Liraglutide improves peripheral perfusion and markers of angiogenesis and inflammation in people with type 2 diabetes and peripheral artery disease: An 18-month follow-up of a randomized clinical trial
- Design
- Randomized trial · 55 participants · Biomarker outcome
- Match to healthy normal-weight adults aged 55–75
- Different population[Auto] Participants had type 2 diabetes (age/BMI not reported in abstract).
- Could weight loss explain it?
- Possibly[Auto] Participants had obesity and/or type 2 diabetes and the abstract does not separate direct drug effects from weight loss or glycaemic improvement.
- Study tier
- Study tier 3[Auto] Small or short randomized trial (auto-provisional).
- Assessment
- Version 1 · automatic, not yet reviewed by a person · Sep 13, 2026
Study facts come from the paper. Population match, weight-loss explanation and study tier are our judgments, made against the reference group of healthy normal-weight adults aged 55–75.
[Auto, unreviewed; quoted from abstract conclusions] In people with type 2 diabetes and PAD, liraglutide increased peripheral perfusion, with amelioration of markers of angiogenesis and inflammation over an 18-month follow-up.
01Findings
What the study reported
- Drugs
- Liraglutide
- Dose
- 1.8 mg
- Comparator
- the control group
- Primary outcome
- Not extracted
- Effect
- Not extracted
- 95% confidence interval
- Not extracted
- Follow-up
- 18 months
- Adverse events
- Not extracted
- Limitations
- Auto-classified from abstract only; effect estimates, adverse events and limitations not extracted. Requires manual review.
Who was studied
- Diabetes status
- type 2 diabetes present (all or most)
- Sample size
- 55
Study quality details
- Study design
- Randomized controlled trial
- Sample size
- 55
- Randomization
- yes
- Blinding
- not stated
- Comparator
- not stated
- Follow up duration
- 18 months
- Outcome type
- biomarker
- Replication
- not assessed (auto)
- Consistency with other evidence
- not assessed (auto)
- Population applicability
- INDIRECT
- Statistical precision
- CI reported: 95% CI
- Funding conflicts
- no
- Peer review status
- yes
01bRisk of bias
How much the result can be relied on
Risk of bias is judged per result, not per paper: the same study can report one marker at low risk and another at high. Two reviewers assess each result independently against a written guide, and every domain judgment below carries its reasoning and the sentences it rests on.
These words are not quality scores. On the RoB 2 scale, High risk of bias is the worst rating a result can receive and Low risk of bias is the best — the opposite of how the same words read on some other scales. Levels are always written out in full here for that reason.
C-reactive protein (high sensitivity), change from baseline vs control · 18 months
Some concerns about risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No domain is High. Three domains are Some concerns (D1 randomization method and concealment undescribed in a small single-centre academic trial; D2 completers-only analysis of an assignment effect in an open-label trial with no placebo injection and unreported by-arm co-intervention intensity; D5 prespecification of the 18-month biomarker result asserted but unverifiable). RoB 2 algorithm: some concerns in at least one domain, none High -> Some concerns.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | The paper states participants "underwent randomization" and reports a registration number, but this is a small single-centre academic trial and neither the sequence-generation method nor any allocation-concealment mechanism is described in the retrieved sources. Under the guide's D1 rule, small academic trials must describe their method or the answer is NI; the v1.2 item-6 parent-trial presumption is not applied, because it is a presumption drawn from the character of large industry registration trials and STARDUST is neither large nor industry-run (the paper is itself the parent trial's extended follow-up). Baseline table shows no imbalance beyond chance (age 67.2 vs 67.4; weight 80.9 vs 81.9; CRP 0.4 vs 0.4), so nothing aggravates the NI, but concealment is unverified. NI on concealment -> Some concerns.The 60 individuals who met the inclusion criteria underwent randomization to liraglutide subcutaneous daily injection ( n = 30) or a matching control group ( n = 30). is a single‐centre, open‐label randomized