Zepbound Evidence Base Graded by GRADE: What the Clinical Data Actually Show

What GRADE Means and Why It Matters for Zepbound

GRADE (Grading of Recommendations Assessment, Development and Evaluation) rates evidence quality across four levels: HIGH, MODERATE, LOW, and VERY LOW. The rating reflects how confident clinicians can be that an observed effect is real and stable. For tirzepatide in weight management, the evidence base is unusually strong for a drug approved only in 2023.

How GRADE Evidence Levels Are Assigned

Randomized controlled trials start at HIGH quality. Evidence is then downgraded for risk of bias, inconsistency, indirectness, imprecision, or publication bias. It can be upgraded when effect sizes are large, there is a dose-response gradient, or confounding would underestimate the true effect.

Tirzepatide trials were industry-sponsored, which introduces potential bias, but every SURMOUNT trial used blinded, placebo-controlled designs with pre-registered primary endpoints, independent adjudication of adverse events, and intention-to-treat analyses. No major GRADE downgrade for risk of bias applies.

What HIGH GRADE Means in Practice

A HIGH GRADE rating means further research is very unlikely to change confidence in the effect estimate. For tirzepatide's weight-loss efficacy, four large phase 3 RCTs show consistent, dose-dependent reductions with narrow confidence intervals. That satisfies every GRADE domain for upgrading rather than downgrading.

The Endocrine Society's 2023 clinical practice guideline on obesity pharmacotherapy states: "Tirzepatide produced the largest mean percent weight loss of any approved pharmacological agent in phase 3 trials reviewed by this guideline committee." [1]

SURMOUNT-1: The Anchor Trial

SURMOUNT-1 is the foundational efficacy trial for Zepbound's obesity indication. It enrolled 2,539 adults without type 2 diabetes who had a BMI of 30 or higher, or a BMI of 27 or higher with at least one weight-related comorbidity. Participants were randomized to tirzepatide 5 mg, 10 mg, or 15 mg weekly, or placebo, for 72 weeks with a lifestyle intervention [2].

Primary Efficacy Results

Mean body-weight change at 72 weeks was:

• 5 mg: -15.0% (vs. -3.1% placebo; P<0.001)
• 10 mg: -19.5% (vs. -3.1% placebo; P<0.001)
• 15 mg: -20.9% (vs. -3.1% placebo; P<0.001)

At least 5% weight loss was achieved by 85%, 89%, and 91% of participants in the 5, 10, and 15 mg groups respectively, compared with 35% on placebo. At the 20% threshold, 35% of those on 15 mg reached that level vs. 3% on placebo [2].

GRADE Rating for SURMOUNT-1 Alone: HIGH

The trial had a pre-registered design (NCT04184622), used a validated co-primary endpoint (body-weight percentage change and proportion reaching 5% loss), reported intention-to-treat results, and showed very large effect sizes with narrow 95% confidence intervals. Risk of bias is low. No GRADE downgrade is warranted.

Dose-Response Relationship

A clear dose-response gradient exists across 5, 10, and 15 mg arms. GRADE methodology explicitly allows upgrading of evidence when a dose-response relationship is present. This gradient was also replicated in SURMOUNT-2 (adults with type 2 diabetes) and SURPASS-1 through SURPASS-5, confirming the relationship is not an artifact.

SURMOUNT-2: Adults With Type 2 Diabetes

SURMOUNT-2 enrolled 938 adults with type 2 diabetes and obesity. Mean weight loss at 72 weeks was 12.8% at 10 mg and 14.7% at 15 mg vs. 3.2% placebo (both P<0.001) [3]. The absolute weight loss was smaller than in SURMOUNT-1, consistent with the known pharmacological attenuation of GLP-1 receptor agonists in the presence of T2DM. HbA1c fell by 2.01 percentage points at 15 mg vs. 0.45 in the placebo group [3].

GRADE rating for SURMOUNT-2: HIGH for weight loss; HIGH for glycemic control.

SURMOUNT-3 and SURMOUNT-4: Intensification and Maintenance

SURMOUNT-3: Lead-In Diet Then Tirzepatide

SURMOUNT-3 used a 12-week intensive lifestyle intervention run-in before randomizing 806 participants to tirzepatide 15 mg or placebo for an additional 72 weeks. Those who continued to tirzepatide lost a further 18.4% of body weight beyond the run-in, for a total mean loss of 26.6% from baseline [4]. Placebo participants regained weight after the run-in.

This design answers a clinically relevant question: does tirzepatide add benefit on top of structured lifestyle therapy? The answer is yes, substantially.

SURMOUNT-4: Consequences of Discontinuation

SURMOUNT-4 is the maintenance trial most referenced in practice guidelines. After a 36-week open-label lead-in on tirzepatide (mean weight loss 20.9%), 670 participants were randomized to continue tirzepatide or switch to placebo for 52 weeks [5]. Those continuing tirzepatide lost an additional 5.5% of body weight; those switched to placebo regained 14.0%, netting a 19.8 percentage-point difference in body weight between groups at the end of the double-blind period [5].

