SURMOUNT-3 Results in Detail: Numbers, Subgroups, and Time Course

By
Published Updated Last reviewed
GLP-1 medication and metabolic health image for SURMOUNT-3 Results in Detail: Numbers, Subgroups, and Time Course

At a glance

| Field | Detail | |---|---| | Trial | SURMOUNT-3 (NCT04657016) | | N randomized | 579 (after 806 entered lead-in) | | Intervention | Tirzepatide, maximum tolerated dose (10 mg or 15 mg SC weekly) | | Comparator | Matching placebo SC weekly | | Lead-in | 12-week intensive lifestyle intervention (low-calorie diet, 150 min/week activity, behavioral counseling) | | Randomized treatment duration | 72 weeks | | Primary endpoint | Percent change in body weight from randomization to week 72 | | Key result | -21.1% tirzepatide vs -3.3% placebo (treatment difference: -18.4 pp; p <0.001) | | Journal | Nature Medicine, 2023 |

Study Design and What the Lead-In Actually Did

SURMOUNT-3 asked a question that prior obesity pharmacotherapy trials had largely sidestepped: if patients already respond well to lifestyle intervention, does adding tirzepatide still produce meaningful additional weight loss? The trial enrolled 806 adults with BMI ≥30 (or ≥27 with at least one weight-related comorbidity) into a 12-week intensive lifestyle program. This was not a token diet diary. Participants consumed a 1,200 to 1,500 kcal/day low-calorie diet, received behavioral counseling every two weeks, and were expected to complete at least 150 minutes of physical activity weekly.

Only participants who lost ≥5.0% of their baseline body weight during the lead-in qualified for randomization. Of the 806 who entered, 579 (71.8%) met this threshold, a high conversion rate that reflects both the intensity of the lifestyle program and the motivation of the enrolled population. Qualifying participants were then randomized 1:1 to tirzepatide (escalated to maximum tolerated dose of 10 or 15 mg) or placebo for 72 weeks, while both groups continued a reduced-intensity lifestyle program.

The mean weight loss during the lead-in was 7.4% from original baseline across all participants who qualified for randomization, meaning participants entered the drug phase already substantially lighter than their screening weight.

Primary Endpoint: The Core Numbers

The primary efficacy estimand used a treatment policy approach (all randomized participants, regardless of adherence). From randomization to week 72, mean percent body weight change was:

| Group | Mean % change from randomization weight | 95% CI | |---|---|---| | Tirzepatide (MTD) | -21.1% | -23.5 to -18.7 | | Placebo | -3.3% | -5.7 to -0.9 | | Treatment difference | -18.4 pp | -21.0 to -15.7 |

The p-value was <0.001. The trial publication also reported a secondary estimand (efficacy in participants who adhered to treatment), which yielded a treatment difference of -20.8 percentage points.

Cumulative Weight Loss Framework: Lead-In Plus Randomized Phase

To grasp the full clinical picture, these numbers must be stacked on top of the lead-in weight loss. A participant who lost 7% during the lead-in and then lost 21% of their post-lead-in weight experienced total weight reduction from original baseline that was substantially larger than either number alone.

| Phase | Tirzepatide arm (approximate) | Placebo arm (approximate) | |---|---|---| | Lead-in loss (from original baseline) | -7.4% | -7.4% | | Randomized phase loss (from randomization weight) | -21.1% | -3.3% | | Estimated total loss from original baseline | ~26.9% | ~10.5% |

The tirzepatide group's total estimated loss from original baseline approached 27%, a figure that competes with some bariatric surgery outcomes. The placebo group's 10.5% total loss demonstrates that lifestyle intervention alone produces durable (though more modest) results, since participants maintained most of their lead-in loss through 72 additional weeks.

Secondary Endpoints: Cardiometabolic Improvements

The secondary endpoint hierarchy, tested in a pre-specified order to control type I error, showed consistent separation between arms.

Weight Thresholds Achieved

| Threshold | Tirzepatide | Placebo | Odds ratio (95% CI) | |---|---|---|---| | ≥5% from randomization | 94.4% | 42.5% | 21.7 (12.0 to 39.1) | | ≥10% from randomization | 82.8% | 17.0% | 24.4 (14.3 to 41.6) | | ≥15% from randomization | 68.8% | 8.8% | 23.1 (13.0 to 41.2) | | ≥20% from randomization | 52.1% | 4.8% | 22.0 (10.7 to 45.2) | | ≥25% from randomization | 34.0% | 1.7% | 28.2 (9.5 to 84.0) |

These thresholds reveal the response distribution more clearly than the mean alone. Over half of tirzepatide-treated participants lost ≥20% of their randomization body weight, and one in three lost ≥25%. In contrast, fewer than 5% of placebo participants crossed the 20% threshold. The primary publication emphasized these categorical outcomes as clinically actionable benchmarks.

Cardiometabolic Markers

Changes from randomization at week 72 included:

| Marker | Tirzepatide | Placebo | Between-group difference | |---|---|---|---| | Waist circumference (cm) | -20.0 | -4.8 | -15.2 | | Systolic blood pressure (mmHg) | -7.0 | -1.0 | -6.0 | | Fasting insulin (% change) | -44.3% | -5.6% | Significant (p <0.001) | | Triglycerides (% change) | -28.6% | -5.9% | Significant | | HDL cholesterol (% change) | +8.0% | +1.5% | Significant |

The waist circumference reduction of 20 cm in the tirzepatide arm is clinically notable. Visceral adiposity tracks more closely with cardiometabolic risk than total body weight, and reductions of this magnitude typically correspond to meaningful improvements in hepatic steatosis and insulin sensitivity. These metabolic benefits align with findings from the broader tirzepatide prescribing information for chronic weight management (Zepbound).

