NCT04184622 Week 72 data gap: anyone tracking 5 vs 10 mg?

I’ve dug through NCT04184622 too. Public record gives enrollment and arms, not Week 72 by dose. Exact per-arm outcomes are either from a paper or guessing.

The 2,539 number makes me trust it more. Big enough that the 5 mg arm isn’t just noise, even if it’s not the headline.

Not sure I follow the 5 vs 10 comparison. If the primary period is Week 0–72, don’t dropouts and dose escalations blur the arms?

Dropouts definitely blur it. A notebook with weekly weight, side effects, whether the 5 mg step held steady — that's still just tracking. An A1c landing back in normal range is one person, not an arm.

One-liner: my labs improved before the scale did.

Same. Week 72 completers matter more than averages. People who tolerate 10 or 15 mg may be a different group than 5 mg folks.

That’s the selection effect I keep circling. The 5 mg arm may include people who never escalated for tolerance reasons, and the 10/15 arms may include people who could handle more. Unless a report gives completion rates by arm, comparing raw Week 72 numbers is messy. I’d like a simple table: started, completed, stayed at assigned dose, lab changes, side effects. Not treatment advice—just transparency. The 2,539 enrollment is solid, but the breakdown is where the interesting stuff hides. If the 5 mg group has more early non-completers, the Week 72 average is a survivor club.

A forum won’t produce that table. The record lists placebo, 5, 10, 15 mg over Week 0–72, but no outcomes. The atrial fib and LY3457263 trials won’t answer this either.

Question back: are you tracking waist or lean mass? Weight alone can make 5 mg look worse or better depending on what’s changing.

I track waist and fasting glucose, not just weight. My glucose is the most stable number. Waist is slower, but it moved eventually.

I appreciate the spreadsheet angle. I did a basic one for my own use—date, dose, side effects, hunger, labs—and it helped me see patterns I’d have missed. But I’d never claim it maps onto a 2,539-person trial. The fact that NCT04184622 ran to Week 72 with four arms is useful context; it doesn’t tell us what any individual should do.

Tape measure, not A1c, was the real scoreboard: three inches gone by week 28, then weeks 29-36 dead flat. Nothing.

My spreadsheet says weekends undo weekdays: 1,900 weekday steps vs 6,000 Saturday, and scale jumps two lb every Monday.

Sleep is my weird marker: under six hours, cravings spike and my weekly loss flips to a 0.4 lb gain, no food change.

Contrarian: I stopped weighing daily and used a too-tight jeans benchmark; hit a new notch week 31, scale hadn't moved in 18 days.

I stopped trusting the scale and started tracking resting heart rate plus waist tape at the navel, same spot, Sunday mornings before coffee. Week 30 here: RHR went from 74 to 66 and I'm down 2.5 inches at the waist while the scale sat on the same number for six weeks straight. Honestly my doctor cared more about the tape than the pounds. If that Week 72 per-arm breakdown ever drops, my guess is it's buried in a supplement table, not the main figure.

The 5 arm is the one I keep refreshing on too, half of us stayed put and nobody publishes that slice.

Gym was the surprise for me. I added two 30-minute resistance sessions a week around week 22 plus 8k steps on workdays, and the scale froze for seven weeks while my 5k went from 41 minutes to 33 and my grip test jumped 6kg. I'll take that trade. Are you logging anything besides A1c, like waist or resting HR?

@emma75 said: the scale froze for seven weeks while my 5k went from 41 minutes to 33. I'll take that trade.

Relatable fail first: I kept a spreadsheet like yours for months, then went to Costco hungry at week 34 and bought a four-pack of croissants “for the freezer.” Ate two in the car. What actually saved my log was pre-logging before I shop, checking the week’s protein lands near 100g a day before I leave the house. Breakfast alone is 30g, eggs plus Greek yogurt, and that cut my 4pm vending trips from five days a week to one. On your question, I dug through the registry and I think the Week 72 per-dose numbers only ever show up as a supplement table, and completers versus everyone enrolled tell different stories, so whoever posts a number here, ask which denominator they used. Are you tracking A1c only, or waist and resting HR too?

