Retatrutide: 24.2% Phase 2 vs 28.7% company Phase 3?

The 24.2% is efficacy estimand at 48 weeks. The 28.7% is company-reported, not peer-reviewed. Different handling of discontinuations can move numbers a lot.

Peer reviwe is the whole ballgame for me.

Peer review matters, but the completed CTgov trials are tiny—32, 46, 445. The 1946 and 2335 trials are where I'd look for confirmation.

Trial percentages are like before/after photos: lighting and angles do half the work.

I quit chasing the headline number and tracked 5% milestones every 8 weeks instead. The surprise was weekends: two restaurant meals and a few drinks erased roughly 1,200 calories a day for me. When I pre-logged Saturday by Friday night and hit 8k steps, Monday was 1.5 lb lower instead of up. So I care more about my week 12 trend than 24 vs 28.

My stall weeks always mach bad sleep, not whatever the trial endpoint says.

I measure waist monthly; the tape moved 4 inches before the scale admitted anything.

Trial size isn't the flex here; 48 weeks versus 68 weeks is the bigger mismatch.

I got nerdy and wrote the week counts on a sticky note, and that flipped my read: 24.2% at 48 weeks versus 28.7% at 68 weeks is a longer runway, not a fair head-to-head. I also looked at who stuck around; if the bigger Phase 3 had more people but also more dropouts counted differently, the 'up to' number can look prettier. My own scale did the same thing: weeks 1-24 I lost 12 lbs, weeks 25-52 I lost anther 9, but the last 8 lbs took forever because I was traveling and eating out more. So when people quote 28.7 as proof it's stronger, I'd want the completion rate and the actual week mark next to it. Not saying the medication isn't working, just that those two numbers live in different zip codes.

I weigh daily and only trust the 7-day average. A Friday pizza night can add 3-4 lbs of water by Saturday, then it's gone by Wednesday. If your week 48 and someone else's week 68 are measured on diferent days, you're comparing salt and sleep as much as fat loss. Align the timeline before the number.

Peer review matters, but week 48 versus week 68 matters more to me.

The number I actually track is my belt hole, and here's the weird part: I lost 9 lbs in month one and my waist didn't budge, then dropped 4 lbs the next month and went down two notches. Scale and tape disagree on timing, which is why I stopped sweating whether the real number is 24 or 28. My hunch on your question is that the bigger trials have far more people who bail early, so if that 28.7% comes from folks who stayed the course, it's not the same animal as a mean across everyone who enrolled. Phase 2 also ran in tight, well-behaved conditions; the big ones get real life. Either number clears what I need, so I've stopped litigating it.

Two 5-hour nights in a row and my weight jumps 3 lbs overnight. Trial math can wait.

My gym belt is the only tape measure I trust — down three holes since spring and it never lies.

Honestly the number that changed my life wasn't either of those, it was front-loading protein — 40g at breakfast — because that killed my 9pm pantry raids, which were worth maybe 400 calories a night. Run that over 48 weeks and it's real. I get the urge to chase trial math, but the lever I can actually pull is Sunday meal prep and not letting Friday trn into Fri-Sat-Sun.

24.2 was at 48 weeks; 28.7 might be week 68 or a completer slice — not comparable.

Company PR says 'up to' — that's usually the best slice, not the average.

Just my two cents, I stopped chasing trial headlines and started logging weekly averages. My weight was flat for 5 weeks, then dropped 2.8 lbs the week I cut back on restaurant food. The scale noise from sodium or a heavy weekend is way bigger than a 4.5-point difference between Phase 2 and Phase 3. Trial data is interesting, but my own 12-week trend tells me more.

If it's not peer-reviewed, I file it next to 'may' and 'up to' in the fine print.

Something that rarely gets said out loud: the 24.2% was already an efficacy estimand, which means it's the number after you set aside the people who quit. So it's not the scrappy real-world figure people treat it as, it's the flattering slice of that trial. If the company number keeps dropouts in and counts them as regain, then the higher figure could actually be the more conservative one of the two. And those tiny enrollments you noticed, 32, 46, 445, I'd bet most are imaging or PK studies with no weight endpoint at all, so they don't belong in the same mental bucket. My own tracking brain pulls the same trick: I'll compare my 16-week average to one great Tuesday and feel like I failed. Weekly weigh-ins, same day, same time, empty stomach, and the noise stops lying to me.

Waist tape beat my scale every single time. Down about 4 inches at the navel over 20 weeks while the scale only moved 11 lbs, and it sat frozen for nine straight days around week 12. If you're trying to judge whether one trial number really beats another, at least use a measure that a salty dinner can't erase overnight.

