A prespecified secondary analysis of the SELECT trial, published in Circulation on 18 August 2026, reports that semaglutide reduced high-sensitivity C-reactive protein by 37.8 percent at 104 weeks. The number that matters more is when it happened: the authors report reductions evident by weeks 4 and 8, before major weight loss, and present in participants who lost no weight at all.

That timing is what separates this from a routine biomarker readout. If inflammation falls before the weight does, weight loss is not the whole mechanism.

A framing note: semaglutide is an approved medicine and SELECT studied patients. Peptra Labs supplies compounds in this class as reference material for laboratory research only, and nothing here is guidance about human use.

What SELECT tested, and what the semaglutide substudy asked

SELECT enrolled 17,604 patients with known atherosclerotic cardiovascular disease and overweight or obesity, but not diabetes. Randomisation to semaglutide significantly reduced the primary outcome of major adverse cardiovascular events, defined as cardiovascular death, non-fatal myocardial infarction or non-fatal stroke, compared with placebo, over a mean follow-up of 39.8 months. The trial is registered as NCT03574597.

This substudy was prespecified, which is worth stating plainly because it is the opposite of the analysis we covered for a different compound yesterday. The authors set out to evaluate whether baseline hsCRP predicted event risk, and how changes in hsCRP related to time to first event, to baseline body weight, to weight loss, and to other clinical measures, across treatment groups over 104 to 208 weeks, using Cox modelling among other approaches.

What the analysis found

Baseline inflammation predicted events. Baseline hsCRP was similar in the two arms, at a geometric mean of 1.96 mg/L on semaglutide and 1.91 mg/L on placebo. The authors report that it was prognostic of future events, with risk rising across the subgroups below 2, 2 to under 10, and 10 mg/L or above, including significant associations with cardiovascular death and all-cause death.

Semaglutide lowered hsCRP. The reduction was 37.8 percent at 104 weeks, and the authors report that event risk was reduced across all hsCRP subgroups.

The reduction did not depend on weight loss. Greater reductions in hsCRP relative to baseline were associated with greater weight loss, but the authors report that the hsCRP changes preceded major weight loss, being evident by weeks 4 and 8, and occurred among participants who did not lose weight.

It was independent of the usual confounders. The authors report that semaglutide-associated hsCRP changes were independent of LDL cholesterol, statin use, and the cardiovascular entry criteria.

Falling hsCRP tracked with fewer events. hsCRP reductions were prognostic of decreased event risk, and the authors state that modelling suggests decreased inflammation contributed in part to the benefit seen in the trial.

What hsCRP actually measures

C-reactive protein is an acute-phase protein made in the liver, largely in response to interleukin-6. The high-sensitivity assay is the same molecule measured with enough precision to resolve the low concentrations that matter in cardiovascular risk work, rather than the high ones seen in acute infection.

It is a downstream summary, not a pathway. A change in hsCRP tells you that systemic inflammatory signalling has shifted, and does not tell you which signal moved or where it came from. That limitation is worth carrying through the rest of this article.

Why the timing is the finding

The obvious explanation for anti-inflammatory effects in this drug class is that losing a large amount of body fat reduces inflammation, because adipose tissue is itself an inflammatory organ. That explanation is almost certainly part of the story.

What the timing data argue is that it cannot be the whole story. A change visible at week 4, in people who have not yet lost meaningful weight, and present in people who never lose weight, needs a mechanism that does not route through fat mass.

The authors do not claim to have identified that mechanism. Their conclusion is carefully hedged: the findings suggest that the event reduction observed with semaglutide in SELECT may have partially involved a decrease in inflammation. Suggest, may, and partially, all in one sentence.

How this sits next to the retatrutide data

We reported a post hoc look at retatrutide and cardiovascular biomarkers yesterday, in which hsCRP fell 54.8 percent in one of two phase 2 trials and did not reach significance in the other.

The two papers are not comparable in weight, and the difference is instructive. The retatrutide analysis was post hoc, in two phase 2 trials, with biomarkers as the endpoint. This one is prespecified, in a 17,604-patient outcome trial, with the biomarker analysed against actual cardiovascular events that the trial was powered to detect.

Both are evidence that this class moves inflammatory markers. Only one of them can say anything about whether that matters for outcomes.

The distinction between prespecified and post hoc is doing real work here, and it is the single most useful thing to check when a biomarker result appears in coverage. Prespecified means the question was written into the analysis plan before anyone saw the data. Post hoc means it was asked afterwards. Both can be honest and well executed. Only the first constrains how many other questions were tried first.

What it still does not show

A prespecified secondary analysis of a positive trial is strong evidence, and it is still not a mechanistic demonstration.

The analysis is associative in structure. hsCRP fell, events fell, and modelling suggests a partial contribution. It does not establish that lowering inflammation caused the event reduction, which would require an intervention targeting inflammation alone.

hsCRP is also a single marker, a downstream acute-phase reactant rather than a specific inflammatory pathway. Ridker, one of the authors, has spent a career establishing hsCRP as a usable risk marker, and its usefulness as a marker is not the same as it being the thing that changed.

For laboratories tracking where this class is going, our ranking of the most-studied research peptides follows the published evidence per compound, and we covered the same molecule in a different setting in a 52-week biopsy trial in MASH.

References

The products referenced on this site are supplied for laboratory research use only. They are not medicines and are not intended for human or veterinary use. This article summarises published research for informational purposes and is not medical advice. Statements about third-party studies belong to their authors.

Newer
GLP-1 and Mitochondria: A 2026 Review Maps the Axis
Back to list
Older MOTS-c Neutralised MRSA in Mice: A 2026 eLife Report

Related Posts