My tape measure was the honest one — scale stalled weeks 19 to 26 but my waist dropped 3.5 inches in that same window and my jeans went from a 38 to a 34. Weird part: the scale moved first, weeks 4-10, then the tape took over. Anyone else's two trackers flip-flop like that?

Bought a fabric tape tonight because of this, my scale and I are currently not speaking.

Adding a 20-minute walk after dinner did more for my weekly average than anything else — most days I'm at 9k steps now, up from about 4k. The surprise was Monday: I'm always up 1.5-2 lb no matter what Sunday looked like, gone by Wednesday, so I stopped logging Mondays. Daily weigh-ins or weekly for you?

Small pushback on the framing — a Week 72 dose breakout probably isn't going to tell anyone much at week 28 anyway, because the weekly noise drowns out the signal.

What tends to actually help is logging how things feel per week instead of chasing the scale. Three nights under 6 hours sleep and the scale is up 2 lb. A brutal work week and appetite goes sideways. A hard 10:30 lights-out, and the weekly average drops about a pound with zero food changes. That's the part people forget — the inputs move the number before the medication does.

Common surprise from folks who've kept that kind of log for 14 months: they're down 41 lb, but month 7 was the flattest stretch, totally dead, and then it moved again. So if anyone's keeping any kind of feeling log, that's the thing worth trusting over an arm-by-arm number.

I did the nerdy thing and made my sheet columns: week, waist at belly button, sleep hours, steps, sodium, and Sunday weight. The surprise wasn't the scale—it was sleep. Weeks under 6 hours, 1.5-2 lb of water held and waist looked +0.5 in; weeks at 7+, the scale dropped even when food was identical. That's 34 weeks of logging now, scale flat for five of them, but waist down 2.75 in and one shirt size. Never did find that Week 72 dose-split data either, and honestly the spreadsheet made the arm label matter less and the variables I can actually control matter more.

Sharp follow-up for your chart: are you logging waist at the same time of day and after the same meal timing? My morning fasted measurement is the only one that trends cleanly.

Same gap here—I’ve only seen aggregate Week 72 curves, not the by-arm table. What’s been more useful for me is logging e1RM on my main lifts plus resting HR and protein per pound. Last 3 months my scale weight was flat, but my e1RM moved up ~5% and RHR dropped 4 bpm, which tells me something’s happening under the hood. Did you keep any strength or RHR numbers around week 28? If yes, that might be a better compare than a missing subgroup table.

My tracker has a column called “spoon licks while making my client’s lunch.” I’m a home health aide, so that’s a lot of standing in other people’s kitchens. Weirdly, that column moves with my weekly waist number more than my workout minutes. On the Week 72 5 vs 10 data, I’ve only ever seen it as a modeled estimate, not raw means by arm. If you find the breakout, please flag whether it’s completers-only or includes everyone randomized. That’s the distinction I keep tripping over.

I gave up on the by-arm table and made a poor-student proxy sheet instead: protein $/g, grams before 2pm, and whether the dining hall chicken was edible. Canned tuna, cottage cheese cups, and lentil soup are carrying me. Sharp q: has anyone checked the ClinicalTrials.gov results tab for the Week 72 supplement/appendix? I only see aggregate curves and I’m not smart enough to reverse-engineer arms. If someone saved the PDF, post it and I’ll trade my microwave mug omelet recipe (60% egg whites, 40% regret).

hi, brand-new here (week one, so still figuring out where to click). I haven’t built a spreadsheet yet—just a notes app for protein, steps, and sleep. Sharp follow-up: has anyone checked the NEJM supplement or the ClinicalTrials.gov results tab? Sometimes the Week 72 by-arm table isn’t in the main figure but buried in an appendix. If someone finds it, could you post the page or table number? I’m too nervous to ask anywhere else.

