High evidence 1Moderate evidence 8Low evidence 6Very low evidence 5
Tirzepatide|semaglutide (head-to-head) is shown as a marketed evidence record where applicable. Findings describe observed research and regulatory records; they do not provide treatment advice.
Tirzepatide|semaglutide (head-to-head)MARKETED
Safety signalsNo change
High evidence
FDA, after a comprehensive review incl. a meta-analysis of 91 trials, found no increased suicidality risk and REQUESTED REMOVAL of the suicidal-behaviour-and-ideation...
SourceFDA Drug Safety Communication, 'FDA Requests Removal of Suicidal Behavior and Ideation Wa...Source
Full findingFDA, after a comprehensive review incl. a meta-analysis of 91 trials, found no increased suicidality risk and REQUESTED REMOVAL of the suicidal-behaviour-and-ideation warning from labels for liraglutide, semaglutide 2.4 mg and tirzepatide.
PopulationFDA pooled clinical-trial meta-analysis + post-marketing/observational review
Fundingregulatory agency (FDA/EMA/MHRA/WHO)
Scope limitsStrongest regulatory conclusion to date (warning REMOVAL). Began as a 2023 FDA investigation into post-marketing suicidality + aspiration reports; interim Jan-2024 DSC already found no evidence of causation.
Comparatorsplacebo
Tirzepatide|semaglutide (head-to-head)MARKETED
Safety signalsDecrease
Low evidence
Narrative review framing: incretin agents cause rapid significant lean-mass loss (~10% of body weight, ~6 kg), characterised as comparable to a decade+ of ageing,...
SourceLocatelli JC et al., Diabetes Care 2024Source
Full findingNarrative review framing: incretin agents cause rapid significant lean-mass loss (~10% of body weight, ~6 kg), characterised as comparable to a decade+ of ageing, raising sarcopenia/frailty concern; resistance exercise proposed to preserve muscle.
PopulationNarrative review across liraglutide, semaglutide, tirzepatide, retatrutide (~15-24% weight loss)
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Scope limits'Ageing-equivalent' framing is the authors' characterisation, not a measured endpoint. Includes retatrutide. Counterpoint in C-CLASS-SAFETY-MUSCLE15.
Tirzepatide|semaglutide (head-to-head)MARKETED
Safety signalsMixed
Low evidence
Animal reproductive-tox studies show adverse offspring outcomes (decreased fetal growth, skeletal/visceral anomalies, embryonic death); no consistent pattern of human...
Full findingAnimal reproductive-tox studies show adverse offspring outcomes (decreased fetal growth, skeletal/visceral anomalies, embryonic death); no consistent pattern of human congenital anomalies to date.
PopulationReview of animal repro-tox + human case reports/cohort/population studies
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Scope limitsanimal data; human relevance uncertain
Comparatorsinsulin
Tirzepatide|semaglutide (head-to-head)MARKETED
Safety signalsIncrease
Very low evidence
FAERS disproportionality: strong post-marketing reporting signal for optic ischaemic neuropathy with semaglutide; weaker but significant for tirzepatide and...
SourceKakkar AK, Payra S, Charaya A, Diabetes Obes Metab 2026Source
Full findingFAERS disproportionality: strong post-marketing reporting signal for optic ischaemic neuropathy with semaglutide; weaker but significant for tirzepatide and liraglutide; none for dulaglutide/exenatide/lixisenatide.
VigiBase (WHO) disproportionality: significantly increased ROR for suicidal IDEATION with semaglutide, liraglutide, tirzepatide, and for 'depression/suicidal' and...
SourceMcIntyre RS et al., J Affect Disord 2024;369:922-927Source
Full findingVigiBase (WHO) disproportionality: significantly increased ROR for suicidal IDEATION with semaglutide, liraglutide, tirzepatide, and for 'depression/suicidal' and suicidal behaviour with semaglutide/liraglutide; BUT significantly DECREASED ROR for suicide attempts and completed suicide.
FAERS unmasking analysis (2005-2023) with metformin/orlistat comparators: per-agent RORs for psychiatric ADRs elevated for semaglutide (2.03), tirzepatide (1.76),...
