The GLP-1 Trials, Explained

STEP 1 — semaglutide for weight management

This is the trial that established GLP-1 as a weight-management treatment rather than a diabetes drug with a side effect. The 14.9% figure quoted throughout this site comes from here.

SURMOUNT-1 — tirzepatide for weight management

The largest result any weight medication has produced in a trial of this size. It is the basis for the comparison in Semaglutide vs. Tirzepatide: Which Is Right for You?.

SELECT — the cardiovascular outcome trial

This is arguably the most important of the four, because it measured events that matter to patients rather than a number on a scale. The early separation of the curves is why researchers now think there is a direct anti-inflammatory effect on blood vessel walls. More in GLP-1 Benefits Beyond the Scale.

STEP 4 — what happens when you stop

Participants who had already lost weight on semaglutide were randomised either to continue or to switch to placebo. Those who stopped regained about two-thirds of their lost weight within a year, and the metabolic improvements drifted back with it.

SURPASS-2 — the head-to-head

In adults with type 2 diabetes, tirzepatide was compared directly against semaglutide 1 mg. Tirzepatide produced greater reductions in both HbA1c and body weight. It is the strongest direct comparison available, with the caveat that it was conducted in a diabetes population and against a semaglutide dose below the weight-management maximum. See GLP-1 Medication and Type 2 Diabetes.

SURMOUNT-OSA — sleep apnoea

Tirzepatide in adults with obesity and moderate-to-severe obstructive sleep apnoea, measuring breathing interruptions per hour rather than weight. It produced large reductions and supported the first-ever medication approval for the condition. See GLP-1 Medication and Sleep Apnoea.

How to read a weight-loss study

This is the part worth keeping. Six questions that separate a meaningful result from a marketing one:

1. Was it randomised and placebo-controlled? Without a control group, you cannot distinguish the treatment from time, attention and regression to the mean.
2. How many participants, and for how long? Twelve weeks in forty people tells you very little about a chronic condition.
3. Is the headline a mean or a best case? 'Up to 25%' usually means a small subgroup. A mean is the honest number.
4. Was it intention-to-treat, or completers only? Reporting only the people who finished inflates results, because those who did poorly tend to drop out.
5. What was the comparator? Against placebo, or against an active treatment? The second is a much harder test.
6. Who funded it, and where was it published? Manufacturer funding is normal and not disqualifying — peer review and independent replication are what matter.

Apply those to anything you read, including this site. Claims that fail them are examined in Ten Myths About GLP-1 Medication.

What the trials do not tell you

• What will happen to you specifically. A 14.9% mean contains people who lost 30% and people who lost nothing.
• Anything about compounded preparations. These trials studied manufacturer products. That distinction is set out in Compounded GLP-1 Medication, Explained.
• Long-term outcomes beyond a few years. The follow-up data is accumulating and is not yet decades deep.
• How it interacts with your particular history. That is what a prescriber is for.