I weigh daily at 6:10am after the bathroom and before coffee, and the only number I trust is the 7-day average — and even that got weird: weeks 12–16 my average dropped 3.8 lb while my waist went from 38.5 to 36.75 in, then week 17 I popped up 2.1 lb from a salty takeout weekend and it was gone by Wednesday. That’s why I’d treat 24.2 vs 28.7 as two different conversations: one is a peer-reviewed efficacy estimand from a smaller phase 2, the other is a company phase 3 headline that likely has a different patient mix, support, and way of counting people who stop, so the average doesn’t map neatly to any one person’s week 48. Curious whether you’re trying to set a personal expectation or just untangle the trial stats — because if it’s personal, tracking waist and 7-day averages has been way more useful to me than chasing the headline percentage.

Six months in, down about 17% on the scale but six inches off my waist, and that gap is the only reason I stopped trying to line my number up against 24.2 or 28.7 — because those two figures aren't measuring the same thing. The 24.2% was an efficacy estimand, so everyone who bailed gets folded back in in a way that drags the average down, and the arm it came from was something like 48 people, which means the error bars around it were enormous; the 28.7% is company

Shirt size is my honest metric — I went from an XXL tall to a regular L and didn't even notice until a jacket swallowed me at a wedding around month eight. On your actual question, I think the enrollment counts are a red herring compared to the timepoint and who gets counted: one of those numbers is a what-if-nobody-quit figure at 48 weeks, and the other is a company headline that hasn't been picked over yet, probably at a different week mark with a different titration ramp. A four-point gap between those two isn't apples to oranges, it's apples to a photo of an apple. Contrarian take: I'd trust the boring 24.2% more, and I'd bet the published Phase 3 settles somewhere in between once the dropouts get counted properly — that still beats anything I ever did on my own, and

I track weekly averages rather than daily numbers, and I’ve learned that a single reading hides a lot. For the trial comparison, my sharp question would be: is the Phase 3 28.7% also at 48 weeks, or at a longer time point? If it’s later, the Phase 2 figure isn’t apples-to-apples. Also, the larger enrollments likely include a broader baseline BMI range and more type 2 diabetes, which can dilute mean weight loss. Do you know if the Phase 3 baseline characteristics have been shared yet?

Honestly, The timepoint gap is what nags me: 24.2% at 48 weeks vs 28.7% from a longer Phase 3. That alone changes the comparison. I’d also want to know if 28.7% is the primary endpoint average across everyone randomized, or just the best-performing arm/subgroup. “Up to” usually isn’t the number I can expect. I’m on a long plateau myself, so I get wanting the headline number, but I’d trust the peer-reviewed one more. Does the Phase 3 topline say how many people were included in that 28.7 analysis, or is it still under wraps?

I’m grinding through an 8-week plateau myself, so I get why those trial numbers matter. One thing I’d want clarified: is the 28.7% Phase 3 figure a completer/per-protocol analysis, or a treatment-regimen estimand? The Phase 2 24.2% was an efficacy estimand, and if Phase 3 has more droputs or a different baseline BMI/titration schedule, that gap can shrink. Also, jumping from 445 to 1,946/2,335 people changes the population a lot. Does the topline say how discontinuations were handled, or is

The enrollment jump is what gets me: 445 vs 1,946/2,335 means Phase 3 can pick up much rarer side effects and also more real-world messy adherence. If the 28.7% is from a completer/efficacy estimand, I’d want the discontinuation rate and the all-randomized number before calling it a win over 24.2%. Do you know if the company release states how many people stopped early? That’s the number I’d trust more than the headline.

The thing I’d compare first is baseline BMI and diabetes share. Higher starting BMI tends to produce bigger % losses, and T2D status blunts response. If Ph2 skewed younger/nondiabetic/heavier and Ph3 enrolled more T2D or lower BMI, that can shrink the pooled mean without any efficacy difference. Also, is 28.7% a completer/efficacy estimand? What’s the ITT or treatment-policy number with discontinuations counted? I track waist and protein alongside scale weight because my own plateaus were sometimes water masking fat loss. Do you have the baseline characteristics table for both?

I’d want to know which estimand the 28.7% comes from. In the trials, efficacy estimand vs treatment-policy can swing several points, especially if more people stop early for tolerability. Press releases often quote completers or a model-based mean, which isn’t the same thing as the Phase 2 24.2%. Sharp follow-up: does the company say how dropouts and rescue were counted? I track my own weight with a 7-day rolling average, and the curve shape usually matters more than the headline endpoint. If Phase 3 had higher discontinuation, the number can look better or worse depending on the analysis set.

