Yes, and by a measured amount. Pooled across nine randomized trials and 6,239 adults with overweight or obesity, semaglutide cut C-reactive protein by 40.90% (95% CI 35.42–46.37) [2]. Weight fell 12.04% in the same analysis. Those two percentages have different denominators, which is the first thing to get straight.
Buyers pay for pounds. Everything on a product page is priced against them, and the expected-loss tool is the question people actually arrive with. The cleanest available evidence that some of what the drug does is not the pounds sits in a randomized heart-failure program.
Why 40.90% is not three times 12.04%
12.04% is a proportion of the person. 40.90% is a proportion of a blood test result, and which blood test matters enormously. C-reactive protein varies by an order of magnitude between people, swings with any infection or injury, and has a distribution skewed hard to the low end.
A large relative drop from a small starting value is a small absolute change, so the two percentages should not be ranked against each other. Waist circumference in the same pooled trials fell 9.36 cm.
What was analyzed in the randomized program
The two STEP-HFpEF trials randomized 1,145 people with obesity-related heart failure with preserved ejection fraction to semaglutide 2.4 mg or placebo for 52 weeks. This secondary analysis pooled them and sorted participants by baseline C-reactive protein — under 2 mg/L, 2 to under 10, and 10 or above. [1]
Seventy-one per cent had CRP of 2 mg/L or more. Those with more inflammation at baseline were younger, more often women, heavier, felt worse on the symptom score and could walk less far in six minutes.
What it found
The drug’s benefits did not depend on how inflamed somebody was. Improvements in symptoms, physical limitation, body weight, six-minute walk distance and the hierarchical composite endpoint were consistent across all three CRP categories. Every interaction test was nonsignificant.
Semaglutide also lowered CRP more than placebo, whatever the starting level, P for interaction 0.32. And the fall in CRP was similar regardless of how much weight the person lost, P for interaction 0.91.
Why this is the credible version
Claims that these drugs work partly through some route other than weight are everywhere. Most of them rest on studies that cannot possibly show it. Single-arm cohorts where everybody took the drug and everybody lost weight leave no comparison to make.
This is different in one respect that matters: randomization. There is a placebo group, the weight-loss distribution within the treated group is wide, and the CRP change can be compared across it. That is the design a mechanism claim needs. It is still a secondary analysis of a question asked after the trial was run.
Where a real inflammatory disease was measured
Inflammatory bowel disease is the nearest thing to a clinical test of this. Matched one to one across 3,042 patients per arm, tirzepatide patients needed intravenous steroids less often than patients on a GLP-1 [3]. The adjusted hazard ratio was 0.81 (95% CI 0.68–0.94).
Hospitalization, emergency visits and bowel surgery did not move. The comparator pooled the entire GLP-1 class rather than naming a drug, so the winner is a product and the loser is a category.
And the measurement that went the other way
Worth holding beside it. In the tirzepatide imaging substudy this site covered, the reduction in heart muscle mass did track the weight loss. Different drug, different trial, different measurement. The honest summary is that some effects of these drugs appear to follow the weight and some do not. That is more interesting and less tidy than either slogan.
What it is worth to a buyer
Two practical things, both modest. If you are buying and the scale moves less than you hoped, that does not automatically mean nothing else is happening. In this trial, symptom and inflammation benefits appeared across the range of weight loss.
And CRP is a biomarker, not a benefit. Nobody feels better because a laboratory value fell. The outcomes in this trial were symptoms and walking distance; the CRP result explains rather than delivers. That distinction is the same one that separates a modeled saving from money in your pocket. It belongs in the worth-it calculation rather than beside a hard cardiovascular endpoint. None of the heart-failure evidence was measured in people without heart failure, which is nearly everybody reading.