Effects and plasma proteomic analysis of GLP-1RA versus CPA/EE, in combination with metformin, on overweight PCOS women: a randomized controlled trial
- Design
- Randomized trial · Mixed outcome
- Match to healthy normal-weight adults aged 55–75
- Different population[Auto] Participants had obesity/overweight (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] Both CPA/EE+Met and GLP-1RA + Met treatment improved reproductive functions in overweight PCOS women. GLP-1RA + Met was more effective than CPA/EE + Met in reducing body weight, BMI, and waist, and improving metabolism, and ovulation in overweight women with PCOS, with acceptable short-term side effects. CPA/EE + Met was more effective in reducing hyperandrogenemia. The novel plasma biomarkers PRDX6, FN1, and SERPINB9, might be indicators and targets for PCOS treatment. TRIAL REGISTRATION CLINICALTIALS.
01Findings
What the study reported
- Drugs
- Liraglutide
- Dose
- 2 mg
- Treatment duration
- 12 weeks
- Primary outcome
- Not extracted
- Effect
- Not extracted
- 95% confidence interval
- Not extracted
- Follow-up
- 12 weeks
- Adverse events
- Not extracted
- Limitations
- Auto-classified from abstract only; effect estimates, adverse events and limitations not extracted. Requires manual review.
Who was studied
- Obesity status
- obesity/overweight present (all or most)
Study quality details
- Study design
- Randomized controlled trial
- Sample size
- not extracted
- Randomization
- yes
- Blinding
- open-label
- Comparator
- not stated
- Follow up duration
- 12 weeks
- Outcome type
- mixed
- Replication
- not assessed (auto)
- Consistency with other evidence
- not assessed (auto)
- Population applicability
- INDIRECT
- Statistical precision
- not extracted
- 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.
CRP (mg/L), between-arm difference in change from baseline to 12 weeks (delta-CRP) · 12 weeks
High risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No domain reaches High. Some concerns in D1 (concealment unreported), D2 (open-label, completers-only, no ITT), D3 (arm-level missingness unreported) and D5 (prespecification unverified, outcome outside the stated primary/secondary lists) accumulate but none of them, alone or together, gives a specific reason to think the null delta-CRP contrast is substantially wrong. Note separately that the comparator is an active drug (CPA/EE) that itself acts on inflammatory markers, which biases the contrast toward null; that is an indirectness/stratum matter under c001-outcomes.md v1.2, not a RoB 2 domain.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Sequence generation is described (table of random numbers) and the authors state the arms were matched at baseline, but nothing is reported about allocation concealment (no sealed envelopes, central allocation, or third-party sequence holder). Item 6's parent-trial presumption does not apply: this is a small single-centre academic trial and its own parent. Concealment is therefore NI, which under the RoB 2 algorithm gives Some concerns.The participants of the study were randomly divided into these two groups using a table of random numbers (Fig. 1 ). The baseline clinical characteristics of the two randomized treatment groups were matched. |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Judged per the guide on discontinuation imbalance and ITT handling, not on unblinding alone (this is an open-label trial with an active comparator, and CRP is a lab biomarker for which awareness rarely biases measurement). The problem is the analysis set: 10 of 70 eligible participants were excluded or dropped out and the analysis is of the 60 completers only, with no arm-level breakdown of who left and no ITT or intention-to-treat statement anywhere. GI adverse events were concentrated in the GLP-1RA arm (diarrhoea/nausea/vomiting 13.33% each) versus weight gain and liver enzyme rises in the comparator arm, so a differential, treatment-related exit from the analysis set cannot be excluded. Not High: no evidence of a large imbalance, and the result is a null.A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. Given the special administration method (I.H.) with a GLP-1RA, participants and investigators were not blinded to the treatment in this trial. |
| D3 · Missing outcome data Some concerns about risk of bias | Missingness relative to the 70 eligible/randomized is 10/70 = 14.3%, below the guide's 20% prong, so that prong does not trip. But the second prong cannot be evaluated: the paper gives no per-arm breakdown of the 10 exclusions/dropouts (the flow numbers sit in Fig. 1, which is not rendered in the retrieved text), so the between-arm difference in missingness is NI rather than shown to be within 5 points. CRP was measured only in completers, there is no sensitivity analysis, and dropout for GI intolerance would correlate with drug exposure. With one prong untestable, the calibration ruling that makes the thresholds decisive cannot deliver Low here.Eighty(80%) patients completed the study. Among them, 70 women matched the inclusion criteria. A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. |
| D4 · Measurement of the outcome Low risk of bias | CRP is an objective laboratory biomarker measured on the same schedule in both arms (fasting samples on days 3-5 of the cycle or the follicular phase), with kits and normal ranges specified. No indication the assay changed partway or differed between arms. Per the guide, measurement bias for such markers is Low; open-label status does not reach an assay readout.Blood samples were collected while the participants were in a fasted state on the 3rd to 5th day of the menstrual cycle or during the follicular stage. All the above indicators were measured before and after treatment. Experimental normal ranges and kit suppliers are listed in Supplementary Table 2 . |
