Tirzepatide on obstructive sleep apnea-related cardiometabolic risk: secondary outcomes of the SURMOUNT-OSA randomized 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?
- Specifically tested[Auto] Abstract addresses weight-loss independence: "Post hoc analyses include changes in a homeostatic model assessment for insulin resistance and mediation analysis to determine the proportion of observed changes attributable to reductions in body weight, apnea-hypopnea index and sleep apnea-specific hypoxic burden."
- 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; last sentences of abstract] Based on the mediation analysis, treating both sleep-disordered breathing and obesity is likely required to optimize the treatment effect on cardiometabolic benefits for patients with moderate-to-severe OSA and obesity. The ClinicalTrials.gov registration number for this study is NCT05412004 .
01Findings
What the study reported
- Drugs
- Tirzepatide
- Comparator
- placebo
- Primary outcome
- Not extracted
- Effect
- Not extracted
- 95% confidence interval
- Not extracted
- Follow-up
- Not stated
- 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)
- Baseline condition
- obstructive sleep apnea
Study quality details
- Study design
- Randomized controlled trial
- Sample size
- not extracted
- Randomization
- yes
- Blinding
- double-blind
- Comparator
- placebo
- Follow up duration
- not stated
- 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.
Percent change in hsCRP from baseline (tirzepatide vs placebo), study 1 (participants unwilling/unable to use PAP) · Week 52
Some concerns about risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No domain is High. Some concerns at D1 (concealment not described), D2 (efficacy estimand with unverifiable discontinuation balance), D3 (missingness unquantifiable) and D5 (estimand/transformation prespecification unverifiable) give an overall judgment of Some concerns. The effect is large and consistent, but that does not enter the risk-of-bias judgment.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Randomization and 1:1 allocation are stated and baseline characteristics are described as comparable, but no sequence-generation method and no allocation-concealment mechanism (IVRS/IWRS or other central system) is described anywhere in the available source. The guide's industry-trial presumption (D1 rule) requires a described central/IVRS system; it is affirmatively absent from this text, so concealment is NI rather than PY. D1 is kept trial-level per calibration ruling 4.Participants were randomly assigned in a 1:1 ratio to receive either tirzepatide at the maximum tolerated dose (10 or 15 mg) or placebo once weekly. The baseline characteristics of investigated cardiometabolic risk measures were, in general, comparable between the tirzepatide and placebo groups across both study 1 and study 2 at baseline. |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Double-blind, placebo-controlled design is stated; per the D2 collection rule, unblinding is not the basis of judgment for a lab biomarker. The judgment rests on discontinuation handling: the assessed result uses the efficacy estimand, which excludes data after permanent discontinuation of study intervention, with multiple imputation by reason of intercurrent event. No per-arm discontinuation counts are reported in the available source (Fig. 1 disposition is referenced but its numbers are not in the text), so imbalance in discontinuation cannot be checked, and the analysis is not an intention-to-treat analysis of the assignment effect. For the adherence effect of interest, the estimand is appropriate in form but its handling of intercurrent events cannot be verified against the disposition.The SURMOUNT-OSA master protocol comprised two, 52-week, randomized, double-blind, placebo-controlled phase 3 studies (study 1 and study 2) The analyses reported here were guided by the ‘efficacy’ estimand; the analysis included data collected before permanent discontinuation of study intervention and was conducted using the efficacy analysis set. |
| D3 · Missing outcome data Some concerns about risk of bias | No analysis denominator is given for the Table 2 hsCRP result and Fig. 1 disposition numbers are not reproduced, so neither D3 prong can be evaluated: percentage missing is NI and the between-arm difference in missingness is NI. Multiple imputation based on the reason for intercurrent events was used, which is a principled handling, but with missingness unquantifiable and 27.1% of the 469 randomized participants absent from the pooled hsCRP mediation set (n = 342), the assessed result cannot be placed on the Low side of the thresholds. NI on a decisive prong yields Some concerns, not Low.For efficacy analysis, missing values were imputed using multiple imputation based on the reason of intercurrent events. A total of 469 participants were randomized to receive either tirzepatide or placebo across study 1 ( Fig. 1 ; 234 participants; tirzepatide n = 114, placebo n = 120) and study 2 ( Fig. 1 ; 235 participants; tirzepatide n = 120, placebo n = 115). |
| D4 · Measurement of the outcome Low risk of bias | hsCRP is an objective laboratory biomarker; under the D4 collection rule such assays are Low unless the assay changed partway through or measurement differed between arms. The design is double-blind and placebo-controlled, both arms followed the same schedule, and nothing in the source indicates any assay change or arm-differential measurement. The specific assay and central-laboratory arrangement are not described (NI), which does not by itself raise the judgment under the rule.Parameters measured included change from baseline in SBP, DBP and levels of hsCRP, HDL-C, non-HDL-C, triglycerides, LDL-C, VLDL-C, fasting insulin and HOMA-IR, in tirzepatide versus placebo. randomized, double-blind, placebo-controlled phase 3 studies |
