Peptides

The short version: the evidence here is almost entirely about semaglutide, a GLP-1 receptor agonist peptide drug for diabetes and weight loss, and it is almost entirely funded by the manufacturer. The burden of proof for any peptide intervention on a healthy body has not been met.

The evidence you retrieved is dominated by the SUSTAIN and PIONEER trial programs for semaglutide, a synthetic peptide that mimics a natural gut hormone [13, 14, 21]. These are large, randomized, placebo-controlled trials. They show semaglutide lowers HbA1c by about 1.5 percentage points and produces weight loss of 3.5 to 6.5 kg over 30 to 56 weeks [14, 21]. The SUSTAIN 6 trial, which ran for 104 weeks in 3,297 patients, found a 26% reduction in a composite of cardiovascular death, non-fatal heart attack, and non-fatal stroke (hazard ratio 0.74, 95% CI 0.58 to 0.95) [19]. That is a real clinical endpoint, not a surrogate.

But the safety picture is not clean. The same SUSTAIN 6 trial found a signal for diabetic retinopathy complications: 3.0% in the semaglutide group versus 1.8% in the placebo group (p = 0.02) [19]. The authors attribute this to rapid blood sugar lowering, not a direct toxic effect, but the signal is there. Gastrointestinal side effects are common: nausea in 20-24% of patients, diarrhea in 11-13%, and discontinuation due to adverse events in 5-13% across trials [16, 21]. A 2025 real-world VA study of nearly 22,000 veterans found broadly similar kidney and cardiovascular outcomes across GLP-1 drugs, but with a potential mortality signal favoring liraglutide over dulaglutide and occasional gallstone signals with semaglutide [22].

The other peptides in the evidence are mostly preclinical or niche. Thymosin alpha 1 was tested in a single 120-patient trial for tuberculosis with diabetes, showing improved immune markers and culture conversion, but this is a single small study in a specific disease population [26]. The gramicidin-inspired peptide LD8Δ is purely in vitro and in animal models for kidney cancer [27]. The thrombospondin mimetic ABT-510 had a half-life of about 1.1 hours in humans and showed only prolonged stable disease in some cancers in phase I trials [28]. None of these are ready for general use.

The fundamental problem is that every major semaglutide trial was funded by Novo Nordisk, the manufacturer [13, 14, 21]. The SUSTAIN and PIONEER programs are well-designed, but they are manufacturer-funded. The long-term safety data beyond two years is thin: the longest trial in this evidence is 104 weeks [19]. The retinopathy signal is real and not fully explained. For a healthy person without diabetes or obesity, there is no evidence here that any peptide intervention provides net benefit. The burden of proof for injecting a synthetic peptide into a healthy body remains unmet.

My call: for people with type 2 diabetes and obesity, semaglutide shows real benefits on hard clinical outcomes, but with a meaningful side effect burden and a retinopathy signal that needs monitoring. For anyone else, the evidence does not support use. Confidence: moderate for semaglutide in diabetes, low for any other peptide in any other population.