SourceGuirguis A et al., Eur Neuropsychopharmacol 2024;82:82-91Source
Full findingFAERS unmasking analysis (2005-2023) with metformin/orlistat comparators: per-agent RORs for psychiatric ADRs elevated for semaglutide (2.03), tirzepatide (1.76), liraglutide (1.64) but lower than metformin; 42 deaths incl. 13 completed suicides; no causal link inferable.
Opposite-pole mechanistic view: emerging preclinical/translational data suggest possible muscle-specific BENEFICIAL effects (attenuated atrophy, improved...
SourceKoceva A, Janez A, Jensterle M, Medicina (Kaunas) 2025Source
Full findingOpposite-pole mechanistic view: emerging preclinical/translational data suggest possible muscle-specific BENEFICIAL effects (attenuated atrophy, improved myogenesis/mitochondrial function, reduced myosteatosis), alongside clinical proportional fat/lean loss.
PopulationNarrative review of preclinical (animal) and clinical evidence on skeletal muscle
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Scope limitsanimal data; human relevance uncertain
Tirzepatide|semaglutide (head-to-head)MARKETED
Safety signalsMixed
Very low evidence
FAERS comparative disproportionality found acute kidney injury reported LESS frequently with tirzepatide than semaglutide.
SourceComparative Renal Safety of Tirzepatide and Semaglutide: FAERS Disproportionality Study 2025Source
PopulationFAERS Jan 2022-Sep 2025, 133,872 reports (92,807 tirzepatide; 41,065 semaglutide)
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Scope limitsidentifier not fully verified
Comparatorssemaglutide
Tirzepatide|semaglutide (head-to-head)MARKETED
Body compositionDecrease
Moderate evidence
Systematic review (Annals of Internal Medicine): across 35 RCTs, incretin therapies consistently reduced fat mass and visceral adiposity; muscle-based-index loss was...
SourceBatsis JA et al., Ann Intern Med, 2026Source
Full findingSystematic review (Annals of Internal Medicine): across 35 RCTs, incretin therapies consistently reduced fat mass and visceral adiposity; muscle-based-index loss was highly heterogeneous, with median 28.3% of total weight loss attributable to muscle-based indices and ~two-thirds of interventions exceeding pre-specified muscle-loss benchmarks. No study reported objective physical-function outcomes.
Population35 RCTs (median duration 26 weeks, median N 78) of liraglutide, semaglutide, tirzepatide or dulaglutide in adults with obesity; BIA/DXA/CT/MRI; Jan 2003-Feb 2026
Scope limitsClass-level synthesis; heterogeneity precluded meta-analysis. Note the FFM/muscle benchmark is measurement-method-dependent (~25% for BIA/DXA, ~15% for CT/MRI). Anchors the comparison matrix. The 'no objective physical function outcomes' gap is a recurring caveat.
Comparatorsplacebo; lifestyle
Tirzepatide|semaglutide (head-to-head)MARKETED
Body compositionDecrease
Moderate evidence
Systematic review and meta-analysis (20 RCTs, 15,782 adults, DXA/MRI) with absolute lean-mass change (kg) and lean fraction of weight lost as CO-PRIMARY outcomes. Lean...
SourceEisa N, Barood O et al., Diabetes Obes Metab, 2026Source
Full findingSystematic review and meta-analysis (20 RCTs, 15,782 adults, DXA/MRI) with absolute lean-mass change (kg) and lean fraction of weight lost as CO-PRIMARY outcomes. Lean mass was ~25-39% of total weight lost across incretin agents and broadly comparable to lifestyle intervention (no significant difference, p=0.42). The only modality that materially lowered the lean fraction was lifestyle PLUS resistance training (17.5%). Retatrutide is NOT included.
Population20 RCTs, 15,782 adults with overweight/obesity; semaglutide, tirzepatide, liraglutide or lifestyle; DXA or MRI; searched to 20 Jan 2026; random-effects, RoB 2, GRADE.
