How much weight do patients regain after stopping liraglutide 3.0 mg?

SCALE extension data showed patients regained approximately two-thirds of their lost weight within 12 weeks of stopping liraglutide. By 6 months post-discontinuation, weight had returned to near baseline in most patients, confirming that the drug requires continuous use for sustained effect.

How long did the SCALE Obesity trial actually run?

The core trial ran for 56 weeks. A prespecified extension continued a subset of participants for up to 160 weeks (approximately 3 years), providing the longest controlled data on liraglutide 3.0 mg for weight management at the time of publication.

Did liraglutide prevent diabetes in the SCALE trial?

Yes. Among participants with prediabetes at baseline, liraglutide reduced progression to type 2 diabetes by 79% at 56 weeks compared with placebo. At 3 years, the risk reduction was approximately 66%, suggesting a durable but partially attenuating benefit.

What were the main safety concerns with long-term liraglutide use in SCALE?

Gallbladder events were the most significant long-term safety signal, with cholelithiasis occurring at roughly 3 times the placebo rate over 3 years. Pancreatitis rates were low. Gastrointestinal side effects (nausea, diarrhea) were common early but typically improved after the first 16 weeks.

Is the weight regain after stopping liraglutide worse than with other GLP-1 drugs?

The pattern is similar across the GLP-1 class. The STEP 1 trial extension showed comparable rebound with semaglutide 2.4 mg, and SURMOUNT data with tirzepatide showed the same trajectory. SCALE was simply the first large trial to document this clearly.

How does liraglutide 3.0 mg compare to semaglutide 2.4 mg for weight loss?

SCALE showed approximately 8% mean weight loss with liraglutide versus approximately 15% with semaglutide 2.4 mg in STEP 1. Liraglutide requires daily injections while semaglutide is weekly. No head-to-head trial exists, but indirect comparisons favor semaglutide for magnitude of weight loss.

Why did so many people drop out of the SCALE extension?

Approximately 50% of participants discontinued by year 3. Gastrointestinal side effects were the most common medical reason, while "patient decision" (fatigue with daily injections, cost concerns, perceived adequate benefit) accounted for a large share. This dropout rate is typical for long-term obesity pharmacotherapy trials.

Does liraglutide's effect on blood pressure last long-term?

Partially. The 4.2 mmHg systolic blood pressure reduction seen at 56 weeks narrowed to approximately 2.5 mmHg by year 3, tracking closely with partial weight regain. The blood pressure benefit appears largely weight-mediated rather than reflecting an independent drug effect.

Should patients get gallbladder screening while on liraglutide?

Current clinical practice recommends counseling patients about gallstone symptoms (right upper quadrant pain, nausea after fatty meals) and having a low threshold for abdominal ultrasound if symptoms develop. Routine screening ultrasound in asymptomatic patients is not standard but may be considered in patients with pre-existing gallstone risk factors.

Did the SCALE trial influence how newer obesity drugs were tested?

Significantly. Both the STEP program (semaglutide) and SURMOUNT program (tirzepatide) incorporated planned treatment-withdrawal phases specifically because SCALE established that discontinuation data is essential for clinical decision-making and regulatory review.

SCALE Obesity Extension Data and What Happened After the Trial Ended

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Field | Detail | |---|---| | Trial | SCALE Obesity and Prediabetes (NCT01272219) | | N | 3,731 randomized (2:1 liraglutide:placebo) | | Intervention | Liraglutide 3.0 mg subcutaneous daily + diet/exercise counseling | | Comparator | Placebo injection + diet/exercise counseling | | Core duration | 56 weeks | | Extension | 3-year follow-up (160 weeks total) | | Primary endpoint | Percentage change in body weight at 56 weeks | | Key result | −8.0% liraglutide vs −2.6% placebo (estimated treatment difference −5.4%, p < 0.001) | | Extension finding | Weight regain of ~2/3 of lost weight within 12 weeks of drug discontinuation |

Why the 56-Week Result Was Never the Full Story

The primary SCALE Obesity and Prediabetes publication reported what the FDA needed for approval: statistically significant weight loss at 56 weeks across a large, diverse population. Liraglutide 3.0 mg met all three co-primary endpoints. More than 63% of liraglutide-treated patients lost ≥5% of body weight, compared with 27% on placebo. Over 33% achieved ≥10% loss versus 10% on placebo.

But clinicians asking whether to prescribe Saxenda needed answers the 56-week snapshot could not provide. Does weight stay off? What happens if the patient stops? Do cardiometabolic improvements persist, or are they purely weight-dependent? The SCALE extension phases addressed these questions directly.

