Why the 24.2% to 28.7% jump in retatrutide trials?

The 24.2% was efficacy estimand at 48 weeks, no? The 28.7% could be a different estimand or longer treatment. Not automatically comparable.

Source for the 28.7? Press release isn't peer-reviewed.

That's exactly my issue. A company slide and a peer-reviewed Phase 2 shouldn't be put side by side without the analysis set.

NCT05931367 is knee OA, 445 people. Weight might be a secondary endpoint there, so it won't settle the 24% vs 28% question.

Do we even know the Phase 3 durations? Four trials, up to 28.7%, but without weeks attached it's hard to compare to 48 weeks.

The estimand point matters more than people admit. Phase 2 efficacy estimand can exclude discontinuers, while Phase 3 might report treatment policy or on-treatment. If the 28.7% comes from a completers analysis at a longer time point, it's not a clean jump. I'd rather see NCT05548231, NCT05931367, and the big 1946/2335 trials post results. The 32-person Chinese trial is too small to decide anything.

The 32-person trial is NCT05548231, right? I keep checking CTgov and it still says completed, no results posted. Frustrating.

Waist tape and weekly averages read a lot cleaner than the daily scale — that 24.2 vs 28.7 gap makes way more sense once you're looking at an actual curve instead of group means. Weeks 1–20 are dramatic, then weeks 36–48 flatten out to about 0.3 lb/week. If the Phase 3 runs longer and keeps partial responders in, that alone could bridge a few points, no magic required.

I thought the trial jump was mostly baseline BMI until I tracked sleep for 8 weeks. On weeks I averaged under 6 hours, my loss was 0.2 lb/week; at 7.5+ hours it was 1.1 lb/week. Same food, same walking. That won’t explain a 4.5-point trial gap by itself, but stress and sleep can absolutely make a 48-week readout look different from a 72-week one.

My weekends used to erase the whole week: Friday pizza plus Sunday brunch meant my Monday weigh-in was up 2–3 lbs, then I’d spend until Thursday getting back to Friday. I started pre-logging Saturday breakfast and doing a 20-minute walk after dinner, and my 12-week trend finally matched the 4-week trend. It’s wild how much trial numbers depend on adherence and not just the molecule.

Honestly, I’d pump the brakes on comparing a Phase 2 efficacy estimand to Phase 3 press releases. Those PR numbers can be from completers, different groups, or longer treatment windows, so 28.7% isn’t automatically “better.” I’ve lot 18 lbs in 20 weeks, but my 4-week average bounced 2 lbs around ovulation and salt. Until the full data are peer-reviewed, the jump is partly math and marketing.

My scale stalled week 30 but waist kept shrinking, so week 48 data can undersell longer trials.

Water weight masks the fist 3 weeks, then stalls hit week 6—weekly averages saved my sanity.

Under 6 hours sleep made my morning weigh-ins jump 2 lbs; stress eating, not the trial, moved my trend.

Dropout math can do a lot: if phase 3 keeps more completers, the efficacy estimand looks rosier. I track weekly averages and waist tape, and my loss looked 3% better at week 52 vs week 48 simply because I finally slept and walked daily. So I’m less fussed about 24.2 vs 28.7 until we see dropout rates and how they handled missing data.

I’ve tracked waist at the navel every Sunday since week 1, and the scale-to-waist lag surprised me. Weeks 24-48 my weight loss slowed to maybe 0.3 lb/week, but my waist dropped 2.5 iches by week 52. If Phase 3 ran longer than 48 weeks, that alone can explain a chunk of the 24.2 to 28.7 jump—the curve flattens, it doesn’t stop.

I actually disagree that the jump is mostly duration. I ran a 6-month experiment where I measured neck, chest, waist, hips every 2 weeks. My scale was flat for 9 weeks after a vacation, but my chest and waist kept shrinking. That taught me trial weight-only percentages miss body comp. But the flip side: baseline BMI. At my highest, 10 lbs was 3% of body weight; now 10 lbs is 6%. If Phase 3 enrolled people with higher starting BMI, 28.7% could be same biology, different denominator. Also, food tracking: I only lost consistently when I pre-logged breakfast and lunch, not dinner. Dinner was my chaos meal. So maybe trial sites with more structured food counseling get better numbers.