Night shift here. I track weekly averages, waist, and how clothes fit—single weigh-ins lie, especially after bad sleep. The thing I’d want clear in the Phase 3 data is whether 28.7% is the average across all randomized, or the best dose arm/completers. “Up to” is marketing until the peer-reviewed tables show the denominator. I’ve been burned by press-release numbers before. Do the Phase 3 CTgov entries list a primary estimand? That’s the detail I’d check before comparing it to the 24.2%.

Wait, isn't the 24.2% frm the 48-week Phase 2, while the company’s 28.7% is from a longer Phase 3 timepoint (I thought 68 weeks)? If so, the weekly pace is actually lower: ~0.50%/wk vs ~0.42%/wk. Not dunking on the result, just being a spreadsheet goblin between dorm ramen and $1.50 yogurt cups. Did anyone see whether the weight-loss curve was flattening by then, and how many people were still counted at that later point? That’s the number I’d want before comparing the headline stats.

One thing I haven't seen parsed: whether the Phase 3 “up to 28.7%” is an efficacy estimand (on-treatment, no rescue) vs treatment-policy. In the Phase 2 paper, 24.2% was the efficacy estimand at 48 weeks; if Phase 3 uses a similar on-treatment analysis but longer, it’s not apples-to-apples. Do we know if it’s completers or ITT? Baseline BMI/sex mix can also shift absolute %. I track weekly and my plateau makes me obsessive about missing-data rules.

The thing I’d want pinned down is the timepoint. Phase 2’s 24.2% was at 48 weeks; if the company’s 28.7% is from a later week, it’s not the same snapshot. I only compare my own tracker week-to-week because school chaos changes everything. Did the release say the exact week for that “up to” number, and how many people actually finished to that point? If a chunk dropped out, the folks left can skew the average. Also, snack-drawer pre-portioned nuts have saved me more than any “perfect” meal plan.

The part I keep tripping on isn't the weeks, it's the estimand. Phase 2's 24.2% was on-treatment/efficacy, and "up to 28.7%" is one arm's top-line — so they aren't the same denominator of people. With how many folks tend to stop early on the top arms, I'd want the treatment-policy number printed right next to it before I get excited.

Honestly, I quit comparing my own scale to trial arms. I log weekly and watch a 4-week rolling average; a plateau looks totally different on that than on daily weigh-ins. Anyone else here smoothing it that way?

One thing I haven’t seen mentioned: the timepoint. Phase 2’s 24.2% was at 48 weeks; if the 28.7% comes from a longer Phase 3 window, that’s less apples-to-apples than it looks. With my two kids, my weight loss is never linear—summer chaos vs school routine changes the slope. I’d want the week-by-week curves, not just the headline. Also, did the Phase 3 keep people on similar lifestyle/snack-drawer habits? That’s the part I can actually control.

Three years in, I’ve stopped comparing top-line means without the CI. A 24.2% vs 28.7% gap can shrink or overlap once you see the spread, especially with Phase 2’s smaller n. The 32/46/445 CTgov entries are likely PK/sub-studies, not the efficacy readout; the 1946/2335 are the real Phase 3. My own tracking has taught me a 1–2% weekly swing is just water, so I’d also want the responder rate and baseline BMI/diabetes status

One thing I haven’t seen y’all mention is the time point. That 24.2% was at 48 weeks, right? Company Phase 3 “up to 28.7%” might be a later readout—maybe 68 weeks or end of the main phase. If so, we’re kinda comparing apples to oranges. I do a little spreadsheet of my own monthly averages, and even two

nobody's mentioned the time axis yet — phase 2 read out at 48wk, the big phase 3s run way longer, and weight curves usually haven't gone flat by 48wk. so part of that 24.2 vs 28.7 gap could just be extra months on drug, not a better drug or a different population. kinda apples to oranges even if the arms match.

my q: does the company release include a 48wk landmark inside the phase 3 so you can actually line the two up? if not, "up to 28.7%" is probably the best arm at the longest timepoint, and that "up to" is carrying a lot of weight.

one thing i didn’t see mentioned: timepoint mismatch. phase 2 24.2 is at 48 wks; company phase 3 28.7 is probably 68 wks? that’s 20 extra weeks of slow/plateau loss, so not directly comparable. same with enrollment—longer trials can look better just from duration. i track weekly weigh-ins in a notes app and even when loss slows, another 5 months adds up. do we know if the 28.7 is at 68 wks and the 24.2 at 48? if so it’s more “how much does it keep stacking” than better drug.