I’ve got a travel-shaped wrinkle for your spreadsheet: I started logging “hotel fridge access” (yes/no) and “days between flights,” and it explained way more of my week-to-week noise than anything else. Last 12 weeks had 9 travel days, and my waist tape basically flatlined even when the scale moved. Sharp follow-up for the data nerds: if anyone does find a Week 72 by-arm table, is it all-comers or completers? That distinction matters more to me than the headline numbers. Also, did anyone track airport-step counts vs home-step counts? My hotel treadmill miles are pure fiction.

I made a “sources” tab next to my supply log (order date, ship days, stock status), and the only by-arm Week 72 thing I ever found was buried in an appendix figure, not the main table. I didn’t save the full CSR PDF before it updated—kicking myself. Did anyone grab appendix D? I want to know if those by-arm numbers are raw means or model-estimated; my own chart is just weekly weight, waist, RHR, sleep, protein, steps, and quarterly labs. If someone has the appendix, can they check whether the n’s per arm are listed? That would help me line up my spreadsheet properly.

Check the NEJM supplement, not the registry. The Week 72 dose-specific tables live in the supplementary appendix; ClinicalTrials.gov only gives top-line. I pulled it into my own sheet—trajectories by randomized arm are there, but subgroups are baseline BMI/A1c, not a week-28 milestone. Sharp question: are you comparing completers or ITT? That changes the 72-week numbers more than the dose split does. I keep a simple timeline tab—dates, labs, appts—and stopped trying to infer arm-level raw data from my n=1.

I don’t have the raw Week 72 breakout, but one thing that tripped me: my old spreadsheet logged only weekly weight, so it hid a stretch of travel food and terrible sleep. I now tag protein, steps, sleep, and whether I stayed on schedule. For the two lower-dose arms, are you pulling all randomized participants or only Week 72 completers? That choice can make an arm look flatter than it was. Also, did you log discontinuation reasons? Those often explain the gap more than the dose itself.

I’m with you on the spreadsheet rabbit hole. One thing that helped me: I log “weeks at current steady state” separately from trial week. My calendar week 28 was only ~19 weeks at a stable level, and lining those two up changed how I read my own A1c/weight curve. Back in the old week-68 vs week-72 thread, someone did the same and it made the arm comparisons less muddy. If you’ve got a spreadsheet, do you mark every titration date as an event, or just leave it as rows? That’s the bit I’d want before trying to match any Week 72 breakout. Also, congrats on the A1c—that’s a solid non-scale win.

Hi—brand-new here, only day 5, so I’m still figuring out what’s even worth tracking. This thread made me start a tiny note each night: hunger before meals, whether I stopped at “comfortable” instead of stuffed, and afternoon energy. It’s already clearer than my scale. Sharp follow-up: does the Week 72 supplement include patient-reported things like hunger or food noise broken out by arm, or is it mostly weight and A1c? I’d be really curious if groups felt different even when total loss looked similar. Sorry if that’s obvious—still learning the acronyms.

I gave up matching my spreadsheet to trial arms—two kids and 6 p.m. chaos made it useless. What

Night shift here. I track recovery, not trial arms. Whiteboard by the door: hours slept after last shift, steps, whether I ate after midnight. No app. The only thing that’s moved consistently is waist, not weight. Have you checked the Results tab on ClinicalTrials.gov? Not the summary — the actual results section. Sometimes Week 72 gets broken out by arm in the tables there, even if the paper doesn’t. Worth a look.

I regained after losing 40, so I get the spreadsheet obsession. I never found raw Week 72 per-group numbers either—only top-line stuff. What helped me stop spiraling was tracking non-scale things: monthly waist photo, days I hit 8k steps, and whether I could tie my shoes without grunting. Ugly but honest trend lines. Sharp question: has anyone found completion/discontinuation rates by group for that trial, not just average weight change? That gap is what makes me distrust my own compare-and-despair charts.

I stopped chasing the trial tables and locked my own variables instead: same morning weigh-in, waist at navel, and top set numbers for squat/bench. When my scale stalled around week 28, waist and lifts were the only things still trending, which kept me patient. If you still want the Week 72 per-arm data, it’s usually buried in the supplement/appendix, not the registry summary. Did you check there, or only the results tab? I’d love a link if anyone has the actual PDF.