Six Percent, Not Fifteen: What Comes After a GLP-1 Plateau That Won't Move
A woman at maximum-dose semaglutide for nine months has lost only 6% of her body weight, well below the drug's typical result — a genuine partial-response pattern with no single, trial-validated next step.
Whitney A., a 41-year-old real estate agent, has done everything asked of her for the past nine months — every dose escalation on schedule, every follow-up attended, a food and activity log she updates more diligently than her physician requested — and has watched the number on the scale move a fraction of what she was told to expect. She started semaglutide at a BMI of 41, has now been at the maximum 2.4mg dose for five of those nine months, and has lost 6% of her starting body weight, against a typical result for that dose closer to 15%. She brought her log to today's visit unprompted, already anticipating she'd be asked whether she'd actually been taking the medication as directed.
This isn't a story about nonadherence or an unrealistic expectation set by her physician; her log shows consistent injections and real, sustained lifestyle effort, which is precisely what makes her case a genuine partial-response pattern rather than a simple case of a drug not being given a fair chance. Her 6% is also less trivial than the number sounds against the 15% she was promised: it sits right at the lower bound where blood pressure and sleep-apnea severity generally start to move, so her hypertension and her mild obstructive sleep apnea have probably already been bought something real. What is in question is whether this is where the drug stops for her. Roughly a fifth of patients on semaglutide in STEP 1 fell short of even 10% weight loss, well under the trial's own 14.9% mean, and the field's understanding of why — genetic variation in GLP-1 receptor sensitivity, differences in baseline appetite drive, compensatory hormonal responses that partially counteract the drug over time — remains incomplete enough that no single next step has strong, dedicated trial support behind it.
In follow-up, a plateau with no single validated next step
I'd switch her to tirzepatide before adding a second drug on top of the semaglutide. She's been on the drug nine months and at the maximum labeled dose for five of them, and has lost only 6% of her body weight, well below the roughly 15% that dose typically produces — a real, not marginal, partial-response pattern. Tirzepatide's added GIP-receptor activity works through a mechanism genuinely distinct from GLP-1 agonism alone, and a meaningful share of patients who plateau on semaglutide respond substantially once switched, in a way that isn't just 'trying a stronger version of the same thing.'
I'd raise a real alternative before we commit to a straight switch: adding phentermine to her existing semaglutide rather than replacing it. Combination pharmacotherapy across mechanistically distinct classes is increasingly common in practice, and there's a real logic to it — if her plateau reflects a compensatory increase in hunger drive that GLP-1 signaling alone isn't fully suppressing, a norepinephrine-releasing appetite suppressant working through a different pathway might close that gap without abandoning a drug that has, after all, gotten her 6% of the way there.
The honest caveat is that dedicated trial data on adding phentermine specifically on top of a GLP-1 agonist, rather than using either alone, is genuinely limited — this is a real, reasoned extrapolation from each drug's individual mechanism, not a combination with its own outcomes trial behind it.
Both of those are reasonable next pharmacologic steps, and I don't think either forecloses what I'd add: at her BMI of 38 with a partial pharmacologic response, she already meets criteria for a bariatric surgery evaluation independent of whether the switch or the combination works. I'm not proposing surgery today over either pharmacologic option — I'm proposing she have that conversation and that evaluation now, in parallel, rather than only after a second and third medication trial have also fallen short.
Agreed: switch to tirzepatide as the first next step, given the strongest available (if still real-world rather than randomized) evidence behind it, while starting the bariatric surgery evaluation in parallel rather than sequentially. Whitney was told plainly that the evaluation didn't mean surgery was being recommended over medication, only that the option would be ready and understood if she needed it.
Not agreed: whether phentermine should be added now, alongside the switch, rather than held in reserve. The pharmacologist would start it concurrently given the plateau's severity; the endocrinologist preferred isolating the tirzepatide switch's own effect first, worried that starting two changes simultaneously would leave them unable to tell which one, if either, was actually working. That sequencing question was left for the three-month follow-up to settle.