a cost. In SELECT, 16.6% of people on semaglutide quit the drug because of side effects, compared with 8.2% on placebo.2

Joints and brains: a weaker case

Knees first. In a 68-week trial of 407 people with obesity and knee arthritis, people on semaglutide lost 13.7% of their weight, versus 3.2% on placebo. Their pain score, on a scale of 0 to 100, fell 41.7 points, versus 27.5.9 That is a real gain. But the trial was built on the idea that losing weight eases knee pain.9 It shows less pain, not proof of an anti-inflammatory effect in the joint.

Brains are the sharpest test. Two trials, called evoke and evoke+, gave 3,808 people with early Alzheimer's disease a daily semaglutide pill of up to 14 mg. After 104 weeks, the drug did not slow the disease.10 Yet it clearly lowered hs-CRP.11 The trial leaders' reading: lowering inflammation in the body "may be insufficient" to help in Alzheimer's.11 They left one door open. People with high inflammation, or drugs that reach the brain better, might still respond.11

The bottom line

The CRP drop is real. It came within weeks in SELECT, showed up across dozens of trials, and was large in heart failure. The heart benefit is real too, and some of it may run through calmer inflammation. But a knee trial that eased pain was a weight-loss trial, and an Alzheimer's trial that lowered CRP did not slow the disease. These are strong drugs with a solid heart record in the people who were studied. They are not proof that lowering one marker helps every organ. A falling CRP is good news on a lab report. It is a clue, not a verdict.