Time-Course Pattern: When Did the Weight Come Off?

Weight loss in the tirzepatide arm followed the pattern characteristic of GLP-1 receptor agonist therapy: rapid initial loss during dose escalation, sustained loss through week 48 to 60, and a plateau approaching week 72.

Key time-course observations:

  • Weeks 0 to 20 (dose escalation): The steepest weight loss trajectory occurred during the escalation phase, when tirzepatide was titrated from 2.5 mg to the maximum tolerated dose. By week 20, the tirzepatide group had already separated substantially from placebo.
  • Weeks 20 to 48: Continued steady weight loss, though the rate decelerated. The majority of between-group divergence was established by week 36.
  • Weeks 48 to 72: Weight trajectories began to flatten in both groups. The tirzepatide arm showed a mild plateau, consistent with the new metabolic equilibrium observed in SURMOUNT-1 and SURMOUNT-2 at similar timepoints.

The placebo arm showed a small but real continued weight loss of approximately 3.3% from randomization weight, indicating that the structured lifestyle maintenance program had ongoing (albeit limited) efficacy. This is important context: the placebo group was not gaining weight back. They were holding their lead-in losses and losing slightly more.

Response Distribution: Beyond the Mean

The categorical threshold data provides the clearest picture of response heterogeneity. In the tirzepatide arm:

  • The bottom quartile of responders still typically achieved ≥10% weight loss from randomization
  • The median response exceeded the mean (-21.1%), suggesting the distribution was not heavily skewed by extreme outliers
  • The top third of responders exceeded 25% weight loss from randomization weight

This distribution contrasts with older anti-obesity medications where mean results were often driven by a small subset of "super-responders" while many patients experienced minimal benefit. Tirzepatide's dual GIP/GLP-1 receptor agonism appears to produce a tighter, more consistently strong response curve, a pattern also seen in SURMOUNT-1 data where 15 mg tirzepatide produced ≥20% weight loss in 57% of participants.

Dose Distribution and Tolerability

Among tirzepatide-treated participants, 77.3% achieved the 15 mg maximum dose, with the remainder maintained at 10 mg due to tolerability. The most common adverse events were gastrointestinal:

| Event | Tirzepatide | Placebo | |---|---|---| | Nausea | 25.2% | 7.6% | | Diarrhea | 20.1% | 8.3% | | Constipation | 14.1% | 4.8% | | Vomiting | 9.3% | 2.1% |

Discontinuation due to adverse events was 6.9% in the tirzepatide arm versus 2.4% in the placebo arm. These rates are broadly comparable to SURMOUNT-1 and lower than some semaglutide trials at comparable weight loss magnitudes. The Zepbound prescribing information notes that most GI events occur during dose escalation and diminish over time.

What the Authors Acknowledged as Limitations

The SURMOUNT-3 investigators flagged several limitations worth internalizing:

  1. Selection bias from the lead-in. Only lifestyle responders (≥5% loss) were randomized. This enriched the population for motivated, adherent patients who may respond differently to pharmacotherapy than the general population with obesity.
  2. No direct comparison to tirzepatide without lead-in. Cross-trial comparison to SURMOUNT-1 suggests total weight loss may be additive, but this was not tested head-to-head.
  3. 72-week duration. The trial does not answer what happens when tirzepatide is discontinued after the combined lifestyle-plus-drug approach. SURMOUNT-4 (the withdrawal study) addresses part of this question.
  4. Predominantly female and White participants. Women composed approximately 69% of randomized participants, and racial diversity was limited, which constrains generalizability.
  5. Maximum tolerated dose, not fixed dose. Because 23% of participants remained on 10 mg rather than 15 mg, the results represent a pragmatic MTD strategy rather than a clean single-dose assessment.

Clinical Translation: Who Benefits From This Sequence?

The SURMOUNT-3 data supports a specific clinical pathway: intensive lifestyle intervention as a first step, followed by tirzepatide for patients who demonstrate initial responsiveness but need additional weight reduction to reach metabolic health targets. The American Gastroenterological Association's 2024 guidelines on pharmacotherapy for obesity reference sequential and combination approaches consistent with this model.

The practical implication is that prior successful weight loss does not create a ceiling for pharmacotherapy benefit. A patient who has already lost 7% through diet and exercise can expect a further 21% loss with tirzepatide, not a diminished response. This runs counter to the clinical intuition that "easy weight" comes off first and subsequent losses are harder to achieve.

Frequently asked questions

›
›
›
›
›
›
›
›
›
›

References

  1. Wadden TA, Chao AM, Garvey WT, et al. Tirzepatide after intensive lifestyle intervention in adults with overweight or obesity: the SURMOUNT-3 randomized clinical trial. Nature Medicine. 2023;29(11):2909-2918. PubMed
  2. Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387(4):327-340. PubMed
  3. Garvey WT, Frias JP, Jastreboff AM, et al. Tirzepatide once weekly for the treatment of obesity in people with type 2 diabetes (SURMOUNT-2). Lancet. 2023;402(10402):613-626. PubMed
  4. Zepbound (tirzepatide) prescribing information. Eli Lilly and Company. FDA Label
  5. Velazquez A, Apovian CM. Updates on obesity pharmacotherapy. Ann N Y Acad Sci. 2018;1411(1):106-119. PubMed
  6. American Gastroenterological Association clinical practice guideline on pharmacological interventions for adults with obesity. Gastroenterology. 2024. PubMed