| D5 · Selection of the reported result Some concerns about risk of bias | The outcome itself is stated to be prespecified in the SAP, which supports the domain. But the specific assessed analysis — percent change on the efficacy estimand, unadjusted for multiplicity and described as exploratory — is a different analysis from the previously reported treatment-regimen-estimand result for the same outcome, and neither the protocol nor the SAP is among the sources available, so prespecification of this estimand/transformation cannot be verified. Under the D5 rule, prespecification unverifiable, plus two estimands and a percent-change transformation for the same outcome with one selected for report here, is Some concerns. Calibration ruling 1 applied: judged on this result, not the paper's whole suite.The SURMOUNT-OSA statistical analysis plan prespecified assessment of changes in SBP and DBP, hsCRP, HDL-C, non-HDL-C, LDL-C, VLDL-C and fasting insulin. Changes in SBP, DBP and hsCRP from SURMOUNT-OSA have already been reported using treatment-regimen estimand analysis |
Percent change in hsCRP from baseline (tirzepatide vs placebo), study 2 (participants using PAP therapy) · Week 52
Some concerns about risk of bias · RoB 2 · Two reviewers agreed on every domain
How the overall was reached: No domain is High; Some concerns at D1, D2, D3 and D5. The additional PAP-discontinuation intercurrent event in study 2 is noted at D2 but does not on its own raise the domain to High absent evidence of arm-differential effect.
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Same trial-level randomization description covers both studies: 1:1 random assignment stated, no sequence-generation or allocation-concealment mechanism described, so concealment is NI. Baseline characteristics described as comparable in both studies. Ruling 4 applied.Participants were randomly assigned in a 1:1 ratio to receive either tirzepatide at the maximum tolerated dose (10 or 15 mg) or placebo once weekly. The baseline characteristics of investigated cardiometabolic risk measures were, in general, comparable between the tirzepatide and placebo groups across both study 1 and study 2 at baseline. |
| D2 · Deviations from intended interventions Some concerns about risk of bias | As for study 1: double-blind placebo-controlled design, so unblinding is not the basis (D2 collection rule for lab biomarkers). The assessed result uses the efficacy estimand, which censors data after permanent discontinuation of study intervention, and no per-arm discontinuation counts are available in this source, so discontinuation imbalance cannot be checked. Study 2 carries an additional intercurrent-event source — PAP discontinuation at week 52 — which the authors themselves treat as confounding for the week-48 BP endpoints; its effect on the week-52 hsCRP endpoint is not addressed.randomized, double-blind, placebo-controlled phase 3 studies (study 1 and study 2) Changes in SBP and DBP were assessed at week 48 to prevent confounding effect of PAP discontinuation at week 52 in study 2, all other changes were assessed at week 52. |
| D3 · Missing outcome data Some concerns about risk of bias | As for study 1: no analysis denominator for the study 2 hsCRP result and no reproduced disposition figures, so both D3 prongs are NI. Multiple imputation by reason of intercurrent event was applied. The pooled hsCRP analysis set (n = 342 of 469 randomized) indicates material attrition programme-wide but cannot be attributed per study or per arm.For efficacy analysis, missing values were imputed using multiple imputation based on the reason of intercurrent events. hsCRP n = 342 |
| D4 · Measurement of the outcome Low risk of bias | Objective central laboratory-type biomarker under a double-blind design; no indication of an assay change or of measurement differing between arms. D4 collection rule applied. Assay details themselves are NI, which the rule does not treat as raising the judgment.Parameters measured included change from baseline in SBP, DBP and levels of hsCRP, HDL-C, non-HDL-C, triglycerides, LDL-C, VLDL-C, fasting insulin and HOMA-IR, in tirzepatide versus placebo. randomized, double-blind, placebo-controlled phase 3 studies |
| D5 · Selection of the reported result Some concerns about risk of bias | Same as study 1: the hsCRP outcome is stated to be SAP-prespecified, but the reported analysis is the efficacy-estimand percent change, unadjusted for multiplicity and labelled exploratory, and the protocol/SAP is not available to verify prespecification of this estimand and transformation. Two estimands exist for the same outcome with one reported here.The SURMOUNT-OSA statistical analysis plan prespecified assessment of changes in SBP and DBP, hsCRP, HDL-C, non-HDL-C, LDL-C, VLDL-C and fasting insulin. Changes in SBP, DBP and hsCRP from SURMOUNT-OSA have already been reported using treatment-regimen estimand analysis |
Assessed, not settled (1)
Recorded rather than omitted. Leaving these out would make the appraisal look cleaner than it is, and a disagreement between two careful readers is itself worth knowing.
