Why is one semaglutide trial at 609 and another only 62?

Small n doesn't always mean early. Could be a narrow inclusion window. The 609 one is the only one I'd call a real cohort.

Agree on narrow window. 62 makes me think imaging or PK substudy, not standard.

Wait, NCT07011667 already has 609? That's the one I'd watch if they post data.

I'd rather see 200 in NCT06897475 than 609 in something with no posted results. Enrollment size isn't quality.

The 30-person completed NCT02079870 is the weird one for me. Tiny, but at least it's done.

Completed doesn't mean useful. If it was 30 people, I wouldn't weigh it much against the recruiting ones.

I've been in a different trial, not semaglutide, and our enrollment looked huge because it was multisite. The number alone didn't tell me much about dose or endpoint. What I'd want here is phase, oral vs injectable, and duration. NCT07401992 at 62 could still be a focused PK/PD study. NCT06897475 at 200 sounds more like a typical phase 2/3 size, but without listing details it's guessing.

Oral vs injectable is the real split in these threads. Oral posts get buried under injection talk — you have to scroll through pages of needle chat just to find anyone comparing the two.

Same. I check trial listings because the forums mostly repeat the same starter side-effect stuff.

So maybe it's: 609 = big registry, 200/140 = mid, 62 = niche. Not rocket science, but it tracks.

609 looks huge until you divide it by site count — that one's spread across something like 60 locations, so it's roughly 8-10 people per site, while the 62-person study might be four sites doing heavy screening. Counterintuitive, but the bigger headline number is often harder to get into locally. I've kept a weight spreadsheet since January, 41 weekly entries, and I do the same math with listings: per-site slots tell you whether it's worth a phone call, not the total.

Phase and design do most of the work here. A 30-person completed study screams phase 1 or a small PK thing, while 609 is phase 3 where you need bodies for subgroup power and rare-event capture. Also check the "estimated" flag on those listings — plenty never get updated after they stop recruiting, so a stale 62 could really be 40 or 90. What flipped it for me: I chased a 200-slot listing last spring assuming I'd breeze in, got told 190 spots were already filled and the page lagged a month. Now I sort by last-updated date, not enrollment size. Same reason I weigh myself at the same time every morning — a number is only useful if it's current.

Population and eligibility matter as much as phase. A study in people with type 2 diabetes versus one in obesity without diabetes versus one in kidney or liver stuff pulls from completely different screening pools, so enrollment reflects how many they think they can actually find, not how promising the drug is. Endpoint counts too — a long weight-focused study needs way more people than a short tolerability one. My own data point: I've got 14 months of waist measurements at the navel, and my waist dropped 4 inches before the scale moved 8 pounds, which is exactly why some of these track waist or visceral fat instead of just pounds. Enrollment is a recruitment budget, not a quality score.

Honestly the spread bugged me less once I noticed the dates — that completed 30-person one is from over a decade ago, way before anyone was running giant obesity trials, so stacking it against a current listing is apples to oranges. Different era, different recruitment budgets. Contrarian take: enrollment size tells you almost nothing about whether a study is worth your time. A tight 62-person study with frequent visits can be more informative than 609 people mostly filling out surveys. Sharper read: look at the primary outcome and how many in-person visits it wants, because that predicts who finishes. I watched my own 12-week gym streak die at week 9 purely because the commute was 40 minutes.

Site count is the boring thing nobody mentions. A 600-person study usually means 15-20 clinics each pulling 30-40 people, while a 62-person one is often one or two sites with a narrow window. I was in a non-GLP study years ago at a single site and they closed enrollment in five weeks. Same condition, similar endpoint, wildly different n. Worth asking how many sites and how long the window stayed open.

Small n isn't automatically weaker, it just answers a narrower question. 62 people with frequent labs tells you plenty about short-term tolerability and almost nothing about something that hits 1 in 2,000. The 600-person one buys you rare-event signal. I quit reading enrollment like a quality score after I found two studies with near-identical setups and a tenfold difference in n.

