Skip to content
Peptides Codex
ENFR

Use the arrow keys to choose a result, Enter to open it, and Escape to close.

← Home
Materials · models · limits

Semaglutide: evidence brief

Selected primary studies, with the material tested, findings and limitations kept together. Follow each source to understand what it establishes.

Semaglutide: evidence brief

Peptides Codex

What was actually studied

Three randomized placebo-controlled trials address different questions: weight change, cardiovascular events and a kidney/cardiovascular composite. Interpret each result with its eligibility criteria, analysis population, follow-up and treatment limitations.

Version
2026-10-05.1
Snapshot updated

Prepared by Peptides Codex. Independent review by a qualified subject specialist is pending.

Scope of this brief

Selected STEP 1, SELECT and FLOW primary reports of subcutaneous trial semaglutide. These Novo Nordisk-funded trials used different protocols and populations; their results are not interchangeable. This is not an exhaustive or systematic review, a dosing guide or authentication of oral, compounded or vendor-labelled preparations.

  1. Wilding et al., 2021 · STEP 1 · PMID 33567185

    Material tested
    Trial semaglutide solution for subcutaneous injection, a modified GLP-1 analogue; historical target arm 2.4 mg versus matching placebo. Both groups received lifestyle support.
    Model or population
    Adults without diabetes: BMI ≥30, or ≥27 with a weight-related condition.
    Study design
    Multicenter randomized, double-blind phase III trial; 2:1 allocation and 68-week weight endpoints.
    Sample and comparison
    1,961 randomized and in the full analysis set: 1,306 semaglutide, 655 placebo. Week-68 weight observations: 1,212 and 577.
    Reported finding
    Coprimary endpoints: modelled weight change −14.9% versus −2.4%; adjusted difference −12.4 percentage points (95% CI −13.4 to −11.5). ≥5% loss was observed in 1,047/1,212 (86.4%) versus 182/577 (31.5%).
    Limits of this evidence
    The treatment-policy estimate includes discontinuation/rescue and imputed missing data; threshold percentages use observed participants. This row does not establish durability after stopping treatment. Gastrointestinal-event discontinuations: 4.5% versus 0.8%.

    Source access: Primary abstract, indexed journal methods/results and registry results inspected; direct journal full text unavailable.

    Primary source ↗: Wilding et al., 2021 · STEP 1 · PMID 33567185

  2. Lincoff et al., 2023 · SELECT · PMID 37952131

    Material tested
    Trial semaglutide for subcutaneous injection, a modified GLP-1 analogue; historical target arm 2.4 mg versus placebo, added to standard care.
    Model or population
    Age ≥45, BMI ≥27 and established cardiovascular disease; no history of diabetes.
    Study design
    Multicenter randomized, double-blind, event-driven superiority trial; mean follow-up 39.8 months.
    Sample and comparison
    17,604 randomized and analyzed as assigned: 8,803 semaglutide, 8,801 placebo.
    Reported finding
    Primary cardiovascular death/nonfatal heart attack/nonfatal stroke composite: 569/8,803 (6.5%) versus 701/8,801 (8.0%); hazard ratio 0.80 (95% CI 0.72–0.90). That is 20% lower hazard; the observed proportions differ by 1.5 percentage points.
    Limits of this evidence
    Secondary prevention, not a population without established cardiovascular disease. Adverse-event discontinuations: 16.6% versus 8.2%. Cardiovascular death alone missed the hierarchical significance threshold; the composite does not establish benefit for every component.

    Source access: Primary abstract, indexed journal methods/results and registry results inspected; direct journal full text unavailable.

    Primary source ↗: Lincoff et al., 2023 · SELECT · PMID 37952131

  3. Perkovic et al., 2024 · FLOW · PMID 38785209

    Material tested
    Trial semaglutide for subcutaneous injection, a modified GLP-1 analogue; historical trial arm 1.0 mg versus placebo, with standard care.
    Model or population
    Adults with type 2 diabetes and albuminuric chronic kidney disease; eGFR 25–75 mL/min/1.73 m² with trial-specific albuminuria thresholds.
    Study design
    Randomized, double-blind phase III trial; median follow-up 3.4 years. Stopped early after a prespecified interim analysis.
    Sample and comparison
    3,533 randomized and in the full analysis set: 1,767 semaglutide, 1,766 placebo.
    Reported finding
    Primary kidney failure/persistent ≥50% eGFR decline/kidney or cardiovascular death composite: 331/1,767 versus 410/1,766; hazard ratio 0.76 (95% CI 0.66–0.88), or 24% lower hazard. Observed proportions: 18.7% versus 23.2%, a 4.5-point difference.
    Limits of this evidence
    The composite includes cardiovascular death, not kidney failure alone. Selected diabetes/kidney-disease population; early stopping limits longer follow-up. The 4.5-point difference is an unadjusted events/randomized calculation, not a survival-adjusted estimate or personal risk prediction.

    Source access: Primary abstract and trial-registry methods/results inspected; direct journal full text unavailable.

    Primary source ↗: Perkovic et al., 2024 · FLOW · PMID 38785209

A finding applies to the material, model and conditions studied. It does not establish safety or effectiveness in other settings.

Help review this exact version

Qualified scientific review is pending. Source-backed corrections are welcome; a correction submission does not itself constitute expert approval.

Review reference: semaglutide · 2026-10-05.1

What a reviewer should check
  1. Does the reported material match the preparation actually studied?
  2. Do population, participant counts and analysis denominators match the source?
  3. Are estimates, endpoints, uncertainty and missing-data methods represented accurately?
  4. Do limitations, source-access boundaries and English/French text preserve the same conclusion?

Study records in this version

In your message, include this brief reference, the study identifier, the source citation and the proposed change. If offering scientific review, include your relevant specialty and conflicts of interest. Omit patient information and confidential research material.

Recording a qualified review requires verified qualifications, an agreed scope and approval of the identified version. Public reviewer attribution requires separate permission. Review status remains pending until that process is complete.

Send source-backed feedback →

To save this brief as a PDF, use your browser’s Print command and choose Save as PDF. The study details and source URLs are included in the print layout.