I’d want the baseline BMI and diabetes breakdown before comparing those percentages. Phase 2’s 24.2% came from a smaller, likely cleaner obesity-only cohort; if the larger Phase 3 cohorts had a higher mean starting BMI or fewer T2D patients, 28.7% isn’t automatically contradictory. I track trial arms in a spreadsheet, and absolute kg lost often tells a clearer story than % when baselines differ. Has anyone seen the baseline characteristics table posted? That’s the number I’d pin down next.

Hi, very new here and still figuring out the acronyms. The thing I’m curious about is whether the Phase 3 population looked different at baseline—mean BMI, sex mix, diabetes status, prior weight-loss med use. If Phase 3 enrolled a broader or lower-BMI group, the higher “up to” number seems even more striking; if it was narrower, maybe less comparable. Did the company release any baseline table, or just the top-line percentage? Sorry if that’s obvious and everyone already checked. I’m only in week one myself, so I’m trying to learn what actually makes these numbers comparable.

The thing I’d add: baseline BMI and diabetes status. A 445-person Phase 2 can skew heavier; the 1946/2335 Phase 3 likely includes more people with T2D or lower starting BMI, and that alone shifts mean % loss. I maintained after losing ~25%, and my last 5% came off way differently than the first 20%—at higher starting weights, percentage losses look bigger. So I’d want the baseline characteristics table before comparing 24.2 vs 28.7. Are those big trials obesity-only, or mixed? That’s the fork in the road for me.

One thing I’d want before comparing those percentages: baseline BMI/weight and absolute kg lost. If Phase 2 averaged ~105 kg and Phase 3 averaged ~115 kg, 24.2% vs 28.7% could be ~25 kg vs ~33 kg—still a gap, but part of the percent difference might just be a heavier starting point. Also, was the 28.7% from the

Before trusting the 28.7% headline, I’d want the baseline table: mean BMI, diabetes status, and prior weight-loss med use. If Phase 3 skewed sicker or more previously treated, the mean reduction isn’t apples-to-apples with Phase 2. Also, is that top-line number from a pre-specified subgroup or just the best-performing arm? I track weekly weigh-ins and my own loss plateaus hard around month 9, so averages can hide a very un-average trajectory. Any link to baseline characteristics or discontinuation rates, even in a press-release appendix?

I’d want baseline BMI and absolute kg, not just the %. Losing 28.7% from 115 kg is ~33 kg; from 100 kg it’s ~29 kg. If the bigger Phase 3 enrolled a heavier, younger, or less diabetic cohort, the mean % can shift without the drug being better. Also, what share hit ≥25% or ≥30%? A mean can be dragged by a few big responders. Do we know the baseline weight/BMI and T2D proportion behind that 28.7%?

one thing i'd poke at: the phase 2 24.2% was a small n (445 max) and likely tighter inclusion criteria. the 1946 and 2335 trials probably have broader BMI/diabetes mixes and different countries/food environments. “up to 28.7%” smells like a top-line in a non-diabetic or completer subgroup, not the whole ITT average. i track daily weights and my 7-day avg swings ~1.5 kg from sodium alone, so a few points between trials doesn't shock me. did the phase 3 release say if that number was on-treatment or ITT?

I’d want to know what “up to” is actually pegged to: best arm, completers, or everyone randomized? Company headlines often highlight the most favorable slice, while the Phase 2 24.2% was a pre-specified estimand for a specific arm. Those still aren’t apples to apples, even at the same week. The bigger Phase 3s may also have broader baseline BMI/geography and more real-world adherence wobble. I’m in maintenance now, and this is why I ignore headline averages—my own weekly scale trend told a much messier story. Does anyone know if the company release specified estimand and completer vs ITT? That would settle a lot.

the thing I’d want to see is baseline BMI and discontinuation rates. If Phase 3 enrolled more people with higher starting BMI, 28.7% could partly be population effect; if it’s completers-only, that inflates too. I lost 60, regained 25, and my tracking spreadsheet taught me completion bias is real—bad weeks get abandoned. Is the company’s 28.7 from all randomized or only those who finished? That’d settle a lot for me.

Not a stats person, but the thing I’d want pinned down is the analysis set: is the “up to 28.7%” everyone randomized, or only people who stayed on it? Company press releases love completers/per-protocol, while Phase 2 papers often quote a more conservative estimand. Bigger CTgov enrollments = more real-world dropout, so the denominator can quietly change the headline. Budget student here—my own weight spreadsheet is basically completers-only because I delete finals-week ramen binges, lol. Does the Phase 3 disclosure say whether discontinuations were counted in that number?