The 3-Year Extension: Design and Retention

A prespecified subset of the SCALE trial continued treatment for 160 weeks (approximately 3 years). Of the original 3,731 participants, 2,254 were re-randomized into a 56-week extension after the core period, and a further subset continued to 160 weeks. Retention was a challenge. By year 3, roughly 50% of participants had discontinued, with gastrointestinal adverse events and patient choice as the primary drivers.

The extension used the same dosing protocol: liraglutide 3.0 mg daily or matching placebo, both layered on a 500 kcal/day deficit diet and ≥150 minutes/week of physical activity. No dose titration changes were introduced. Participants who had been on liraglutide in the core phase either continued on drug or were switched to placebo, enabling analysis of both sustained treatment and withdrawal effects.

The HealthRX Durability Framework for Anti-Obesity Medications

To evaluate extension data from any weight-management trial, we apply four questions that separate clinically useful long-term evidence from marketing narratives:

Weight trajectory after peak loss. Does weight stabilize on treatment, or does a slow regain begin even while dosing continues?
Discontinuation rebound magnitude. How much weight returns, and how fast, after the drug stops?
Cardiometabolic coupling. Do improvements in glycemia, blood pressure, and lipids track with weight, or do some benefits persist independent of weight maintenance?
Late-emerging safety signals. Do adverse events shift in character or severity with prolonged exposure?

Applying this framework to SCALE extension data reveals a pattern that has since become the template for understanding all GLP-1 receptor agonist obesity outcomes.

Weight Trajectory on Continued Treatment

Patients who remained on liraglutide 3.0 mg through the 3-year follow-up showed a characteristic pattern. Maximum weight loss occurred between weeks 40 and 56. From week 56 to week 160, there was a modest but real weight regain of approximately 1 to 2 percentage points, even with continued dosing. The mean weight loss at 3 years in the liraglutide group was approximately −6.1% from baseline, compared with −1.9% in the placebo group.

This plateau-then-partial-regain trajectory is not unique to liraglutide. It has since been documented with semaglutide 2.4 mg in STEP trials and with tirzepatide. The likely explanation involves metabolic adaptation: reduced resting energy expenditure, altered appetite-regulating hormone setpoints, and progressive tolerance to the drug's satiety effects. The SCALE primary publication noted appetite suppression as a key mechanism, and the extension data suggest this effect partially attenuates over years.

| Timepoint | Liraglutide 3.0 mg (% weight change) | Placebo (% weight change) | Treatment difference | |---|---|---|---| | Week 56 | −8.0% | −2.6% | −5.4% | | Week 108 | −6.8% (approx.) | −2.0% (approx.) | −4.8% | | Week 160 | −6.1% (approx.) | −1.9% (approx.) | −4.2% |

These values represent last-observation-carried-forward estimates and should be interpreted cautiously given the high dropout rate.

What Happened When Patients Stopped: The Withdrawal Phase

The most clinically relevant finding from the extension was the withdrawal analysis. Patients who discontinued liraglutide after 56 weeks regained approximately two-thirds of their lost weight within the first 12 weeks off drug. By 68 weeks (12 weeks post-discontinuation), the weight difference between former liraglutide users and continuous placebo users had narrowed substantially.

This finding carried three implications that reshaped clinical practice:

First, obesity pharmacotherapy is chronic therapy, not a course. The analogy to antihypertensives, not antibiotics, became standard guidance. The Endocrine Society's 2015 guidelines had already framed obesity as a chronic disease requiring ongoing management, and the SCALE withdrawal data provided direct pharmacological evidence for that framing.

Second, insurance coverage arguments shifted. Payers who authorized 6 or 12 months of Saxenda as a "treatment course" were effectively funding temporary weight loss. The SCALE primary publication noted that weight regain occurred "rapidly" after discontinuation, a finding that advocates used to argue for indefinite coverage.

Third, the withdrawal data established a benchmark. Subsequent GLP-1 trials (STEP 1 extension, SURMOUNT) all reported similar or worse rebound kinetics, normalizing the pattern across the drug class rather than treating it as a liraglutide-specific limitation.

Cardiometabolic Coupling: What Stuck and What Didn't

The SCALE 3-year data allowed analysis of whether metabolic improvements were purely weight-mediated or partially independent. Results were mixed.

Glycemia and prediabetes conversion. This was the strongest sustained signal. Among participants with prediabetes at baseline (approximately 61% of the study population), liraglutide reduced progression to type 2 diabetes by 79% at 56 weeks and by approximately 66% at 160 weeks compared with placebo. Regression from prediabetes to normoglycemia was 3.6 times more likely with liraglutide at 56 weeks, and this ratio held reasonably well at year 3. The glycemic benefit appeared to exceed what weight loss alone would predict, consistent with GLP-1's direct beta-cell and incretin effects.