clinical trial registered on ClinicalTrials.gov ( NCT04881110 ). |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Discontinuation was small and near-balanced (3/30 liraglutide vs 2/30 control), and no adverse events were reported in either arm, so there is no GI-driven differential discontinuation signal. The domain nevertheless does not reach Low: the analysis is completers-only with no ITT or sensitivity analysis stated, so the effect of assignment is estimated by an analysis that excludes randomized participants. The trial is open-label with no placebo injection and the control arm received only "tailored therapeutic prescriptions"; per the guide, awareness alone rarely biases an objective lab biomarker, but the unblinded design also leaves co-intervention intensity (lipid, BP, antiplatelet titration) potentially differential and it is not reported by arm, and CRP is sensitive to such co-interventions.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up: 55 participants completed the follow‐up until December 31, 2022, and were included in the analysis at 6 months. No participants in either group reported adverse events to any medication administered. |
| D3 · Missing outcome data Low risk of bias | 5 of 60 randomized lack the outcome (8.3%, below the 20% threshold); missingness is 10.0% (3/30) vs 6.7% (2/30), a 3.3-point arm difference, below the 5-point threshold. Neither prong trips, and there is no specific contrary evidence (no adverse events reported, so no GI-exposure-linked attrition). Per calibration ruling 2 the thresholds are decisive: Low. The absence of an MNAR sensitivity analysis is a limitations note, not a downgrade.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up: 55 participants completed the follow‐up Parameters Liraglutide (27) Control (28) |
| D4 · Measurement of the outcome Low risk of bias | hs-CRP measured by a named commercial ELISA, in duplicate with a re-run rule, identically in both arms, with laboratory staff and outcome assessors masked to assignment. No indication the assay changed partway through or differed between arms. Per the guide, central-lab hs-CRP is objective and measurement bias is Low.CRP (high sensitivity), TNF‐α, IL‐6 and VEGF‐A (Quantikine; R&D Systems, Minneapolis, USA) were assessed by enzyme‐linked immunosorbent assay (ELISA). Samples were assayed in duplicate and re‐run if duplicates differed by >20%. the laboratory staff did not know the patients' group assignments. The trial staff who assessed outcomes and analysed the data were also masked to group assignment. |
| D5 · Selection of the reported result Some concerns about risk of bias | The paper asserts the biomarker analysis was a pre-specified secondary outcome, and CRP is named in the outcomes section alongside IL-6 and TNF-alpha, with both timepoints (6 and 18 months) and both a null and a positive marker reported - no sign of selective emphasis among analyses. But the protocol/SAP and the registry entry were not retrieved, so prespecification of this specific result (the 18-month timepoint, which is an extension beyond the trial's 6-month primary outcome, and the between-group change-difference model) is unverifiable rather than verified. Under the domain rule, prespecification unverifiable -> at least Some concerns. Calibration ruling 3 (High) does not apply: prespecification is not affirmatively absent, because no searchable protocol or SAP was available to establish absence.we conducted a vascular biomarker analysis as a pre‐specified secondary outcome of the STARDUST trial Change in TcPO 2 at 6 months was one of the primary outcomes of the trial. Here, we report data of people who have reached the 18‐month follow‐up for the coprimary outcome and also for additional secondary outcomes, including markers of inflammation (C‐reactive protein [CRP], tumour necrosis factor α [TNF‐α], interleukin‐6 [IL‐6]) |
Interleukin-6, change from baseline vs control · 18 months