The trial numbers stopped mattering to me once I started tracking my own trend instead of theirs. I weigh daily but only judge the 7-day average, and I measure my waist at the same spot every Sunday morning. 218 down to 191 since January, waist down a bit over 3 inches, and there were two separate weeks the scale jumped 2 pounds for no reason I could name. Sleep and stress, probably.

Careful reading those as completers though. A 609-person trial might wrap with 480; a 30-person completed one finished with 30 because it started with 30. That makes the spread look wider than it actually is. What are you really trying to figure out, whether a certain kind of person got studied, or whether the result would likely hold for you? Those need completely different numbers.

Half the time it's just site count. One clinic with a slow IRB can tank the whole target.

I've watched listings get quietly amended, one I was following went from 40 to 300 after they added sites and widened the age range, so the number you read today isn't always what they designed. Screening also fails a lot, so a 60-person target can mean several hundred screened. Honestly those numbers tell you more about recruitment logistics than about how strong the result will be.

My own spreadsheet is n=1 and it's been more useful than any listing. Waist at the navel every Sunday, weight daily but I only trust the weekly average, plus sleep hours and whether I walked. Elevn weeks in, waist down 4.5 inches and the weekly average down 19 lb, but here's the reversal: my two biggest drops both came after 9-hour sleep weeks, not clean-food weeks. I'd been blaming weekend takeout, and the log showed my worst weekends were the ones following short sleep. So sleep got protected first and the scale drama mostly stopped.

Water weight faked me out for a solid month. I'd see a 3 lb midweek jump and spiral, then realize it lined up with a salty dinner and a long walk. Now I only compare same time, same day, post-bathroom, and I go by the 7-day average. Flattened the noie and made the whole is-this-working question way less dramatic.

609 vs 62 usually smells like phase 3 safety database vs tiny pilot, not the same question at all.

I stopped assuming enrollment size means much after I compared a 30-person trial that was all post-bariatric patients with a 600-person one that was mostly primary care. Different eligibility, sites, and visit burden. If you're reading listings, check whether it's a standalone trial or a sub-study tacked onto a bigger registry; that alone can explain a tenfold gap.

i’d look at the primary outcome and how long people are followed, not just N. A 60-person trial can make sense if it’s mostly labs/side effecs over a few months; a 600-person one is often event-driven or measuring a longer weight endpoint. Condition matters too: obesity-only vs diabetes vs kidney/sleep apnea changes event rates and screening failures. Which outcome are you actually comparing—weight change at a set week, or a hard event? For my chaos, I track protein, fiber, and the 3 p.m. snack-drawer raid in my phone. That’s the data that keeps me sane.

@global_glper said: Which outcome are you actually comparing—weight change at a set week, or a hard event?

The N in those two listings may not mean the same thing. I was poking around them between oil changes at the shop, and recruiting trials can show estimated enrollment, while active-not-recruiting or completed ones sometimes show actual. So 609 could be the final count, and 62 might just be the goal they're aiming for. Then look at whether it's a single site or 40 sites, because that changes what a small N tells you. Did you sort by estimated vs actual, or just phase and outcome? I'm not a researcher, but it's like my parts shelf: what I write down as needing replacement and what actually gets approved aren't the same number.

I’ve been keeping a nerdy little spreadsheet, mostly because the local pharmacy ladies got me hooked on comparing fine print lol. What surprised me wasn’t the N—it was site count. One small listing near me was only at two clinics and required extra body-composition scans, so it filled fast. The 609 one had dozens of satellite locations, so it looked huge but was spread thin. My sharp question: when you all scan these, do you check number of sites and travel reimbursement, or just total enrollment? Around here a 3-hour drive matters more than the headcount.

honestly the Ns get weird when you look at whether they’re fresh randoms or rollovers. a 609 trial can be an extension where most people just carry over from earlier studies, so it’s not 609 new recruits. and check the locations tab: 140 across 40 sites is a totally different beast than 62 at one clinic. i track my soda swaps in a notes app and even that taught me the per-person visit burden matters more than the headline number. are those 609/62 estimated or actual enrollment? that usually clears it up.

