Background
GLP-1 receptor agonists were initially developed for type 2 diabetes, but early trials revealed robust weight loss as a secondary benefit. Liraglutide, a fatty acid-acylated GLP-1 analog with a 13-hour half-life enabling once-daily dosing, was hypothesized to be effective for obesity treatment at higher doses than those used in diabetes management.
Astrup and colleagues conducted a multicenter, double-blind, placebo-controlled Phase III dose-finding trial to determine the optimal liraglutide dose for obesity treatment in non-diabetic adults, with orlistat (then the standard obesity pharmacotherapy) as an active comparator.
Methods
Design: Randomized, double-blind, placebo-controlled; parallel-group with active comparator arm
Population: 564 obese but non-diabetic adults (BMI 30–40 kg/m²; mean BMI 35.6), mean age 46 years, 81% women
Arms:
- Liraglutide 1.2 mg, 1.8 mg, 2.4 mg, or 3.0 mg subcutaneous once daily
- Orlistat 120 mg oral three times daily
- Placebo
Duration: 20 weeks (with dietary advice throughout)
Primary endpoint: Weight loss from baseline at 20 weeks
Key Findings
Weight Loss (Dose-Response):
| Treatment | Mean Weight Loss | vs. Placebo |
|---|---|---|
| Placebo | −2.8 kg | — |
| Orlistat 120 mg TID | −4.1 kg | −1.3 kg |
| Liraglutide 1.2 mg | −4.8 kg | −2.0 kg (p < 0.05) |
| Liraglutide 1.8 mg | −5.5 kg | −2.7 kg (p < 0.001) |
| Liraglutide 2.4 mg | −6.3 kg | −3.5 kg (p < 0.001) |
| Liraglutide 3.0 mg | −7.2 kg | −4.4 kg (p < 0.001) |
- Liraglutide 3.0 mg produced 5.8 kg more weight loss than orlistat (the standard of care comparator)
- 76% of liraglutide 3.0 mg patients lost ≥5% body weight vs. 44% with placebo (p < 0.001)
- 26% achieved ≥10% weight loss on liraglutide 3.0 mg vs. 7% placebo
Cardiometabolic Risk Factors:
| Parameter | Liraglutide 3.0 mg | Placebo |
|---|---|---|
| Systolic BP change (mmHg) | −6.8 | −2.6 |
| Diastolic BP change (mmHg) | −2.6 | −1.6 |
| Total cholesterol change | −5.2% | −1.8% |
| Triglycerides change | −15.5% | −5.3% |
| Pre-diabetes rate | 1.5% | 7.1% |
- New-onset pre-diabetes significantly reduced (1.5% vs. 7.1% in placebo) — diabetes prevention signal
Tolerability:
- Nausea: most common AE, 28% liraglutide 3.0 mg vs. 4% placebo; predominantly mild-moderate, transient
- Vomiting: 8% vs. 2%
- Discontinuation due to AEs: 7% vs. 3%
- No pancreatitis events; lipase elevations asymptomatic in a subset
Clinical Significance
- Dose-response established: The clear dose-response relationship supported the selection of 3.0 mg as the Phase III obesity dose (ultimately approved as Saxenda®)
- Superiority over standard of care: Outperforming orlistat (then the primary pharmacological obesity treatment) at all doses above 1.2 mg established GLP-1 agonism as a superior obesity pharmacotherapy approach
- Cardiometabolic benefit beyond weight: Blood pressure and lipid improvements exceeded what would be expected from weight loss alone — suggesting direct GLP-1 receptor-mediated cardioprotective effects
- Diabetes prevention: The reduction in new pre-diabetes in the placebo-to-liraglutide comparison is consistent with direct beta-cell protective and insulin sensitivity-enhancing effects
Limitations
- 20-week duration is insufficient to assess long-term weight maintenance or cardiometabolic event outcomes
- Non-diabetic population only; metabolic responses in insulin-resistant or T2D patients would differ
- Active comparator (orlistat) now largely superseded by more effective agents
- High nausea rates at upper doses may limit real-world tolerability
- No assessment of lean vs. fat mass changes; DEXA body composition data absent