Some concerns about risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No High domain; D1, D2 and D5 are Some concerns for the same trial-level reasons as O1. Algorithm -> Some concerns.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Trial-level, identical to O1: randomization is asserted and the trial is registered (NCT04881110), but sequence generation and allocation concealment are not described anywhere in the retrieved sources, and this is a small single-centre academic trial to which the v1.2 large-industry-trial presumption does not extend. Baseline IL-6 is comparable (44.7 vs 43.9 pg/mL), so there is no rebutting imbalance, but concealment remains NI.The 60 individuals who met the inclusion criteria underwent randomization to liraglutide subcutaneous daily injection ( n = 30) or a matching control group ( n = 30). IL‐6, pg/mL 44.7 (36.9, 63.9) 43.9 (29.4, 60.5) |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Same as O1: near-balanced small attrition (3 vs 2) and no reported adverse events, but the estimate of the assignment effect comes from a completers-only analysis with no ITT or sensitivity analysis, and the open-label design without a placebo injection leaves co-intervention intensity unreported by arm.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up is a single‐centre, open‐label randomized clinical trial |
| D3 · Missing outcome data Low risk of bias | Identical denominators to O1: 8.3% of randomized participants lack the outcome and the arm difference is 3.3 points; neither the 20% nor the 5-point prong trips, so Low by calibration ruling 2.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up: 55 participants completed the follow‐up |
| D4 · Measurement of the outcome Low risk of bias | IL-6 measured by the same named commercial ELISA kit, in duplicate with a re-run rule, identically in both arms, with masked laboratory staff and masked outcome assessors. Objective biomarker, no assay change reported.CRP (high sensitivity), TNF‐α, IL‐6 and VEGF‐A (Quantikine; R&D Systems, Minneapolis, USA) were assessed by enzyme‐linked immunosorbent assay (ELISA). Samples were assayed in duplicate and re‐run if duplicates differed by >20%. the laboratory staff did not know the patients' group assignments. |
| D5 · Selection of the reported result Some concerns about risk of bias | As for O1: IL-6 is named in the outcomes list and prespecification of the biomarker analysis is asserted, and both timepoints are reported, but the protocol, SAP and registry entry were not retrieved, so prespecification of the 18-month result cannot be verified. Unverifiable -> Some concerns; not High, since absence is not affirmatively established.we conducted a vascular biomarker analysis as a pre‐specified secondary outcome of the STARDUST trial additional secondary outcomes, including markers of inflammation (C‐reactive protein [CRP], tumour necrosis factor α [TNF‐α], interleukin‐6 [IL‐6]) |
Tumour necrosis factor alpha, change from baseline vs control (null result) · 18 months
Some concerns about risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No High domain; D1, D2 and D5 Some concerns on the same trial-level grounds. Algorithm -> Some concerns.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Trial-level and identical to O1/O2: randomization asserted, registration given, but no description of sequence generation or allocation concealment in any retrieved source, and the v1.2 presumption is not extended to a small single-centre academic trial. Baseline TNF-alpha medians are comparable (9.1 vs 9.6 pg/mL), though with very wide IQRs in both arms.The 60 individuals who met the inclusion criteria underwent randomization to liraglutide subcutaneous daily injection ( n = 30) or a matching control group ( n = 30). TNF‐α, pg/mL 9.1 (4.1, 88.5) 9.6 (3.8, 79.1) |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Same trial-level facts as O1/O2: small, near-balanced attrition and no reported adverse events, but a completers-only analysis standing in for the assignment effect, with no ITT or sensitivity analysis, in an open-label trial with no placebo injection and no by-arm reporting of co-intervention intensity.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up No participants in either group reported adverse events to any medication administered. |
| D3 · Missing outcome data Low risk of bias | Same denominators: 5/60 (8.3%) missing, 3.3-point arm difference. Neither threshold prong trips -> Low.Three individuals in the liraglutide group and 2 in the control group withdrew from the study and/or were lost during follow‐up: 55 participants completed the follow‐up |