Fundingacademic / investigator-initiated; no industry sponsor found in indexed metadata (Europe PMC fundingList empty)
Scope limitsNEW source. Most direct head-to-class lean-fraction comparison and the closest thing to an absolute-lean-kg anchor. Decision-relevant: incretin lean fraction is no worse than diet-induced loss of similar magnitude (p=0.42); resistance training is the lever, not the drug. CRUCIAL GAP: retatrutide is ABSENT, so the class benchmark exists but no reta point sits in it. Meta-analysis graded moderate by default. DXA/MRI lean is not muscle function. VALIDATOR: confirm PMID 41877354 resolves to this DOM 2026 meta and the per-drug fractions (sema 35.2%, tirz 25.4%). Cross-ref T5-014, T5-010.
Comparatorslifestyle intervention; lifestyle plus resistance training
SourceLocatelli JC et al., Diabetes Care 2024Source
Full findingCAUTION-POLE anchor (both poles preserved): narrative review argues incretin agents (liraglutide, semaglutide, tirzepatide, retatrutide named) inducing ~15-24% weight loss also cause rapid lean-mass loss of ~10% / ~6 kg, threatening muscle mass/function and raising sarcopenia/frailty risk, motivating resistance exercise and muscle-preserving adjuncts. The same DXA numbers the adaptive camp reads as benign are read here as a threat: the disagreement is INTERPRETIVE, not over the raw number.
PopulationNarrative review; adults with overweight/obesity on incretin therapy; older-adult sarcopenia/frailty framing.
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Scope limitsCAUTION pole. Names retatrutide in the lean-loss list but cites NO reta-specific function data (see C-RETATRUTIDE-BODYCOMP-03 gap). Still a MASS argument, not strength data. The ~10%/~6 kg figure is a CLASS generalisation; NO reta-specific lean-kg supports reta's inclusion (the review names reta but does not measure it; Council Methodologist). The 'decade of ageing' is a framing analogy. Two further caution reviews fold here to avoid duplicates: Stefanakis/Mantzoros Metabolism 2024 (PMID 39481534, '>25% of loss is fat-free mass') and the JAMA viewpoint Conte/Hall/Klein 2024 (PMID 38829659, which frames the harm-vs-adaptive question as explicitly UNRESOLVED). Counter-pole: C-CLASS-BODYCOMP-05.
Tirzepatide|semaglutide (head-to-head)MARKETED
Insulin and beta-cell functionMixed
Moderate evidence
SURPASS-2 biomarker substudy (week 40): the tirzepatide-minus-semaglutide differential isolates a GIP-attributable islet increment — greater beta-cell function...
SourceFrias JP et al., J Clin Endocrinol Metab 2024;109(7):1745-1753Source
Full findingSURPASS-2 biomarker substudy (week 40): the tirzepatide-minus-semaglutide differential isolates a GIP-attributable islet increment — greater beta-cell function (HOMA2-B), greater insulin-sensitisation (HOMA2-IR / fasting insulin), and greater fasting-glucagon suppression at higher doses.
Scope limitsReused from corpus T11-S2-08/09/10/03. CONFOUNDS (flagged, not resolved): (a) GIP-or-different-GLP-1 (sema dosed 1 mg not 2.4 mg); (b) the larger HOMA2-IR/insulin improvement is substantially WEIGHT-confounded (tirzepatide lost more weight); (c) greater glucagon suppression is mechanistically counter-intuitive for GIP (GIP is glucagonotropic at fasting), so likely GLP-1-led or downstream of greater glycaemia/weight. HOMA2-B is the axis where GIP attribution is most coherent. COI (sponsor framing).
Comparatorssemaglutide 1 mg
Tirzepatide|semaglutide (head-to-head)MARKETED
Insulin and beta-cell functionIncrease
Moderate evidence
SURMOUNT-5 prediabetes subgroup (post-hoc, n=425): tirzepatide produced greater HbA1c reduction, higher reversion to normoglycaemia, and greater improvements in...
SourceGalindo RJ, Aronne LJ, Horn DB et al., J Endocrinol Invest 2026Source
Full findingSURMOUNT-5 prediabetes subgroup (post-hoc, n=425): tirzepatide produced greater HbA1c reduction, higher reversion to normoglycaemia, and greater improvements in fasting insulin and HOMA2-IR than semaglutide, alongside greater weight loss.