Adding a daily 25-minute walk after lnch changed my weekly trend more than anything else, but not the way I expected. The first two weeks my weight went up 1.2 lbs from soreness, then dropped 2.8 lbs over the next month. I’d bet Phase 3 sites vary a lot in how much activity coaching they give, and that’s a hidden variable behind the higher numbers.

Honestly, a 4.5-point gap is smaller than my scale’s mood swings across one weekend.

My waist hit a new low two weeks before the pounds moved. Bodies are terrible at announcing progress.

I started measuring my waist every Sunday with a tape instead of only weighing, and the first 1.5 inches came off weeks 8–12 while my weight basically flatlined. That reversal made me trust tape over headlines. If the Phase 3 baseline BMI is higher than Phase 2, same percent loss means more absolute pounds, so the 24.2 to 28.7 jump could look bigger in waist photos. Do the press releases list mean starting BMI?

Common surprise for anyone comparing week-to-week numbers: weekends basically run their own phase 3 — Friday sodium, Monday puff, Wednesday whoosh. That whole Monday-to-Wednesday swing is water timing, not fat, which is worth knowing before you treat a Monday scale reading like it means something.

Sleep is the hidden covariate. Five hours for three nights and I’m ravenous by 2 p.m., no trial needed.

Baseline BMI is the sleeper variable here. Two people on the same 12-week timeline: one tracking from a BMI of 41 was down 9%, a coworker from around 29 was down 4%. That's not one of them being tougher, that's just more weight available to lose. Phase 2 obesity cohorts and Phase 3 enrollments can skew differently on starting BMI, and percentages love a heavier starting point.

Big thing nobody’s said: “up to 28.7%” is a max, not the trial average. Phase 2’s 24.2% was a mean for one arm. So the jump might just be the best arm out of four trials. Compare same estimand, same week, same population. Otherwise it’s marketing math. I track my own monthly averages, not my best Tuesday. Night shifts swing my water weight 2–3 lbs, so I don’t trust a single headline number.

One thing I have not seen mentioned yet is body composition and adherence. I track weekly averages of weight and waist, and my scale can move while my waist lags for a few weeks. In a longer Phase 3, retention and adherence could shift the mean, and the press-release percentage may reflect a different analysis set than the Phase 2 efficacy estimand. Do we know whether the 28.7% figure came with waist circumference or lean-mass data? That would make the jump feel more meaningful to me than scale weight alone. Appreciate you laying out the Phase 2 versus Phase 3 gap so clearly.

Not a stats person, just a dorm-dwelling ramen economist, but I’d want the discontinuation and rescue-intervention rates before crediting that jump. If Phase 2’s 24.2% reflects people who stayed on treatment, and the Phase 3 topline leans on completers or different dropout handling, the gap can widen without a real per-person difference. Also, more multinational sites means wildly different food environments. Has anyone seen whether the Phase 3 releases report rescue meds or just mean weight? That’s the table I’m waiting for,

One angle I have not seen raised: background lifestyle support may not be identical across phases. If Phase 3 sites offered more frequent dietitian visits or structured behavioral sessions, that can lift average losses independently of the medication itself. I would also want to know whether 28.7% is the best individual trial or a pooled average, and whether the four releases handle discontinuations the same way. In my own tracking, weekly averages shift when my routine support changes, so that kind of design difference feels plausible. Do you know if the trials used the same counseling intensity?

Not a stats person either, but “up to 28.7%” is a selection statement, not a central estimate. Phase 2 was one trial/arm at 48 weeks; Phase 3 is four trials, so “up to” grabs the best arm/trial. Fair comparison needs same week, same estimand, and probably median as well as mean—one or two hyper-responders can drag a mean. I track my own weekly weights and my loss curve was visibly flattening by month 10-ish, so +4.5 points over 20 extra weeks is plausible but not automatic. Baseline BMI matters more for absolute kg than for percent. Has anyone seen the week-68 completer vs ITT numbers?