| D4 · Measurement of the outcome Low risk of bias | Same named commercial ELISA platform, duplicate assay with re-run rule, identical across arms, masked laboratory staff and masked assessors. Objective marker; no assay change reported.CRP (high sensitivity), TNF‐α, IL‐6 and VEGF‐A (Quantikine; R&D Systems, Minneapolis, USA) were assessed by enzyme‐linked immunosorbent assay (ELISA). Samples were assayed in duplicate and re‐run if duplicates differed by >20%. The trial staff who assessed outcomes and analysed the data were also masked to group assignment. |
| D5 · Selection of the reported result Some concerns about risk of bias | TNF-alpha is named in the outcomes list on the same footing as CRP and IL-6, and the null result is reported in full at both timepoints in the tables and the text - no evidence of suppression of an unfavourable marker. But, as for O1 and O2, no protocol, SAP or registry entry was retrieved, so prespecification of the 18-month analysis is unverifiable -> Some concerns.No significant differences between groups were found for TNF‐α concentrations. additional secondary outcomes, including markers of inflammation (C‐reactive protein [CRP], tumour necrosis factor α [TNF‐α], interleukin‐6 [IL‐6]) |
02Funding
Funding and conflicts
- Funding
- PhD program of Translational Medicine at the University of Campania "Luigi Vanvitelli", Naples, Italy (Dr. Caruso)
- Industry funded
- No
- Manufacturer
- Novo Nordisk, Eli Lilly, Lilly, AstraZeneca, Sanofi, Roche
- Sponsor role
- no manufacturer funding identified
- Author conflicts
- DG received a consultancy fee from Eli Lilly and has given lectures from Eli Lilly, Sanofi, Novartis, Astrazeneca and Novo Nordisk. MIM has given lectures for Novo Nordisk, Eli Lilly, and Sanofi. KE received a consultancy fee from Eli Lilly and has given lectures from Eli Lilly, Sanofi, Novo Nordisk, Roche, Bayer and Lifescan. All other authors declare that there are no relationships or activities that might bias, or be perceived to bias, their work.
- Independent replication
- unknown
- Notes
- Authors declare relationships with the drug's manufacturer: Novo Nordisk
Funding is shown on every study and never used to score it.
04Source
The source, as retrieved
Abstract
[AIMS] In a six-month randomized clinical trial, improved peripheral perfusion has been shown with liraglutide, associated with favourable vascular effects in people with type 2 diabetes and peripheral artery disease (PAD). We aimed to evaluate the durability of these benefits and to elucidate some mechanisms underlying liraglutide's effect over an 18-month follow-up. [METHODS] STARDUST was a randomized clinical trial which compared liraglutide up to 1.8 mg/day with tailored therapeutic prescriptions to manage cardiovascular risk factors in 55 participants with type 2 diabetes and PAD. We report data of people who have reached the 18-month follow-up for the primary outcome (transcutaneous oxygen pressure, TcPO2) and also for additional secondary outcomes (markers of inflammation, angiogenesis and kidney function), as well as glycemic and metabolic parameters. TcPO2 was assessed with transcutaneous oximetry. Circulating levels of angiogenic progenitor cells and serum inflammation markers were evaluated by flow cytometry and enzyme-linked immunosorbent assay, respectively. [RESULTS] Compared with the control group, significant differences favouring the liraglutide group were observed at 18 months for TcPO2 [estimated treated difference (95% CI), 10.9 mmHg (7.6 to 14.1 mmHg), p < 0.001]. At 18 months of follow-up, participants in the liraglutide group, as compared with those in the control group, had a significant reduction in urine albumin to creatinine ratio (estimated difference, -103.9 mg/g Cr, 95%CI, -170.8 to -37.1, p = 0.003), C-reactive protein (-0.5 mg/dL, 95%CI, -0.8 to -0.2, p = 0.002), as well as interleukin-6 (-32.6 pg/mL, 95%CI, -54.6 to -10.5, p = 0.004). Compared with the control group, participants in the liraglutide group showed significantly higher concentrations of circulating progenitor cells and endothelial progenitor cells at both 6 and 18 months, for CD34+, CD133+, KDR+, CD34+/KDR+ and CD34+/CD133+/KDR+. Liraglutide was also associated with a higher increase in vascular endothelial growth factor A at 18 months (70.1 pg/mL, 95%CI, 44.7 to 95.4, p < 0.001). [CONCLUSIONS] In people with type 2 diabetes and PAD, liraglutide increased peripheral perfusion, with amelioration of markers of angiogenesis and inflammation over an 18-month follow-up.
Where this record came from
| Source | Retrieved | Identifier |
|---|---|---|
| pubmed | Sep 13, 2026 | 40276845 first ingestion |