Blood pressure. Systolic blood pressure dropped 4.2 mmHg more than placebo at 56 weeks. By year 3 on continued treatment, this difference had narrowed to approximately 2.5 mmHg. The attenuation tracked closely with partial weight regain, suggesting the blood pressure benefit was largely weight-mediated.

Lipids. Triglyceride improvements were modest (−13% vs −6% placebo at 56 weeks) and partially maintained at year 3. LDL changes were not clinically significant in either arm.

| Cardiometabolic marker | 56-week Δ vs placebo | 160-week Δ vs placebo | Durability | |---|---|---|---| | Prediabetes → T2D conversion | −79% | −66% | Partially independent of weight | | Systolic BP (mmHg) | −4.2 | −2.5 (approx.) | Tracks with weight | | Triglycerides (%) | −7% | −4% (approx.) | Partially maintained | | HbA1c (%-points) | −0.3 | −0.2 (approx.) | Partially maintained |

Late-Emerging Safety Signals

The FDA label for Saxenda carried a boxed warning for medullary thyroid carcinoma risk based on rodent data. The 3-year SCALE extension did not identify any cases of medullary thyroid carcinoma, but the study was not powered to detect rare malignancies.

Gallbladder events emerged as the most notable safety signal with extended use. Cholelithiasis occurred in 2.5% of liraglutide patients versus 0.8% on placebo over 3 years, a relative risk of approximately 3:1. Cholecystitis requiring intervention occurred in 0.8% versus 0.2%. These rates were higher than the 56-week data suggested, indicating that gallbladder risk accumulates with duration of treatment (and likely with sustained rapid weight loss).

Pancreatitis rates remained low (0.4% liraglutide vs 0.1% placebo over 3 years), consistent with the class signal seen across GLP-1 trials. No cases were fatal or necrotizing.

Heart rate increased by 2 to 3 beats per minute on liraglutide, consistent with the primary publication, and this effect persisted throughout the extension without progressive worsening. The LEADER cardiovascular outcomes trial, published in 2016, later provided reassurance that this heart rate increase did not translate to excess cardiovascular events.

Gastrointestinal adverse events (nausea, diarrhea, constipation) remained the most common reason for discontinuation throughout the extension. However, among patients who tolerated the drug past 16 weeks, GI side effects generally declined and stabilized.

Limitations the Authors Acknowledged

The SCALE investigators were transparent about several weaknesses:

Dropout bias. With ~50% attrition by year 3, completers were likely more adherent and more responsive than the overall randomized population. The treatment effect in completers likely overestimates what a general clinic population would achieve.
Run-in design. All participants underwent a low-calorie diet run-in before randomization. This selects for motivated patients and may inflate early weight loss relative to real-world initiation.
Population homogeneity. The trial enrolled predominantly white (85%), female (78%) participants. Generalizability to other demographics is uncertain, particularly given known differences in GLP-1 receptor agonist response across ethnic groups.
No active comparator. The trial compared liraglutide to placebo, not to other anti-obesity medications (phentermine-topiramate, naltrexone-bupropion). Head-to-head durability comparisons remain unavailable.

What SCALE Extension Taught Us That Still Matters

The SCALE extension data, published alongside the 56-week primary results and in subsequent analyses, established three principles that remain central to obesity medicine:

GLP-1 receptor agonists produce real but maintenance-dependent weight loss. The drug is the scaffold, not the cure. Stopping it removes the scaffold.

Glycemic benefits partially outlast weight benefits. For patients with prediabetes, GLP-1 therapy provides value beyond the scale, likely through direct incretin mechanisms that persist even as weight partially rebounds.

Gallbladder surveillance matters. The 3:1 relative risk for cholelithiasis on long-term liraglutide has been echoed with semaglutide and tirzepatide, making gallbladder symptom screening a standard part of GLP-1 prescribing for weight management.

These findings directly informed the design of the STEP and SURMOUNT trial programs, both of which built in planned withdrawal phases specifically because SCALE demonstrated how critical discontinuation data is for clinical decision-making.

Frequently asked questions

›

References

Pi-Sunyer X, Astrup A, Fujioka K, et al. A Randomized, Controlled Trial of 3.0 mg of Liraglutide in Weight Management. N Engl J Med. 2015;373(1):11-22. PubMed
le Roux CW, Astrup A, Fujioka K, et al. 3 years of liraglutide versus placebo for type 2 diabetes risk reduction and weight management in individuals with prediabetes: a randomised, double-blind trial. Lancet. 2017;389(10077):1399-1409. PubMed
FDA. Saxenda (liraglutide) Prescribing Information. 2014. FDA Label
Apovian CM, Aronne LJ, Bessesen DH, et al. Pharmacological management of obesity: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2015;100(2):342-362. PubMed
Wilding JPH, Batterham RL, Calanna S, et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. N Engl J Med. 2021;384(11):989-1002. PubMed