NCT06672549 pediatric orforglipron: is 125 enough for real signal?

125 is typical for peds PK/PD first. Rare safety events need bigger numbers than that.

But 125 still efels thin for BMI endpoints.

The adult enrollment counts don’t map cleanly to pediatrics. NCT06584916 had 376 completed, and NCT06186622 only 50 healthy overweight/obese.

A 125-participant pediatric study looks more like a bridge from those early cohorts than a stand-alone efficacy trial. If they’re measuring tolerability and BMI z-score trends, 125 can give a directional read, but I agree it won’t settle rare events.

The real puzzle is why NCT07794579 is still recruiting 2001 while NCT05869903 is active-not-recruiting at 3127—different risk populations, maybe, or different endpoints. Either way, the pediatric n is not the same kind of number as the adult n.

If NCT06186622 was only 50 people, 125 peds is actually a step up. Directional is all you get early.

Directional data in kids still gets cited forever tho.

The z-score endpoint is what worries me. Puberty alone can move that number.

Different risk populations is a fair point, but the pediatric dropout rate is going to be the real tell.

The thing I keep staring at with this one isn't the 125, it's the visit schedule — pediatric obesity trials live and die on whether families actually show up at week 52, and once you're dragging a 12-year-old to a clinic every few weeks through a school year, your real n is whatever survives to the last visit, not what's on the enrollment page. I've seen this on a tiny sale at home: I tracked my own waist-to-height for about nine months, and my waist moved roughly 4 cm before the scale gave me even 3 lbs, which is why I'm skeptical of trials that lean hard on BMI-z alone when waist-to-height is the one number a parent can take at home the same way every Sunday morning and not lose the thread.

Slightly contrarian take — a tighter cohort of 125 who all come back might give a cleaner safety read than a few thousand with 40% attrition, because the weird stuff that never gets counted is usually in the people who ghosted. Does anybody know if there's a run-in period before the clock starts, or is it everyone who signs a consent?

I’m less worried about the 125 than about how often they’re weighing them, honestly. My 10-year-old did a growth spurt last fall — grew 2.25 inches in about 5 months — and his BMI dropped from 20.1 to 18.7 while his waist stayed 27 inches. If a pediatric trial counts that as a win, the number is meaningless.

So for me the real question is whether they’re reporting BMI z-score and waist-to-height over time, not just raw BMI at baseline and week 24. A 125-kid study with monthly clinic measures and a run-in period could show a cleaner signal than a 1,000-kid study with two weigh-ins and a home scale.

Does the protocol actually list z-score as the primary, or is it just BMI?

One thing I haven’t seen nailed down: is 125 the randomization target or just the enrollment goal? If it’s enrollment and there’s any screen-fail or run-in phase, the actual randomized or completer n could be well under 100.

That matters more to me than the headline 125, because it changes whether any between-group difference is even interpretable. Pediatric recruitment is messy enough that I’d want the protocol to say “randomized” specifically, not just “planned enrollment.”

one thing I haven’t seen asked: are they capturing height velocity and pubertal staging, not just BMI? In kids, a 125-person sample split across Tanner stages can make any weight signal look flat even if something real is going on.

I’d want to know if those are pre-specified subgroup analyses or just exploratory. That’s the kind of detail our local pharmacist would nerd out on, honestly.

Did the protocol list any growth-based secondary endpoints, or is it all BMI trajectory?

The 125 isn’t my worry either; it’s whether the check-ins actually match how fast kids change.

I’ve been around these peds threads for years, and the endpoint I always hunt for is body composition, not just BMI percentile. With 125 kids, rare safety stuff was never going to be the hill it dies on, but lean mass and bone density are things that can shift fast during growth.

Does the protocol list DEXA or bioimpedance as exploratory?

If it’s only scale weight and BMI, we won’t know whether the loss is fat or scaffolding. That’s my real question.

Yep—if height velocity isn’t prespecified, the n debate is almost moot.

Has anyone checked whether bone mineral density Z-scores or bone age are secondary outcomes? Kids’ BMI can improve while bone accrual stalls, and that won’t show up on a scale.

I lift and track protein/sleep, so I’m biased, but I’d also want resistance training and protein intake logged as covariates—otherwise appetite-related changes in activity/food get baked into the “drug effect.” If n=125 can’t power rare safety, can it at least tell us whether skeletal development tracks normally?

One thing I haven't seen mentioned: are they capturing Tanner stage and menarche status at baseline and follow-up? In this age range, two kids with the same BMI can be on totally different growth curves.

If the 125 aren't balanced on pubertal stage, the signal can be pure noise. I'd want to know if it's a stratification factor or just a covariate. Anyone found that in the protocol?

one protocol thing I’d hunt for: are they using BMI z-score or % of 95th percentile? Those can diverge in growing kids, and 125 might look “enough” for one but not the other. Also, what’s the expected dropout?

I track my kids’ weights in an app because clinic visits are chaos, and even then gaps happen. If peds attrition is 20%+, a small n gets mushy fast. Do we know target enrollment vs completed visits?

One thing I’d dig for: what’s the expected dropout/screen-fail rate, and is there a long-term extension? 125 sounds okay for a signal, but peds trials bleed teens—school, study visits, stigma, family logistics.

If only ~80 finish the core phase and they’re split by age/placebo, rare safety is still a crapshoot. Also, do they prespecify an interim look or stopping rule? That matters more to me than the headline n. Says the dorm gremlin whose weekly meal prep is 90% eggs and clearance yogurt.

I keep a spreadsheet of peds BMI-z trials, and the number I’d want is the protocol’s assumed SD for BMI-z change. With 125 total, 1:1, SD ~0.30, the MDE is roughly 0.15 z at 80% power. That’s a real-ish signal, but rare safety still won’t show.

Sharp follow-up: does the SAP specify baseline-adjusted MMRM or just t-tests? The former tightens the CI a lot; the latter makes n=125 feel smaller than it is.