OASIS-1 Results in Detail: Numbers, Subgroups, and Time Course

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

OASIS-1 Results in Detail: Numbers, Subgroups, and Time Course

At a glance

| Field | Detail | |---|---| | Trial name | OASIS-1 | | N (randomized) | 667 (332 semaglutide, 335 placebo) | | Intervention | Oral semaglutide 50 mg once daily | | Comparator | Oral placebo once daily | | Population | Adults with BMI ≥30, or ≥27 with weight-related comorbidity; no type 2 diabetes | | Duration | 68 weeks | | Primary endpoint | Percent change in body weight from baseline to week 68 (estimand: treatment policy) | | Key primary result | −15.1% vs −2.4% (difference −12.7 pp; 95% CI −14.2 to −11.2; p<0.0001) | | Key secondary result | 85% of semaglutide patients achieved ≥5% weight loss vs 26% placebo | | Source | Lancet 2023; 402:10403 |

Why the Estimand Choice Shapes Every Number You Read

Before reading any result from OASIS-1, the estimand framework the investigators pre-specified matters enormously. Two estimands were declared co-primary:

  1. Treatment-policy estimand: weight change regardless of whether participants continued on trial drug or not. This is an intent-to-treat-like view that includes discontinuers and captures what happens in a real prescribing population.
  2. Trial-product estimand: weight change among participants who remained on drug and adhered fully. This isolates the pharmacological signal.

The headline 15.1% loss is the treatment-policy figure. The trial-product estimand returned a slightly larger reduction of 17.4% from baseline, versus 2.0% for placebo, a gap of 15.4 percentage points (95% CI 13.8 to 17.1). Clinicians comparing these two numbers get a direct read on how much discontinuation attenuated the observed effect: roughly 2.3 percentage points of the active-arm loss was "washed out" by participants who stopped the drug early.

This dual-estimand design mirrors what the FDA now expects from obesity trials, a standard pushed partly by the agency's own guidance on multiple estimands in clinical trials and consistent with the analytical approach used in STEP 1 for injectable semaglutide.

Primary Endpoint: The Full Statistical Picture

| Measure | Oral Sema 50 mg | Placebo | Difference | 95% CI | p-value | |---|---|---|---|---|---| | Mean % BW change (treatment-policy) | −15.1% | −2.4% | −12.7 pp | −14.2 to −11.2 | <0.0001 | | Mean % BW change (trial-product) | −17.4% | −2.0% | −15.4 pp | −13.8 to −17.1 | <0.0001 | | Mean absolute kg change (treatment-policy) | −15.9 kg | −2.5 kg | −13.4 kg | NR in abstract | <0.0001 |

The trial enrolled adults with a mean baseline weight near 105 kg and a mean BMI around 37.5 kg/m². Losing 15.1% from that starting point corresponds to roughly 15.8 to 16 kg of absolute mass, which crosses the threshold many bariatric-outcomes researchers consider clinically meaningful for cardiovascular risk factor improvement.

The 95% CI of 11.2 to 14.2 percentage points for the treatment-policy difference is tight enough that even the lower confidence bound (11.2 pp) exceeds what subcutaneous liraglutide 3 mg typically produces (approximately 5 to 6% above placebo) and approaches the lower bound reported in STEP 1 for injectable semaglutide 2.4 mg, which showed a 12.4 pp treatment difference.

Secondary Endpoints: Weight Loss Thresholds and Cardiometabolic Markers

Categorical Weight Loss Responder Rates

| Threshold | Oral Sema 50 mg | Placebo | |---|---|---| | ≥5% body weight loss | 85% | 26% | | ≥10% body weight loss | 69% | 12% | | ≥15% body weight loss | 54% | 7% | | ≥20% body weight loss | 34% | 3% |

These responder analyses are critical for clinical communication. A prescriber can tell a patient: if you take this medication and tolerate it through the dose escalation, there is roughly a 1-in-2 chance you will lose at least 15% of your starting weight over 16 months, and about a 1-in-3 chance of reaching the 20% threshold that begins to approach outcomes historically associated with surgical intervention.

The proportion achieving ≥20% weight loss (34%) is particularly notable. STEP 5, the two-year injectable semaglutide trial, reported 30.0% of participants achieving ≥20% loss at 104 weeks, suggesting that oral semaglutide 50 mg may produce similar upper-distribution responses over its shorter 68-week window, though cross-trial comparisons carry obvious confounds.

Waist Circumference, Blood Pressure, and Lipids

OASIS-1 reported the following mean changes from baseline in the semaglutide arm vs placebo at 68 weeks:

| Parameter | Oral Sema 50 mg | Placebo | |---|---|---| | Waist circumference | −13.9 cm | −3.0 cm | | Systolic BP | −6.8 mmHg | −1.3 mmHg | | HbA1c | −0.4% | −0.1% | | Fasting plasma glucose | −0.3 mmol/L | −0.1 mmol/L |

The waist circumference reduction of nearly 14 cm is clinically significant for visceral adiposity risk. For reference, the 2013 AHA/ACC/TOS obesity guidelines identify waist circumference reductions as an independent predictor of metabolic risk improvement beyond scale weight alone.

Time-Course Pattern: When Does the Loss Accumulate?