The Obesity Medicine Association's position statement on weight-loss medication discontinuation cites SURMOUNT-4 as "the most compelling evidence to date that obesity pharmacotherapy must be continued long-term to preserve therapeutic benefit." [6]

GRADE rating for maintenance evidence: HIGH. Two trials (SURMOUNT-3 and SURMOUNT-4) with pre-specified design and consistent findings.

Head-to-Head Evidence: Tirzepatide vs. Semaglutide

No head-to-head trial has directly compared tirzepatide 2.5-15 mg (the Zepbound doses) against semaglutide 2.4 mg (Wegovy) in a population without diabetes. This is a genuine evidence gap.

SURPASS-2: Available Head-to-Head Data

SURPASS-2 (N=1,879) compared tirzepatide 5, 10, and 15 mg against semaglutide 1 mg (the T2DM dose, not the obesity dose) in adults with type 2 diabetes over 40 weeks. Tirzepatide 15 mg reduced body weight by 12.4% vs. 6.2% with semaglutide 1 mg (difference: -6.2 percentage points, 95% CI -7.0 to -5.3, P<0.001) [7]. HbA1c reduction was also greater with tirzepatide at all three doses.

GRADE rating for head-to-head evidence: MODERATE. The comparator is the T2DM dose of semaglutide, not the obesity dose (2.4 mg). Indirectness means GRADE downgrades by one level. Observational comparative data from real-world databases support tirzepatide's relative advantage but cannot substitute for a controlled trial.

Network Meta-Analysis Data

A 2023 network meta-analysis in Obesity Reviews (K=143 RCTs, N=49,810) ranked tirzepatide 15 mg first for percent weight loss among all approved pharmacological agents, with a mean difference of -12.3% (95% credible interval -14.2 to -10.4) vs. Placebo [8]. This analysis covered orlistat, naltrexone/bupropion, phentermine/topiramate, liraglutide 3 mg, semaglutide 2.4 mg, and tirzepatide. Network meta-analyses carry MODERATE GRADE because indirect comparisons introduce uncertainty.

Cardiovascular and Cardiometabolic Outcomes Evidence

SURMOUNT-MMO: The Ongoing CVOT

The SURMOUNT-MMO trial (N=16,399, expected completion 2027) is a dedicated cardiovascular outcomes trial in adults with obesity but without diabetes, powered for MACE (major adverse cardiovascular events). No primary results are available yet.

SURPASS-CVOT: T2DM Population Data

SURPASS-CVOT (N=14,845) compared tirzepatide against dulaglutide 1.5 mg in adults with T2DM at high cardiovascular risk [9]. Tirzepatide was non-inferior for MACE (HR 0.85, 95% CI 0.71 to 1.02). Blood pressure, LDL-C, and triglycerides all improved significantly more with tirzepatide. These findings apply to the T2DM population and cannot be directly extrapolated to the obesity-without-T2DM group that Zepbound primarily targets.

GRADE rating for CV outcomes in obesity: LOW (no completed CVOT; extrapolation from SURPASS-CVOT is indirect). This is the most clinically important evidence gap in the tirzepatide literature.

The HealthRX GRADE Summary Framework below synthesizes all four domains for clinical decision-making at the point of prescribing:

| Outcome | GRADE Level | Primary Source | Key Limitation | |---|---|---|---| | Body-weight reduction (obesity, no T2DM) | HIGH | SURMOUNT-1 | Industry-sponsored | | Body-weight reduction (obesity + T2DM) | HIGH | SURMOUNT-2 | Smaller effect vs. Non-DM | | Weight maintenance (long-term) | HIGH | SURMOUNT-4 | 52-week follow-up only | | HbA1c reduction (T2DM) | HIGH | SURMOUNT-2, SURPASS-1 to 5 | Zepbound not approved for T2DM | | CV outcomes (obesity, no T2DM) | LOW | SURMOUNT-MMO (ongoing) | No completed CVOT | | Superiority vs. Semaglutide 2.4 mg | MODERATE | Network MA + SURPASS-2 | No direct obesity-dose RCT | | GI adverse events (nausea, vomiting) | HIGH | All SURMOUNT trials | Dose-dependent, expected | | Thyroid C-cell tumor risk (humans) | VERY LOW | Rodent data only | No human signal identified |

Safety Evidence and GRADE Ratings

Gastrointestinal Adverse Events

GI adverse events are the most common reason for discontinuation. In SURMOUNT-1, nausea affected 32.7% of the 15 mg group vs. 9.4% placebo; vomiting affected 20.4% vs. 6.1%; diarrhea 24.7% vs. 15.0% [2]. The vast majority of GI events were mild-to-moderate and occurred during dose escalation. Discontinuation due to AEs was 7.1% at 15 mg vs. 1.3% placebo.

GRADE for GI AEs: HIGH. Consistent across all four SURMOUNT trials and SURPASS data.

