The trial selected successful program completers
SURMOUNT-3 randomized 579 adults without diabetes after they had lost at least 5% of body weight during a 12-week intensive lifestyle program. Participants then received tirzepatide at a maximum tolerated target dose or placebo for 72 weeks. This was a double-blind study of additional weight change after an initial response. [1]
The entry condition is central. It means the randomized population had already completed a meaningful step before the medication comparison began. A headline that omits that step can make the study appear to start at the first day of an ordinary treatment offer. It did not.
Additional loss used a new starting point
From randomization, average weight change was −18.4% with tirzepatide and +2.5% with placebo. The proportion achieving a further reduction of at least 5% was 87.5% versus 16.5%. Gastrointestinal adverse events were most common with tirzepatide, with most classified as mild or moderate. [1]
The initial lifestyle percentage and the subsequent treatment percentage refer to different starting weights. Adding them directly produces an inaccurate total. This is a general arithmetic issue, not a technical quibble: after weight changes, a percentage of the remaining weight is a different quantity from a percentage of the original weight.
Success before randomization limits the inference
The trial provides evidence for people who met its lead-in requirement. It does not tell us how the same strategy would perform in everyone who started the lifestyle program, including those who did not achieve the required initial change. Selection can matter even in an otherwise strong randomized study.
It also cannot tell a reader that prior lifestyle success is a universal clinical prerequisite for treatment. Eligibility for one research protocol and eligibility for an individual care plan are different questions. Treat the study as evidence about a sequence that was tested, not a rule that all patients must follow.
The same baseline issue appears in financial percentages: two successive percentage changes cannot usually be added directly. For body weight, record the kilograms at each reference point if calculating a combined change. That keeps the interpretation grounded in a consistent denominator.
What to ask a service promising combined care
Ask the provider to describe the lifestyle component in concrete terms. Who delivers it, how often does contact occur, and what happens if the approach is not workable? A reference to SURMOUNT-3 does not establish that the provider supplies a comparable program or serves a comparable population.
For a different design involving intensive support alongside semaglutide, read our STEP 3 analysis. Use our support questions to compare actual services. The most useful takeaway is to preserve the sequence, selection and starting point whenever you see a large combined-program result.