In a 28-week randomised phase 1 clamp trial, tirzepatide improved the clamp disposition index more than placebo and more than semaglutide, reflecting greater total...
SourceHeise T et al., Lancet Diabetes Endocrinol 2022;10(6):418-429Source
Full findingIn a 28-week randomised phase 1 clamp trial, tirzepatide improved the clamp disposition index more than placebo and more than semaglutide, reflecting greater total insulin secretion rate and greater clamp insulin sensitivity (M-value), with greater glucagon lowering. The tirzepatide-minus-semaglutide differential isolates a human GIP increment on insulin secretion and sensitivity.
Scope limitsReused from corpus T3-003/T3-004. Drug-level decomposition is combination-only; the human tirz-minus-sema differential is the permitted GIP isolation (caveat C1 not triggered — human, not mouse). No glucagon-agonist arm, so does not model retatrutide's GCGR offset.
Comparatorsplacebo; semaglutide
Thin or bounded evidence here means uncertainty remains visible. It is not evidence that an effect has been ruled out.
Tirzepatide|semaglutide (head-to-head)MARKETED
comparative-efficacyNot directional
Moderate evidence
FOLK-CLAIM VERDICTS (strength hierarchy): (1) 'reta > tirz > sema' = TRUE directionally, OVERSTATED on certainty - the tirz>sema rung is head-to-head proven...
SourceAronne LJ, et al. N Engl J Med 2025 (SURMOUNT-5); presented ECO 2025Source
Full findingFOLK-CLAIM VERDICTS (strength hierarchy): (1) 'reta > tirz > sema' = TRUE directionally, OVERSTATED on certainty - the tirz>sema rung is head-to-head proven (SURMOUNT-5), but reta's #1 rank rests entirely on CROSS-TRIAL inference from a single phase-2 trial (no reta-vs-tirz head-to-head exists); the direction is very likely right, the certainty is not established (TRIUMPH phase-3 will settle it). (2) 'tirz is stronger than sema' = TRUE (SURMOUNT-5). (3) 'sema stopped working' = FALSE as a mechanism - it is a sustained plateau + a lower efficacy ceiling than tirzepatide, not tachyphylaxis; switching benefit is real but is ceiling-access, not tolerance-escape. (4) 'reta is DIFFERENT, not just stronger' = TRUE - it is a TRIPLE agonist; the glucagon arm adds effects the dual/mono agents lack (liver-fat clearance, energy expenditure, the HR rise, dysesthesia), so it is not 'tirzepatide at a higher dose'.
PopulationCross-trial comparison of the pivotal weight-loss trials + the one head-to-head (SURMOUNT-5).
FOLK-CLAIM VERDICT ('tirzepatide is gentler than semaglutide despite being stronger', attributed to GIP): MIXED, leaning TRUE - the robust HUMAN finding is LOWER...
SourceAronne LJ et al. Tirzepatide vs semaglutide for obesity (SURMOUNT-5). N Engl J Med 2025 (...Source
Full findingFOLK-CLAIM VERDICT ('tirzepatide is gentler than semaglutide despite being stronger', attributed to GIP): MIXED, leaning TRUE - the robust HUMAN finding is LOWER GI-driven discontinuation with tirzepatide (SURMOUNT-5, 2.7% vs 5.6%); the nausea gap itself is modest and partly dose/exposure-confounded. The GIP-antiemetic mechanism is real but mostly ANIMAL-proven. So tirzepatide is MODESTLY better tolerated (clearest on vomiting/discontinuation), not dramatically 'gentle'.
PopulationSURMOUNT-5 randomised head-to-head (tirzepatide vs semaglutide, obesity) + indirect/network meta-analyses + preclinical GIP-antiemetic mechanism (shrew/rat area postrema).
Fundingindustry-Eli Lilly
Scope limitsanimal data; human relevance uncertain; conference/abstract-level
Comparatorssemaglutide
Thin or bounded evidence here means uncertainty remains visible. It is not evidence that an effect has been ruled out.