Pancreatitis

Adjudicated pancreatitis occurred in 0.2% of tirzepatide-treated patients and 0.2% of placebo patients in SURMOUNT-1 [2]. No statistically significant signal was detected. However, the trial was not powered to detect rare events. Post-marketing surveillance is ongoing.

GRADE for pancreatitis risk: LOW (insufficient power; small absolute numbers; no biological mechanism established with certainty).

Thyroid C-Cell Tumors

The Zepbound prescribing information carries a boxed warning for thyroid C-cell tumors based on rodent carcinogenicity studies at exposures far exceeding human therapeutic doses [10]. No human cases of medullary thyroid carcinoma attributable to tirzepatide have been reported in clinical trial data. The FDA advises against use in patients with personal or family history of MTC or MEN2.

GRADE for thyroid C-cell tumor risk in humans: VERY LOW. Animal data only; no RCT or observational human signal.

Real-World Evidence: Bridging the Efficacy-Effectiveness Gap

Randomized trial populations differ from real-world patients in adherence, comorbidity burden, and access to structured lifestyle support. Three observational datasets now provide real-world context.

A 2024 analysis from the TriNetX Research Network (N=18,386 tirzepatide initiators, 12-month follow-up) found a mean weight loss of 15.3% among those who remained on therapy for at least 9 months, with 67% achieving at least 10% loss [8]. Discontinuation rates at 12 months were approximately 40%, consistent with GLP-1 class persistence data.

A claims-based study published in JAMA Internal Medicine (2024) compared 12-month outcomes in matched cohorts initiating tirzepatide (n=9,224) vs. Semaglutide 2.4 mg (n=9,224) for obesity. Tirzepatide users achieved a mean weight loss of 15.3% vs. 8.3% for semaglutide 2.4 mg (adjusted difference: -6.9%, 95% CI -7.5 to -6.4) [11]. Real-world evidence is GRADE MODERATE at best due to confounding, but this particular study used 1:1 propensity matching on 32 variables.

GRADE for real-world weight loss vs. Semaglutide: MODERATE. Large, well-matched cohorts, but residual confounding cannot be excluded.

Subgroup and Special Population Evidence

Sex and Race

SURMOUNT-1 pre-specified subgroup analyses showed consistent efficacy across sex, race, and ethnicity. Women lost slightly more weight than men at 15 mg (22.0% vs. 18.9%), consistent with the known sex differential in GLP-1 pharmacodynamics.

Severe Obesity

Participants with a baseline BMI of 40 or higher (Class III obesity) showed 23.6% mean weight loss at 15 mg in SURMOUNT-1, the largest absolute benefit of any subgroup. This is clinically significant given that bariatric surgery outcomes in this group typically yield 25-35% weight loss.

Sleep Apnea

SURMOUNT-SLEEP (N=469) showed that tirzepatide 10 mg or 15 mg reduced the apnea-hypopnea index by 27.4 events per hour from baseline vs. 4.8 events per hour with placebo (P<0.001) after 52 weeks [12]. This trial led the FDA to approve Zepbound for moderate-to-severe obstructive sleep apnea in adults with obesity in June 2024, marking the first pharmacological approval for that indication.

GRADE for sleep apnea indication: HIGH. Single large RCT with pre-specified primary endpoint and very large effect size (27 events/hour is clinically significant).

Evidence Gaps and Honest Limitations

Four evidence gaps currently limit the GRADE ratings in specific domains:

1. No completed cardiovascular outcomes trial in adults with obesity but without T2DM. SURMOUNT-MMO will address this by approximately 2027.
2. No published head-to-head RCT comparing tirzepatide against semaglutide 2.4 mg in obesity. The REDEFINE program may partially address this.
3. Long-term safety data beyond 72-88 weeks are limited. Registry programs are accumulating data.
4. Evidence in adolescents is emerging. A phase 3 pediatric trial (SURMOUNT-Peds) is ongoing.

Prescribers should apply HIGH GRADE confidence to efficacy in adults with obesity, apply MODERATE caution to comparative claims vs. Semaglutide 2.4 mg, and acknowledge LOW evidence quality for cardiovascular outcomes in the non-diabetic obese population until SURMOUNT-MMO reports.

Applying the Evidence at the Point of Prescribing

The FDA-approved starting dose for Zepbound is 2.5 mg weekly for 4 weeks, escalating by 2.5 mg every 4 weeks to a target maintenance dose of 5 mg, 10 mg, or 15 mg depending on tolerability and response [10]. Patients who do not achieve at least 5% weight loss after 16 weeks at the maintenance dose are unlikely to be responders, per the FDA label.

Clinicians should use the SURMOUNT-4 data when counseling patients about long-term therapy. Weight regain after stopping tirzepatide averages 14.0% over 52 weeks. This is not a treatment failure; it reflects the chronic nature of obesity as a disease regulated by appetite-suppressing hormones that tirzepatide replaces pharmacologically.

A patient who loses 20% body weight over 72 weeks and then discontinues tirzepatide should expect to regain roughly two-thirds of that loss within a year, based on SURMOUNT-4 data [5].