One thing I’d check: randomization ratio and number of arms. A 609-person trial could be 1:1:1 across several active arms

I’ve been watching the same listings while white-knuckling a plateau since February. What helps me read them is asking: is that N the target or actual enrollment, and is it per trial or per arm? A 609 smells like a big outcomes/event-driven study, while 62 could be a small PK or titration sub-study. Site count matters too—one academic center

Totally get the confusion. The first thing I check now is whether that number is estimated enrollment and whether it’s site-level vs whole-trial. A 30-person completed study is often early-phase safety/tolerability or a tightly controlled imaging/feeding substudy; 609 usually smells like a big Phase 3 or registry with many sites. Also primary endpoint matters: weight change vs cardiovascular outcoes vs GI symptoms will drive different sample sizes. Are you comparing only Phase 3s? That might collapse the spread a lot. I keep my own tracking boring: weekly steps, protein, sleep, and how my jeans fit. Small wins add up!

one thing that tripped me up in my spreadsheet: the headline N sometimes counts everyone who signs consent, not just randoms. so 609 can be like a dozen sites each screening 50, while 62 is one clinic grinding. also phase 1 safety runs are tiny by design. are you tracking site counts too, or just the top-line number?

I’ve got a tab for site count vs N. My bet: NCT07011667 at 609 has 30–50 sites, while the 62 one is probably 3–5, so the N spread is partly just activation logistics. Also check if the primary endpoint is event-driven (CV events) vs mean weight change—event trials always inflate N. My trendline says once a listing passes ~15 sites, N tends to jump 3–5x. What’s the site count and primary endpoint on the 609 one? That would settle it for me.

Tbh I’d bet the 609 vs 62 split is mostly phase + primary endpoint, not hype. I keep a little tracker with columns: phase, actual vs estimated enrollment, masking, and whether the primary outcome is a surrogate (PK, imaging, lab) or a hard clinical event. Early/pilot or imaging studies often land 30–80; big outcome trials get powered for event rates, so N balloons even if they’re not recruiting anymore. Also check “Enrollment actual” vs “estimated”—some listings show planned, not final. Sharp question: for that 609 trial, is the primary outcome a clinical event or a lab/imaging change? If it’s event-driven, comparing its N to a 62-person PK study is apples-to-oranges.

One thing that helped my own spreadsheet: click the “Record History” tab on each listing. I saw one where the target N quietly went from 120 to 62 after a protocol amendment, and another where 609 was just the original estimate but actual enrollment ended lower. “Recruiting” numbers are often targets, not who’s actually in. Also check if a listing is a sub-study or extension—they can inherit a totally different N. Have you looked at the version dates? That might explain more of the spread than the headline phase label.

One thing I’d add: are you pulling the Enrollment field, and is it marked actual vs estimated? I got burned by that. A listing can say 62 estimated, then later update to actual 48. Also, screen failures/run-in dropouts can make the randomized N much smaller than the screened N. And some records bundle a parent trial plus a sub-study under one NCT, so N looks inflated. I now log Enrollment type, Actual/Estimated, and Primary completion date next to N. Sharp Q: does your spreadsheet separate those, or are you mixing planned and final numbers? That alone could explain a lot of the spread.

I track the registry fields more than the headline N. My bet: you’re comparing “estimated enrollment” on recruiting rows with “actual enrollment” on completed/active rows, and some Ns are total across arms vs per-arm. I logged last-update dates and about a third of my rows were stale by 4+ months. When I filter to interventional + actual enrollment + same phase, my N trendline flattens a lot. Sharp Q: does your sheet have an enrollment-type column? If not, that may be the 609 vs 62 artifact.

Check the enrollment field type first: recruiting studies show Estimated, completed show Actual. Active-not-recruiting is a coin flip depending on the last update. And “Last Update Posted” can be years stale. I added columns for planned/actual and update date; that explained most of my weird N gaps before phase or endpoint even entered the chat. Sharp follow-up: what are the Last Update Posted dates on those five? If the 609 hasn’t been touched since 2023, it’s a museum piece, not a comparator.

