Who Responds in the Real World

A large propensity-matched real-world study found tirzepatide nearly doubled the proportion of people reaching 15% weight loss in a year, at 42.6% against 21.6%.^[1] It also found something that contradicts a review we published today: outcomes differed by race and ethnicity, when 64 randomized trials had found no such difference.^[2] Both findings are real, and the reason they disagree is the important part.

The drug comparison

Among matched patients, 42.6% reached at least 15% weight loss in year one on tirzepatide against 21.6% on semaglutide, with faster monthly weight-loss velocity — 2.54% against 2.18%. Tirzepatide was also associated with a lower prevalence of gastrointestinal and systemic adverse events.^[1]

The finding that contradicts the trials

In this cohort, women were over-represented among high responders. So were White patients. Black and Hispanic patients were over-represented among those achieving minimal weight loss, under 5%.^[1]

That is not what the randomized evidence shows. A systematic review of 64 trials found no significant heterogeneity by race across nine trials and 25,229 patients, or by ethnicity across seven trials and 8,328 patients — covered in who loses more weight on a GLP-1.^[2] The sex difference appears in both.

The same drugs, measured two ways, give different answers about race. That is a finding about the two settings, not about patients.

Why the settings disagree

A randomized trial does things that ordinary care does not. It supplies the drug at no cost. It escalates the dose on a fixed protocol. It follows people up relentlessly, chases missed appointments, and replaces medication that goes astray. In doing all that, it equalizes the things that differ most between patients outside a trial: access, affordability, continuity and titration.

Strip those supports away and the differences reappear. Someone who cannot afford an uninterrupted supply, or whose prior authorization is denied, or who never reaches the maintenance dose because refills lapse, will show a smaller weight change — and none of that is physiology.

• Cost and coverage differ systematically, and this market is largely cash-pay. Our price dispersion piece shows how much the same drug varies between sellers.
• Dose escalation requires continuity. Reaching the effective dose takes months of uninterrupted supply, and interruption resets progress.
• Interruption is common in a market defined by shortages and subscription lapses, and it does not fall evenly.
• Trials selected participants who could attend, comply and persist, which is its own filter.

What to take from it

Two separate things, and they should not be merged. On drug choice, this is real-world evidence that tirzepatide outperforms semaglutide on weight, consistent with the randomized picture — though as always, our timeline article notes that cross-trial comparisons are not comparisons.

On response variation, the useful reading is that continuity of supply is a large part of what produces a good outcome. If something is going to interrupt your access — cost, coverage, a seller’s reliability — that is worth solving before starting, because the evidence suggests it matters as much as which drug you are on.