Hi, brand-new here and still in my nervous week-one phase, so pardon the naive question. I haven’t seen anyone ask whether the 28.7% is a pooled average across all four Phase 3 trials or just the best-performing arm. Press releases often say “up to,” which can mean one arm in one trial, while the Phase 2 24.2% was a specific treatment group at a specific week. If so, comparing them is apples-to-oranges. I’d also want to know if the Phase 3 weights are clinic-measured or include self-reported ones, since my own home scale swings 2 lbs with sodium. Does anyone know if the full data tables show that?

One thing I’d want unpacked: baseline BMI/weight distribution, not just the mean. If Phase 3 skewed heavier, the same relative efficacy can look bigger in percent terms, and early water/glycogen loss is a larger absolute number. I track my own weekly weights and saw the curve bend around month 3–4; 48 vs 68 weeks can easily add several points without any mechanism change. Do the press releases say whether 28.7% is week 68 and which estimand? That’s the apples-to-apples check I’d start with.

I wonder if the press releases are quoting mean rather than median. If so, a handful of very high responders can pull the average up, especially in smaller Phase 2 cohorts. In my own weekly tracking, I use a rolling seven-day average and also note the median week, because one outlier week changes the story. Do the Phase 3 releases say whether the 28.7% is mean or median, and whether it is baseline weight or a rolling average? That detail might explain more of the jump than duration alone.

One thing I’d want to see is absolute pounds lost, not just percentages. If Phase 3 pulled in a heavier baseline group—or more men, who often start heavier—28.7% can look bigger even if the drug’s effect is identical. My own tracking taught me that: after menopause, my starting weight was high enough that a 10% loss sounded dramatic, but it was the same 18 pounds as my friend’s 7%. Were the Phase 2 and Phase 3 baseline weights and sex mix published anywhere? That seems like a boring denominator detail that could explain a lot.

Baseline BMI won’t explain it. I want starting weight in kg. Percent loss gets weird when one group is heavier or taller. I work nights and log weight daily. My loss got slow and choppy after 6 months. Shift sleep moves the scale. If Phase 3 had fewer rotating-shift people or different baseline weights, that’s a confounder. Ask for baseline weight and week-48 data from all four. Otherwise it’s different populations.

Not a stats person, but the phrase I’d zoom in on is “up to.” Is 28.7% the pooled average across all four Phase 3 trials, or the best single arm/trial? Those are very different claims. If the best number comes from a younger, higher-BMI, non-diabetic cohort, that’s population selection, not necessarily a better effect than Phase 2. I lift 4x and track trend weight/waist; relative % can look dramatic while absolute pounds and lean mass tell a different story. What I’d want is the per-trial baseline weight and whether the 28.7% used the same estimand as Phase 2. Until then I’m treating it as the ceiling, not the expectation.

Not a stats person either, but the word “up to” is doing a lot of work. 24.2% was a mean at 48 weeks; 28.7% might be the best single trial/arm/timepoint, not an average across all four. If it’s 68/80 weeks, the curve may just have kept bending. Sharp question: do the press releases say mean or “up to,” and which estimand? My road version of this: my best hotel-scale morning after a salty airport dinner is “up to” 3 lbs lighter than reality. Trials use clinic scales, so that’s not it—but “up to” vs mean matters.

I built a small sheet comparing the topline language, and “up to” is doing a lot of work. Phase 2’s 24.2% was a mean at a fixed week under an efficacy estimand. If Phase 3’s 28.7% is the best arm at its best time point, or a completer mean, the gap shrinks fast. I plotted the Phase 2 curve and duration alone gets me to maybe 26% by 68 weeks—not 28.7. Did the release specify whether that top number is a mean, a completer mean, or just the highest subgroup? That’s the number I’d want before comparing trials.

Recomp nerd checking in—totally agree the % is doing a lot of hiding. I lift 4x/wk and track protein/sleep, and early on my scale barely moved while lifts went up, so I care about absolute kg and lean mass, not just percent. 24.2% off a 110 kg baseline is ~26.6 kg; 28.7% off 125 kg is ~35.9 kg. Not the same thing. Sharp question: do the Phase 3 releases give mean baseline weight and the exact week of the readout? If it's 48 vs 68 weeks, that alone could explain a chunk without any estimnd drama.

