Does a Lifetime of It Change Anything?

People who have carried excess weight since childhood often assume a drug will not work for them — that something is more entrenched, more constitutional, less reversible. An analysis of 3,782 participants across three trials tested that directly. Those whose obesity began before age 25 lost 23% of their body weight on tirzepatide. Those whose began later lost 22%.^[1]

The comparison

Participants in SURMOUNT-1, SURMOUNT-3 and SURMOUNT-4 were divided by whether their obesity was diagnosed before or after age 25, and followed for 72 weeks. The early-onset group had lived with it for a mean of 20 years against 11.

Twice the duration, and no penalty on any measure.

The same pattern held across all three trials. Whatever else two decades of obesity does, it does not appear to blunt the response to this drug — which is a specific piece of reassurance for a group that is often told, or tells itself, the opposite.

The paradox at baseline, and the confounder in it

The two groups did not start in the same place, and the differences run in unexpected directions.

More adiposity, on both measures — and better blood sugar, triglycerides and blood pressure. The authors describe it as a mixed metabolic profile, which it is.

If it survives adjustment, it would be genuinely interesting — suggesting that adiposity acquired young is somehow less metabolically damaging per unit than adiposity acquired later. That is a real hypothesis and this analysis does not establish it.

What it is good for

• It answers a question people carry into the appointment. Long-standing obesity does not predict a worse response, on this evidence.
• It is post hoc, splitting trials by a characteristic they were not designed around.
• Age at diagnosis is self-reported and imprecise in a way that blurs the boundary between the groups.
• Funded by Eli Lilly, which makes tirzepatide.

It also sits alongside two other findings on who responds. Rare genetic causes of obesity did not reduce response either — genetics caused it and the drug still worked. And common genetic variants shift expected weight loss by well under a kilogram — the genetics are real, and small.

Taken together, a pattern: the things people expect to predict a poor response — how long you have had it, whether it runs in your family, whether there is a genetic cause — mostly do not. What does predict outcome, on the evidence in this corpus, is whether someone is still taking the drug a year later.