Proportion of the tirzepatide effect on hsCRP mediated by change in AHI and SASHB (OSA metrics) independent of change in body weight — i.e. the hsCRP reduction not explained by weight change · Baseline to end of study (week 52)
Assessed, not settled · RoB 2 · Two reviewers reached different judgments and nothing has settled it
This result has no overall judgment, because the two independent reviews did not reach one. Both readings are shown. The domains they did agree on are below.
| Where they differ | The two readings |
|---|---|
| D3 · Missing outcome data | One reviewer: Some concerns about risk of bias The other: High risk of bias |
| Domain | Judgment and reasoning |
|---|---|
| D1 · Randomisation process Some concerns about risk of bias | Trial-level judgment carried from the two parent trials: randomization stated, concealment method not described (NI). Per calibration ruling 4, the non-randomized character of the mediation decomposition is not written into D1 — it is carried at D5, in the overall judgment and in weight_independence.Participants were randomly assigned in a 1:1 ratio to receive either tirzepatide at the maximum tolerated dose (10 or 15 mg) or placebo once weekly. |
| D2 · Deviations from intended interventions Some concerns about risk of bias | Double-blind placebo-controlled parent trials, so awareness is not the basis (D2 collection rule). The mediation analysis is run on the efficacy estimand, which excludes post- discontinuation data, and pools two trials; per-arm discontinuation counts are not available in this source, so discontinuation imbalance cannot be checked for the pooled analysis set.The analyses reported here were guided by the ‘efficacy’ estimand; the analysis included data collected before permanent discontinuation of study intervention and was conducted using the efficacy analysis set. The mediation analyses were performed on the pooled study 1 and study 2 populations. |
| D3 · Missing outcome data Some concerns about risk of bias | The hsCRP mediation analysis uses n = 342 of the 469 randomized participants — 27.1% of randomized participants are absent, so the first D3 prong (>20% missing) trips and the answer to "data available for nearly all" is PN. The second prong (per-arm difference in missingness) cannot be evaluated: no per-arm analysis-set counts are reported. Missingness could plausibly depend on the true value, since discontinuation for gastrointestinal events tracks drug exposure, and no sensitivity analysis for value-dependent missingness is reported (multiple imputation by reason of intercurrent event is a modelling assumption, not such a sensitivity analysis). Under calibration ruling 2, High requires both prongs to trip; only one is demonstrated and the other is NI, so this is Some concerns rather than High.hsCRP n = 342 A total of 469 participants were randomized to receive either tirzepatide or placebo across study 1 ( Fig. 1 ; 234 participants; tirzepatide n = 114, placebo n = 120) and study 2 ( Fig. 1 ; 235 participants; tirzepatide n = 120, placebo n = 115). |
| D4 · Measurement of the outcome Low risk of bias | The outcome measured is hsCRP, an objective laboratory biomarker, measured identically under a double-blind design; the mediators (body weight, AHI, SASHB) are likewise objective measurements. No indication of assay change or arm-differential measurement. D4 collection rule applied.Analyses on SBP and DBP were performed on the original scale, and analyses on hsCRP, HDL-C, non-HDL-C, triglycerides and HOMA-IR were performed on the log scale. randomized, double-blind, placebo-controlled phase 3 studies |
| D5 · Selection of the reported result High risk of bias | The mediation analysis is the load-bearing derived analysis for O4 and the paper itself states it is post hoc, i.e. affirmatively not prespecified. The Results section enumerates what the SAP did prespecify (a list of change-from-baseline endpoints) and the mediation analysis is not in that list. The specific analytic choices are numerous and unconstrained by any prespecification: pooling of two separate trials, log-scale outcome, choice of three mediator scenarios, choice of which post-treatment confounders to control in each scenario, and the covariate set. The estimated quantity (proportion mediated) is also not adjusted for multiplicity across the seven cardiometabolic measures. Under calibration ruling 3, absent prespecification on a load-bearing derived analysis is High, not Some concerns; the "at least Some concerns" in the D5 domain rule is a floor.Post hoc analyses include changes in a homeostatic model assessment for insulin resistance and mediation analysis to determine the proportion of observed changes attributable to reductions in body weight, apnea-hypopnea index and sleep apnea-specific hypoxic burden. The SURMOUNT-OSA statistical analysis plan prespecified assessment of changes in SBP and DBP, hsCRP, HDL-C, non-HDL-C, LDL-C, VLDL-C and fasting insulin. The changes in HOMA-IR were analyzed post hoc. |