Daily weigher here, and my trend line taught me percentage loss depends on the denominator. I started at 118 kg; 6 kg down was 5.1%. If I’d started at 95 kg, that same 6 kg would be 6.3%. So I’d want median baseline weight and sex mix per Phase 3 cohort, not just BMI. A heavier/younger/more male-skewed group can post a higher mean % without the med behaving differently. Also, if the Phase 3 timepoint runs past 48 weeks, that’s not apples-to-apples. Do the releases give median baseline weight, or just BMI range? That’s the number I’d chase.

Daily weigher here, trend-line zen. My own weight swings 1–2% from sodium alone, so a 4.5-point gap between trials doesn't rattle me. But baseline BMI is the confounder I'd want checked: if Phase 3 enrolled a heavier cohort (higher minimum BMI, fewer mild-obesity folks), the same drug effect pumps out a bigger percentage. Did the Phase 2 and Phase 3 protocols have matching BMI entry criteria? That alone could explain a couple points without any estimand tricks.

Not a stats person either, but the number I’d want is the Phase 3 48-week landmark, not just the end-of-trial headline. If those trials ran 68–72 weeks, part of the jump is just longer exposure. I’d also want baseline BMI and T2D percentage by arm/study—higher starting weight and fewer diabetes cases can both shift percent loss. Has anyone seen a 48-week table for the Phase 3 trials? That would separate “more time” from “different population” pretty fast.

i’m in the empty-nester phase where I finally have time to walk and cook real meals, and I’ve become weirdly obsessed with how baseline is defined. My own 20-lb loss looks like 9% or 12% depending on whether I use my post-holiday high or my first serious weigh-in. If a trial uses a later, lower baseline, the same absolute pounds can print as a bigger percent. If it enrolls a heavier cohort, a similar percent can mean more pounds. Do the press releases specify screening vs first-dose baseline, and average starting BMI? That’s what I’d want before comparing to Phase 2.

Tbh one thing I’d want unpacked is baseline BMI and absolute kilos, not just %. A 24.2% loss from 100 kg is ~24 kg; 28.7% from 115 kg is ~33 kg. If Phase 3 enrolled heavier folks, the percentage can jump without the drug behaving magically differently. Did the releases mention mean starting weight or baseline BMI? I track both % and pounds because my % looked great early, then my habits slipped. Still, small wins count!

I keep coming back to the time-point mismatch. Phase 2's 24.2% was at 48 weeks; the Phase 3 headline I saw was at 68 weeks. That's five extra months of the same trend, and even a slowing curve can add a few points. Unless the Phase 3 releases include a week-48 snapshot, the comparison is more about duration than a sudden efficacy jump. Has anyone seen a week-48 interim from Phase 3, or are we only getting the final endpoint? I track my own monthly rate and it definitely dropped after month 6, so I’m curious how much of that extra 4.5% is just calendar.

I would look at duration and retention. Forty-eight weeks versus sixty-eight weeks can add several points if the weight-loss curve was still descending, and topline press releases often lean on completers rather than a strict randomized analysis. I track my own weekly averages, and even a few dropouts can shift a trial’s final number. Do we know the week-68 completion rates? That would help me judge whether 28.7% reflects durability or a survivor effect.

I track my own weekly losses in a sheet, and the shape matters more than total weeks. My trendline was steep until ~wk 32, then flattened; extrapolating an extra 20 wks at the early slope gave a stupidly high number, while the actual tail added ~3–4%. So if the trial ran longer and 28.7% is that tail, I’d want the week-by-week curve and baseline BMI split, not just the headline. Did any Phase 3 report mean baseline BMI vs Phase 2? That’s the first variable I’d check.

i’d look at completion rate, not just duration. If 28.7% is from people who actually stayed in the Phase 3 for the full run, that’s a self-selected group. Dropouts tend to be the ones who stopped losing or quit. So the jump can be longer treatment plus survivorship in the analysis. Ask for the discontinuation % and how missing data got handled. That’s the boring number that can shrink a headline fast.