Tirzepatide|semaglutide (head-to-head)MARKETED
cns-neuropsychiatricMixed
Low evidence
ADDICTION (alcohol) - a within-class gradient: in a target-trial-emulation EHR study, tirzepatide and semaglutide (but NOT liraglutide or dulaglutide) were associated...
SourceHenney AE, Riley DR, Heague M et al., Diabetes Obes Metab, 2025Source
Full findingADDICTION (alcohol) - a within-class gradient: in a target-trial-emulation EHR study, tirzepatide and semaglutide (but NOT liraglutide or dulaglutide) were associated with lower incident AUD vs DPP-4 inhibitors, suggesting potency or GIP co-agonism may matter (an open question, not a verdict).
PopulationHenney target-trial-emulation observational cohort (US EHR >120M): T2D without prior AUD, propensity-matched vs DPP4i; tirz n=7165, sema n=20198, lira n=6565, dula n=19061; 18 months.
Fundingacademic / investigator-initiated (Henney et al., University of Liverpool; EHR target-trial emulation)
Thin or bounded evidence here means uncertainty remains visible. It is not evidence that an effect has been ruled out.
Tirzepatide|semaglutide (head-to-head)MARKETED
health-economics-adherenceNot directional
Low evidence
In the largest US real-world EHR cohort, MOST patients with overweight/obesity discontinued GLP-1 RA within 1 year, with discontinuation substantially higher in...
Full findingIn the largest US real-world EHR cohort, MOST patients with overweight/obesity discontinued GLP-1 RA within 1 year, with discontinuation substantially higher in obesity-without-diabetes (64.8%) than in type 2 diabetes (46.5%); discontinuation was associated with cost/coverage proxies (income) and GI adverse events, while greater weight loss protected against stopping.
PopulationRetrospective cohort, Truveta EHR (collective of US health systems); 125,474 adults (BMI>=27) newly initiating liraglutide, semaglutide or tirzepatide 2018-2023; mean age 54.4, 65.4% women; 61.0% with T2D; obesity vs T2D strata; followed up to 2 yr; USA.
Fundingacademic/independent analysis (manufacturer not study sponsor; per-study COI not individually audited)
Thin or bounded evidence here means uncertainty remains visible. It is not evidence that an effect has been ruled out.
Tirzepatide|semaglutide (head-to-head)MARKETED
Gastrointestinal tolerabilityMixed
Moderate evidence
SURMOUNT-5 head-to-head: GI AEs most common with both; GI-AE-driven discontinuation occurred more often with semaglutide than tirzepatide.
SourceAronne LJ et al., SURMOUNT-5, N Engl J Med 2025Source
PopulationSURMOUNT-5: 751 adults with obesity, no diabetes; 72 weeks; phase 3b open-label; max-tolerated tirzepatide 10/15 mg vs semaglutide 1.7/2.4 mg
Fundingindustry - Eli Lilly (disclosed in publication)
Scope limitsidentifier not fully verified; open-label (unblinded)
Comparatorssemaglutide
Thin or bounded evidence here means uncertainty remains visible. It is not evidence that an effect has been ruled out.
Tirzepatide|semaglutide (head-to-head)MARKETED
Appetite and food intakeNo change
Moderate evidence
Tirzepatide 15 mg gave greater fat-mass loss than semaglutide 1 mg in T2D, but the difference was NOT explained by ad libitum lunch energy intake, and 24-h intake /...
SourceHeise T et al., Diabetes Care 2023;46(5):998-1004Source
Full findingTirzepatide 15 mg gave greater fat-mass loss than semaglutide 1 mg in T2D, but the difference was NOT explained by ad libitum lunch energy intake, and 24-h intake / substrate utilisation / energy expenditure were not measured. No clean human tirzepatide-vs-semaglutide appetite/satiety differential isolating a GIP appetite contribution was located.
PopulationT2D, tirzepatide 15 mg vs semaglutide 1 mg
Fundingindustry - Eli Lilly and Company
Scope limitsReused from corpus T9-GIPR-06. Measured-null on the single acute intake readout; broader endpoints unmeasured. The within-molecule GIP appetite contribution in humans is proxy-only/unresolved.
The Echo Compendium