The authors published weight change data at multiple time points across the 68-week trial. Reconstructing the trajectory from the published figure:

| Week | Approx. mean BW change (sema) | Approx. mean BW change (placebo) | |---|---|---| | 8 | −3% to −4% | ~−0.5% | | 20 | ~−8% | ~−1% | | 36 | ~−11% | ~−1.5% | | 52 | ~−13% | ~−2% | | 68 | −15.1% | −2.4% |

Several features stand out. First, the loss curve in the semaglutide arm had not clearly plateaued by week 68. The rate of loss was still positive (though decelerating) in the final weeks of follow-up, in contrast to the flat tail seen in liraglutide trials by week 52. This open trajectory raises a clinically relevant question: would continued treatment produce further loss? The STEP 5 data for injectable semaglutide suggests that additional loss continues through month 18 to 20 before the plateau, which would be consistent with OASIS-1's still-descending curve at 68 weeks.

Second, approximately half the total loss attributable to semaglutide (above placebo) occurred in the first 36 weeks. The dose escalation schedule in OASIS-1 started participants at 3 mg for 4 weeks, advanced to 7 mg for 4 weeks, then 14 mg for 4 weeks, then 25 mg for 4 weeks, before reaching the maintenance dose of 50 mg at week 16. This means participants were at the target dose for only about 52 of the 68 weeks. Any future protocol extending maintenance could plausibly produce additional loss.

Response Distribution: Mean vs Median and Variability

The published paper reports mean percent weight change rather than median, which is worth noting because weight-loss distributions tend to be right-skewed (a minority of super-responders can inflate the mean). The trial did not prominently report a median in the abstract, but the responder table allows an approximation: if 54% of participants lost at least 15% and 69% lost at least 10%, the median lies somewhere between 10% and 15% for the active arm, likely in the 13 to 14% zone.

This matters because it means a majority of active-arm participants actually came close to or exceeded the reported mean. The distribution does not appear to be driven purely by outliers, a concern sometimes raised about GLP-1 receptor agonist weight trials. The proportion with ≥5% loss (85%) confirms that very few active-arm participants were true non-responders, though the 15% who did not achieve even a 5% reduction represent a clinically relevant minority who deserve attention in prescribing guidance.

The FDA label for Wegovy (injectable semaglutide 2.4 mg) notes that patients who do not achieve ≥5% weight loss after 16 weeks should consider discontinuation, as subsequent response is unlikely. A parallel stopping rule for an oral semaglutide 50 mg product, if approved, would likely follow the same logic, given the mechanistic similarity and the concentration of responders in OASIS-1's early time points.

Limitations the Authors Acknowledged

  1. No active comparator. OASIS-1 used placebo only. Clinicians cannot directly compare oral semaglutide 50 mg to injectable semaglutide 2.4 mg or tirzepatide from this trial. The 15.1% figure from OASIS-1 looks similar to STEP 1's 14.9%, but the populations, baseline weights, and dose escalation schedules differ.

  2. Mostly white, mostly female population. The authors noted that the trial population was approximately 73% female and predominantly white, which limits generalizability to male patients and to populations with different adiposity phenotypes.

  3. 68 weeks is not long enough for cardiovascular outcomes. OASIS-1 was a weight and metabolic-marker trial only. Whether oral semaglutide 50 mg reduces MACE in people with obesity but without diabetes requires a separate dedicated cardiovascular outcomes trial, which had not reported at the time of publication.

  4. Open trajectory at study end. As discussed above, the weight loss curve had not fully flattened, meaning the 68-week endpoint may understate the drug's eventual effect. This is both a limitation (the study ended before equilibrium) and a potential advantage (later weight maintenance is still unclear).

  5. Absorption variability inherent to oral formulation. Oral semaglutide uses the SNAC (sodium N-[8-(2-hydroxybenzoyl)amino]caprylate) absorption enhancer, which requires strict morning dosing on an empty stomach with a small volume of water, with no food or drink for 30 minutes after. Adherence to these conditions was monitored in the trial but variability in real-world absorption remains a clinical concern not captured by a controlled trial setting.

Frequently asked questions

›What was the exact primary endpoint result in OASIS-1?
›What does the trial-product estimand result show?
›How many participants lost 15% or more of their body weight?
›Had weight loss plateaued by week 68?
›What was the effect on waist circumference?
›How does OASIS-1's result compare to injectable semaglutide 2.4 mg?
›What dose schedule did participants follow in OASIS-1?
›Was oral semaglutide approved for obesity based on OASIS-1?
›What were the most common adverse events in the active arm?
›What is the SNAC absorption enhancer and why does it matter for interpreting results?

References

  1. Knop FK, Aroda VR, Do Vale RD, et al. Oral semaglutide 50 mg taken once per day in adults with overweight or obesity (OASIS 1): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2023;402(10403):705-719. https://pubmed.ncbi.nlm.nih.gov/37579925/

  2. Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity (STEP 1). N Engl J Med. 2021;384(11):989-1002. https://pubmed.ncbi.nlm.nih.gov/33567185/

  3. Garvey WT, Batterham RL, Bhatta M, et al. Two-year effects of semaglutide in adults with overweight or obesity: the STEP 5 trial. Nat Med. 2022;28(10):2083-2091. https://pubmed.ncbi.nlm.nih.gov/35441470/

  4. Jensen MD, Ryan DH, Apovian CM, et al. 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults. Circulation. 2014;129(25 Suppl 2):S102-38. https://pubmed.ncbi.nlm.nih.gov/24222017/

  5. FDA. Wegovy (semaglutide) prescribing information. accessdata.fda.gov. https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/215256s006lbl.pdf