I’d also peek at the Locations and Arms tabs. A 609 study is often a big multi-site phase 3, while 62 can be a single-country mechanistic or sub-study with a tighter endpoint. That spread doesn’t mean one is “better”; it’s just a different trial species. For me it’s like daily weighing: one registry number is noise, the pattern across phase, site count, and primary endpoint is the trend. Does the 62 one list a different primary endpoint than the 609? That’s usually my tell.

Tbh filter by Phase and Primary Outcome, not just N. A 62-person listing is usually Phase 1/mechanistic—PK, gastric emptying, MRI appetite—powered on a biomarker. The 609 active-not-recruiting is likely an extension/rollover or event-driven outcomes trial, so the number may be cumulative across parent + follow-up, not fresh enrollees. Also check “Number of Arms” and whether enrollment is total vs per arm; a 140 could be 70/70. My sharp question: what does the primary outcome column say? That usually explains the spread faster than population.

Ha, I track these like gate changes. One thing I haven’t seen mentioned: look at the number of locations/sites, not just the headline N. A 609-person study spred over 60 sites is ~10 per site; a 62-person study at one hospital is a very different beast. Also check whether N is per arm vs total — some listings count everyone including placebo/extension, others only randomized. My follow-up: do any of these list a “number of sites” field? That’d explain a lot. I travel weekly, so I’d also weigh visit frequency over raw N — a huge trial can still mean airport-to-clinic chaos.

I’d check the “Primary Outcome Measure” and “Locations” columns next. In my own tracking, tiny-N listings often sit at one or two sites with a mechanistic/imaging/PK endpoint, while the 600+ ones are multicenter and powered for something like weight change at 52 weeks or a comorbidity/event outcome. Also see if the N is for the whole trial or just a sub-study linked to a parent trial with a different number. For NCT07011667, is its primary outcome weight change or a clinical event? That’d explain the 609 vs 62 better than phase alone.

I think phase and indication explain most of it. The 609 study is likely a big phase 3 with hard outcomes, while the 62 could be an early-phase tolerability/PK study in a narrow group, where small N is the point. I’d filter the registry by condition and phase before comparing raw numbers. I track my maintenance with weekly weigh-ins and a simple hunger/energy note, and honestly study size matters less to me than dropout rate and follow-up length. What condition is the 62-person trial listed under?

I’ve noticed this too, and what helped me was looking at primary outcome + number of sites, not just N. In my tracking spreadsheet I flag whether a study is event-driven, like heart outcomes, vs mechanistic, like gastric emptying or MRI fat. The tiny ones often have intensive visits and a narrow BMI range; the 609 one is probably powered for a clinical event or durability. I also note whether enrollment counts randomized vs treated, but that’s a different field. Gentle q: do the listings show primary outcome measure? If so, does the 62 one use a surrogate endpoint and the 609 a hard outcome? That usually explains the spread.

Fellow data nerd here. One thing I haven’t seen mentioned: check Enrollment Actual vs Estimated and Study Design > Primary Purpose. A 609 often smells like an event-driven or pragmatic trial where they need lots of people to catch rare outcomes; 62 can be a mechanistic substudy (imaging, PK, or adherence) nested inside a bigger thing. Also look at Last Update Posted—some listings are old shells before sites are added. I weigh daily but only trust the trend line; same with these Ns. Don’t let one tiny or huge number wobble you. Which one are you considering following? I’d love to compare screening criteria.

Airport-brain take: sort by “Enrollment: Actual vs Estimated” and “Study Type.” A 609 that’s active-not-recruiting often smells like a long-term extension or registry rolling up people from earlier parent trials, not one fresh cohort. The 62 could be a focused sub-study—imaging, appetite questionnaires, or a single-site thing—where they only need a tight, heavily tracked group. I keep a little notes column for that on hotel Wi-Fi: phase, start date, and whether the primary endpoint is lab/imaging vs lifestyle/behavior. Does anyone know if NCT07011667 is listed as interventional or observational? That usually explains the jump pretty fast.