02Funding
Funding and conflicts
- Funding
- NHLBI NIH HHS; NIA NIH HHS
- Industry funded
- No
- Manufacturer
- Novo Nordisk, Eli Lilly, Lilly, AstraZeneca, Boehringer Ingelheim, Amgen, Pfizer, Zealand Pharma, Structure Therapeutics
- Sponsor role
- no manufacturer funding identified
- Author conflicts
- Competing interests: A.M. is funded by the NIH. A.M. reports income from Eli Lilly, Livanova, Zoll and Powell Mansfield. ResMed gave a philanthropic donation to USCD. R.G. reports income from serving on the advisory board for Apnimed, the steering committee for SURMOUNT-OSA, Eli Lilly and lecture fees from Somnomed Department. He conducts sponsored studies with Eli Lilly, Alkermes, Takeda and Bod Science: Lambert Initiative. A.A. serves as a consultant for Respicardia, Eli Lilly, Inspire, Cerebra and Apnimed. Apnimed is developing pharmacological treatments for Obstructive Sleep Apnea. A.A.’s interests were reviewed by Brigham and Women’s Hospital and Mass General Brigham in accordance with their institutional policies. S.S. received grant support from Apnimed, Prosomnus and Dynaflex and has served as a consultant for Apnimed, Nox Medical, Inspire Medical Systems, Eli Lilly, Respicardia, LinguaFlex and Achaemenid. S.S. receives royalties for intellectual property pertaining to combination pharmacotherapy for sleep apnea via his Institution. S.S. is also the co-inventor of intellectual property pertaining to wearable sleep apnea phenotyping also via his Institution. He has received equity in Achaemenid, a company commercializing biosensor technology for monitoring oral appliance treatment efficacy. S.S. is also co-inventor of intellectual property pertaining to wearable sleep apnea phenotyping also via his Institution. His industry interactions are actively managed by his Inst
- Independent replication
- unknown
- Notes
- Authors declare relationships with the drug's manufacturer: Eli Lilly
Funding is shown on every study and never used to score it.
04Source
The source, as retrieved
Abstract
Obstructive sleep apnea (OSA) is associated with obesity and cardiovascular risk. The SURMOUNT-OSA master protocol comprised two, 52-week, randomized, double-blind, placebo-controlled phase 3 studies (study 1 and study 2) and demonstrated a significant reduction of a number of cardiometabolic risk measures in participants with OSA and obesity following treatment with tirzepatide. Here we report prespecified analysis of cardiometabolic risk measures in SURMOUNT-OSA. Post hoc analyses include changes in a homeostatic model assessment for insulin resistance and mediation analysis to determine the proportion of observed changes attributable to reductions in body weight, apnea-hypopnea index and sleep apnea-specific hypoxic burden. In both study 1 and study 2 of SURMOUNT-OSA, tirzepatide treatment was associated with greater alleviation of cardiometabolic risk factors than placebo. Independent mediation effect of changes in OSA metrics was observed on high-sensitivity C-reactive protein, homeostatic model assessment for insulin resistance and triglycerides. The combination of changes in weight and OSA metrics, as well as weight alone, had a significant mediation effect on systolic blood pressure, but there was no significant mediation effect of weight or OSA metrics observed on diastolic blood pressure. Based on the mediation analysis, treating both sleep-disordered breathing and obesity is likely required to optimize the treatment effect on cardiometabolic benefits for patients with moderate-to-severe OSA and obesity. The ClinicalTrials.gov registration number for this study is NCT05412004 .
Where this record came from
| Source | Retrieved | Identifier |
|---|---|---|
| pubmed | Sep 13, 2026 | 41540105 first ingestion |