| D5 · Selection of the reported result High risk of bias | Item 7, second limb. The stated primary outcomes are reproductive hormones plus glucose and lipid metabolism, and the stated secondary outcomes are anthropometric changes plus proteomics; inflammatory markers appear in neither list, yet the paper makes no statement that the inflammatory analyses were exploratory, post hoc, or unadjusted for multiplicity. The High limb of item 7 is triggered by an author statement of that kind, and there is none; ruling 3's High is for prespecification affirmatively absent from a fully searchable protocol, and here neither the protocol/SAP nor the NCT03151005 registry record was retrieved. Omission from the paper's own outcome list is not the same object as a searched protocol, so prespecification is unverified rather than affirmatively refuted, and D5 sits at the Some concerns floor. Some concerns is reached comfortably, not marginally: the markers are reported in an unlabelled "Inflammatory indicators" block alongside IL-8 and WBC with no multiplicity handling, and the paper leans on the within-group IL-6 fall in the abstract while the assessed between-arm contrast is null.The primary outcomes were changes in reproductive hormone levels along with glucose and lipid metabolism associated with obesity. The secondary outcomes included anthropometric changes as well as plasma proteomics analyses related to PCOS. P values < 0.05 were considered statistically significant. |
IL-6 (pg/ml), between-arm difference in change from baseline to 12 weeks (delta-IL-6) · 12 weeks
High risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: Some concerns in D1, D2, D3 and D5, none High. The selective-emphasis problem is real but transparent: the between-arm null is printed in Table 2 and acknowledged in the Results text, so the assessed result itself is recoverable and not distorted. As with O1, the active CPA/EE comparator's own inflammatory effects push this contrast toward null and are handled outside RoB 2.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | As O1: sequence generation described (random number table), allocation concealment NI, no parent-trial presumption available to this small single-centre academic trial.The participants of the study were randomly divided into these two groups using a table of random numbers (Fig. 1 ). |
| D2 · Deviations from intended interventions Some concerns about risk of bias | As O1: open-label with active comparator, completers-only analysis with no ITT statement and no arm-level accounting of the 10 exclusions/dropouts, against a GI-adverse-event profile concentrated in the liraglutide arm.A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. Given the special administration method (I.H.) with a GLP-1RA, participants and investigators were not blinded to the treatment in this trial. |
| D3 · Missing outcome data Some concerns about risk of bias | As O1: 14.3% missing (below the 20% prong) but the per-arm difference is unreported and so untestable against the 5-point prong; no sensitivity analysis for value-dependent missingness.Among them, 70 women matched the inclusion criteria. A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. |
| D4 · Measurement of the outcome Low risk of bias | IL-6 is an objective assayed biomarker, measured before and after treatment in both arms on the same fasting schedule with kits and reference ranges documented; no evidence of assay change or between-arm measurement difference.All the above indicators were measured before and after treatment. Experimental normal ranges and kit suppliers are listed in Supplementary Table 2 . |
| D5 · Selection of the reported result High risk of bias | Item 7, second limb, as for O1: IL-6 is in neither the stated primary nor the stated secondary outcome list, and the paper makes no exploratory/post hoc/multiplicity statement; protocol, SAP and registry are unretrieved, so prespecification is unverified, not affirmatively refuted. IL-6 additionally attracts selective-emphasis concern: the abstract and Results highlight the within-group IL-6 fall and name IL-6 among the parameters on which GLP-1RA+Met "was more effective", while the assessed between-arm contrast in Table 2 is flatly null (P=0.877) and the Results text concedes it. This is within-group versus between-group framing of the same marker, which is the same shape as the guide's "multiple transformations with one chosen for emphasis" trigger. It does not rise to High under item 7 because the High limb requires an author statement of non-prespecification, which is absent.The primary outcomes were changes in reproductive hormone levels along with glucose and lipid metabolism associated with obesity. The secondary outcomes included anthropometric changes as well as plasma proteomics analyses related to PCOS. GLP-1RA + Met was more effective than CPA/EE + Met in reducing body weight, BMI (Body Mass Index), and waist circumference, FBG(fasting blood glucose), AUCI(area under curve of insulin),TC (Total Cholesterol), IL-6(Interleukin-6) and improving insulin sensitivity |
TNF-alpha (pg/ml), between-arm difference in change from baseline to 12 weeks (delta-TNF-alpha) · 12 weeks
High risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: Some concerns in D1, D2, D3 and D5, none High; no specific reason to believe the null delta-TNF-alpha contrast is substantially wrong, though it is very imprecise (n=30 per arm, large SDs). Active-comparator bias toward null is carried outside RoB 2.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | As O1: random number table for sequence generation, allocation concealment NI, no parent-trial presumption available.The participants of the study were randomly divided into these two groups using a table of random numbers (Fig. 1 ). |
| D2 · Deviations from intended interventions Some concerns about risk of bias | As O1: open-label active-comparator design analysed in completers only, with no ITT statement and no arm-level dropout accounting.A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. Given the special administration method (I.H.) with a GLP-1RA, participants and investigators were not blinded to the treatment in this trial. |
| D3 · Missing outcome data Some concerns about risk of bias | As O1: overall missingness 14.3% does not trip the 20% prong, but the per-arm split is unreported so the 5-point prong is untestable, and no sensitivity analysis addresses value-dependent missingness.Among them, 70 women matched the inclusion criteria. A total of 10 participants were excluded or dropped out from further analyses, and the remaining 60 patients completed the study. |
| D4 · Measurement of the outcome Low risk of bias | TNF-alpha is an objective assayed biomarker measured identically in both arms before and after treatment; no assay change or between-arm measurement difference reported.All the above indicators were measured before and after treatment. Experimental normal ranges and kit suppliers are listed in Supplementary Table 2 . |
| D5 · Selection of the reported result High risk of bias | Item 7, second limb, as for O1. TNF-alpha is in neither stated outcome list, there is no exploratory/post hoc/multiplicity statement, and no protocol, SAP or registry record was retrieved to verify prespecification, so D5 rests at the floor. Note in passing the odd within-arm variance pattern (baseline TNF-alpha SD 9.83 post-treatment in the CPA/EE arm in Table 1 versus a delta SD of 0.70 in Table 2), which the paper does not reconcile; that is a reporting-quality note, not a D5 trigger.The primary outcomes were changes in reproductive hormone levels along with glucose and lipid metabolism associated with obesity. The secondary outcomes included anthropometric changes as well as plasma proteomics analyses related to PCOS. |
02Funding
Funding and conflicts
- Funding
- Chongqing Natural Science Foundation; National Science Fund for Distinguished Young Scholars
- Industry funded
- No
- Manufacturer
- None identified
- Sponsor role
- no manufacturer funding identified
- Author conflicts
- The authors declare no competing interests.
- Independent replication
- unknown
Funding is shown on every study and never used to score it.
04Source
The source, as retrieved
Abstract
[PURPOSE] Polycystic ovary syndrome (PCOS) is characterized by reproductive dysfunctions and metabolic disorders. This study aims to compare the therapeutic effectiveness of glucagon-like peptide-1 receptor agonist (GLP-1RA) + Metformin (Met) versus cyproterone acetate/ethinylestradiol (CPA/EE) + Met in overweight PCOS women and identify potential proteomic biomarkers of disease risk in women with PCOS. [METHODS] In this prospective, open-label randomized controlled trial, we recruited 60 overweight PCOS women into two groups at a 1:1 ratio to receive CPA/EE (2 mg/day: 2 mg cyproterone acetate and 35-μg ethinylestradiol,) +Met (1500 mg/day) or GLP-1 RA (liraglutide, 1.2-1.8 mg/day) +Met (1500 mg/day) for 12 weeks. The clinical effectiveness and adverse effects were evaluated, followed by plasma proteomic analysis and verification of critical biomarkers by ELISA. [RESULTS] Eighty(80%) patients completed the study. Both interventions improved menstrual cycle, polycystic ovaries, LH(luteinizing hormone) and HbA1c(hemoglobin A1c) levels after the 12-week treatment. GLP-1RA + Met was more effective than CPA/EE + Met in reducing body weight, BMI (Body Mass Index), and waist circumference, FBG(fasting blood glucose), AUCI(area under curve of insulin),TC (Total Cholesterol), IL-6(Interleukin-6) and improving insulin sensitivity, and ovulation in overweight women with PCOS, with acceptable short-term side effects. CPA/EE + Met was more effective in improving hyperandrogenemia, including T(total testosterone), LH, LH/FSH(Luteinizing hormone/follicle-stimulating hormone), SHBG(sex hormone-binding globulin) and FAI (free androgen index). By contract, GLP-1RA+Met group only improved LH. Plasma proteomic analysis revealed that the interventions altered proteins involved in reactive oxygen species detoxification (PRDX6, GSTO1, GSTP1, GSTM2), platelet degranulation (FN1), and the immune response (SERPINB9). [CONCLUSIONS] Both CPA/EE+Met and GLP-1RA + Met treatment improved reproductive functions in overweight PCOS women. GLP-1RA + Met was more effective than CPA/EE + Met in reducing body weight, BMI, and waist, and improving metabolism, and ovulation in overweight women with PCOS, with acceptable short-term side effects. CPA/EE + Met was more effective in reducing hyperandrogenemia. The novel plasma biomarkers PRDX6, FN1, and SERPINB9, might be indicators and targets for PCOS treatment. TRIAL REGISTRATION CLINICALTIALS. [GOV TRIAL NO] NCT03151005. Registered 12 May, 2017, https://clinicaltrials.gov/ct2/show/NCT03151005 .
Where this record came from
| Source | Retrieved | Identifier |
|---|---|---|
| pubmed | Sep 13, 2026 | 37653215 first ingestion |