Semaglutide proteomics: 16 studies reviewed

A new review published in the Journal of Endocrinological Investigation in July 2026 compiled 16 proteomics studies^[1] that tracked what semaglutide does to proteins across human blood, fat tissue, heart muscle, kidney, lung, and other tissues. The short answer is that the drug does more at the protein level than earlier clinical-trial summaries captured. The longer answer is worth understanding if you follow GLP-1 research closely.

What proteomics adds to the picture

Standard clinical trials measure outcomes: weight lost, HbA1c dropped, cardiovascular events avoided. Proteomics asks a different question: what proteins changed, and in which direction? The 16 studies in this review^[1] used platforms including SomaScan, TMT-LC-MS/MS, and data-independent acquisition (DIA) proteomics to measure thousands of proteins simultaneously across serum, plasma, adipose tissue, myocardium, aorta, hippocampus, and skeletal muscle.

The result is a catalogue of biological pathways that move when semaglutide is present. Three pathways came up repeatedly across the included studies: PPAR signalling (which governs fat storage and fatty acid burning), oxidative phosphorylation (the core of mitochondrial energy production), and fatty acid metabolism itself. Those three are consistent with what the GLP-1 receptor does pharmacologically, but seeing them confirmed at the protein level across multiple tissue types and multiple species adds evidence that these are genuine systemic effects rather than artefacts of a single study design.

Tissues and pathways beyond the expected

The review also flagged effects in tissue types that are not the primary focus of diabetes or weight-loss labelling. Extracellular matrix (ECM) remodelling appeared as a recurring theme in cardiac and vascular tissue. Complement and inflammatory cascade proteins shifted in several studies. mTORC1 signalling, a nutrient-sensing pathway with relevance to muscle mass and ageing, was also identified. A separate 2026 systems-medicine review^[2] noted that circulating proteins including adiponectin, fibroblast growth factor 21, and apolipoprotein C-III change with semaglutide treatment, as do matrix metalloproteinases involved in tissue remodelling.

These findings sit alongside, not instead of, the established clinical evidence. Semaglutide is authorised by the EMA as Ozempic for type-2 diabetes^[3] and as Wegovy for chronic weight management^[4]. The randomised-trial evidence for those two indications remains the primary basis for prescribing decisions. Proteomics research is mechanism science, not a signal that the drug does something new in a clinical sense.

Human data versus animal data

Of the 16 included studies, four used human participants (population-based designs) and 12 used animal models or in vitro experiments. That split matters. Animal and cell-culture proteomics can run at much higher peptide concentrations than a patient ever receives, can sample tissues that are impractical to biopsy in humans, and can control variables that clinical studies cannot. But the dose and tissue context are also not identical to clinical use.

The authors are direct about this limitation: small sample sizes in the human studies, significant variation between proteomic platforms, and high subject heterogeneity all limit how confidently findings can be generalised. Four human studies is a thin base for strong conclusions about protein-level effects in the broader population of people using semaglutide.

What we do not yet know

Proteomics research on semaglutide is at an early mapping stage. The field can now say with reasonable confidence which pathways shift and in which direction, but cannot yet say which protein changes drive clinical outcomes versus which are downstream markers of weight loss or glucose improvement that would occur with any effective treatment. Disentangling semaglutide-specific effects from generic caloric-restriction effects in human studies is technically difficult and has not been systematically addressed in the literature reviewed.

There are also open questions about tissue accessibility. Much of the most interesting proteomics data comes from cardiac and hepatic tissue in animal models, which cannot be collected non-invasively in humans at scale. Plasma and serum proteomics are more tractable but capture a filtered, systemic signal rather than local tissue effects.

Why this matters for people following GLP-1 research

If you are a patient or clinician tracking semaglutide for its labelled indications, this review does not change the practical picture. The drug's efficacy and safety profile are established by the large randomised trials. What the proteomics layer adds is a growing mechanistic explanation for why GLP-1 receptor agonists appear to have effects in cardiovascular, renal, and possibly neurocognitive tissue, effects that the SELECT and FLOW trials have started to quantify clinically but that were not fully explained by glucose control or weight loss alone.

For researchers, the review is a useful map of where the gaps are. More human-tissue proteomics studies, with larger samples and standardised platforms, would substantially improve confidence in the pathway picture. Until then, the catalogue is suggestive rather than definitive.

Context: the broader semaglutide evidence base

Proteomics sits alongside a large and well-established clinical evidence base. The STEP-1 obesity trial reported a mean body-weight reduction of 14.9% at 68 weeks in adults using semaglutide 2.4 mg plus lifestyle support. The SUSTAIN-6 cardiovascular-outcomes trial reported a statistically significant reduction in major adverse cardiovascular events in patients with type-2 diabetes at elevated cardiovascular risk. The SELECT trial (2023) extended that cardiovascular signal to adults with overweight or obesity without diabetes. These are the findings that inform prescribing guidance. The proteomics layer described in this review is a mechanistic complement to that evidence, not a replacement for it. Understanding why a drug works helps researchers design better follow-on compounds and identify patient populations who may respond differently, but the clinical trial data still defines what is authorised and what is recommended.

For a full summary of the randomised-trial evidence and the regulatory status across jurisdictions, see the semaglutide peptide page, which covers the STEP and SUSTAIN programmes, the SELECT cardiovascular-outcomes data, and the current prescription status across the EU, UK, and US.