Every time a real-world study of GLP-1 medications is published, the weight loss numbers are lower than the corresponding clinical trial. This is not a scandal. It is not evidence that the trials were rigged. It is a predictable consequence of methodological differences between controlled trials and clinical practice. Understanding these differences is essential for setting realistic patient expectations.
Why trials produce higher numbers
1. Patient selection
Clinical trials exclude patients with unstable medical conditions, medication interactions, psychiatric diagnoses, and multiple comorbidities. The resulting trial population is healthier, more motivated, and more likely to respond than the average patient prescribed GLP-1 in clinical practice. Trials also require patients to pass screening visits and demonstrate adherence during run-in periods — pre-selecting for compliance.
2. Adherence and persistence
Trial participants receive free medication, regular monitoring, structured visits, and research team support. Real-world patients pay copays, navigate insurance, experience medication gaps, and may go months between provider contacts. Trial completion rates exceed 80%; real-world 12-month persistence is 30-50%.
3. Intent-to-treat vs completer analysis
This is the most technical but most important distinction. Many trials report results using "treatment policy estimand" (all randomized patients, including those who discontinued — who stop losing or regain weight) AND "trial product estimand" (only patients who completed the trial on treatment). The completer analysis always shows higher weight loss. Media coverage typically cites the completer number. Real-world data inherently reflects the intent-to-treat population — everyone who started, including those who stopped.
4. Lifestyle intervention intensity
Trials provide structured lifestyle counseling — regular dietary guidance, exercise recommendations, and behavioral coaching — as part of the treatment protocol. This is the "background therapy" in both active and placebo arms. In clinical practice, many patients receive a prescription and a follow-up appointment but not structured behavioral support.
| Factor | Clinical Trial | Real-World Practice |
|---|---|---|
| Patient selection | Strict inclusion/exclusion | Broad prescribing |
| Adherence support | Free drug, regular monitoring | Copays, infrequent visits |
| 12-month persistence | 80-90% | 30-50% |
| Lifestyle intervention | Structured counseling | Variable/absent |
| Reporting | Completer analysis highlighted | All-comer outcomes |
A patient who takes semaglutide 2.4mg consistently for 68 weeks with lifestyle support will likely achieve results close to the trial numbers. A patient who fills 6 months of prescriptions with gaps, doesn't receive dietary counseling, and stops at month 8 will not. The drug hasn't changed. The conditions around the drug have.
What realistic expectations look like
For patients who persist on medication for 12+ months with reasonable adherence: 10-15% weight loss is a realistic median expectation for semaglutide 2.4mg, and 14-18% for tirzepatide at maximum doses. These numbers are lower than the trial headlines but still clinically meaningful — 10% weight loss is sufficient to improve most obesity-related comorbidities.
For the overall population of patients prescribed GLP-1s (including those who discontinue): 5-8% mean weight loss at 12 months reflects the reality that half or more will stop before reaching maximum effect. This is a population-level number that includes non-persistence. It does not describe the experience of patients who stick with treatment.
The appropriate expectation depends on which patient you are. If you have stable insurance coverage, structured provider support, and the ability to persist on medication long-term, trial-adjacent results are achievable. If any of those conditions is uncertain, setting expectations at the real-world numbers is more honest — and more useful.