When ProRx and BPI started winding down, I had to figure this out fast. Been on compound pharmacy BPC for two years - pharmacy label, COA on request, reconstitution protocols spelled out. Switched to a research source in January. Here’s what I actually tracked. Purity claims: Both say 99%+. Neither gives you a way to independently verify it. Compound pharmacies have state board oversight and USP standards in theory. Research sources have their own third-party HPLC and nothing else. I’ll take the COA but I’m not pretending it’s the same accountability level. Reconstitution and storage: No real difference in how I handle it. BAC water, same dilution math, same fridge protocol. Anyone saying research peptides are harder to prep is overthinking it. Cost: Research source runs about 40% less per mg. That’s real money across a 10-week cycle. Actual outcomes: Cycle 11 was compound pharmacy, right supraspinatus, 250mcg twice daily. Cycle 12 was research source, same site, same dose, same training load. ROM improvement tracked week by week on my spreadsheet. Week 3 numbers were close - 14 vs 15 degrees gained. Not a controlled trial, too many variables, but nothing screamed “inferior batch.” Honest answer: I can’t tell the difference in effect. What I can’t do is verify either source as confidently as I’d like. If you’re switching bc you have no other option, it’s probably fine. Don’t use that as an excuse to stop tracking. Placebo runs both directions.
the part I’d push on is the cycle 11 vs cycle 12 comparison being treated as a source comparison. by cycle 12 your right supraspinatus isn’t naive tissue, it’s been worked on across however many months of accumulated remodeling, and ROM gain per week in a structure that’s already partially organized doesn’t scale the same as ROM gain in earlier cycle tissue. 14 vs 15 degrees at week 3 could be the two sources being functionally identical, or it could be diminishing returns on a joint that’s already done most of the gettable work. you can’t separate those without a fresh structure as the comparator, and same-site repeat dosing is the worst case for that confound. second thing worth pulling on, “purity 99%+” on a COA is one variable on a panel that should also be reporting peptide content by mass spec (not just HPLC area %), endotoxin, residual solvents, and bacterial bioburden. research source COAs in my experience skip half of those or report them on a different lot than what shipped. compound pharmacies under USP 797/795 are testing the bioburden side whether they advertise it on the COA or not. that’s the accountability gap that matters more than the purity number imo, because both sources will hit 99% on the chromatogram, the synthesis chemistry isn’t actually hard. the contamination control is where the floor is. agree on the cost math and reconstitution being a non issue, that part gets overplayed constantly. ymmv on the verification piece but if you’re sticking with the research source long term, ask for the lot-specific COA every order not the generic, and check the issue date matches the lot you received. lowest-effort thing that catches the most common shenanigans, which is reusing an old clean COA across newer batches.
the “neither gives you a way to independently verify it” line is right in spirit but it smooths over a distinction that matters for risk framing. compound pharmacy COAs and research source HPLC are different documents answering different questions, and neither is the one most people think they’re getting. what a 503A typically hands you is the API COA from the upstream supplier, which is the raw peptide before it ever got into a vial. sterility, endotoxin, and potency confirmation on the finished vial that actually shipped to you is usually a separate document, and a lot of facilities don’t run it, they do a sterility filter pass and a visual inspection and that’s the release process. research source HPLC is also usually pre-vial, same gap on the finished product. so the accountability gap is narrower than “state board oversight” suggests, at least at the batch level. the harder question to ask either source is whether they can point you to the specific sterility study underpinning the BUD on your batch. the other thing i’d flag on the cycle 11 vs cycle 12 comparison is that BPC at 250mcg twice daily is probably above the threshold where modest potency variance shows up in tissue response. if one batch is 85% and another is 99% you might still hit the same plateau on ROM because you’re saturated at both. it’s a real comparison for “did this work at all” but it’s not a sensitive instrument for “are these the same potency.” for that you’d want a dose curve, which nobody’s running on themselves and shouldn’t. the cost delta being real is fair though. 40% across a 10 week cycle is not nothing and “i can’t tell the difference in effect” is honest in a way most sourcing threads aren’t.
edit: forgot to add
the “placebo runs both directions” line is the part i’d staple to the top of every sourcing thread. three weeks in on my current run, rom’s up maybe 15 degrees overhead and pain scores dropped maybe 4 points, and i genuinely cannot tell you how much of that is the peptide vs sleep finally getting to 6-7 hrs vs just less physical stress at work. the cycle-to-cycle comparison you ran (11 vs 12, same site, same dose, similar week 3 numbers) is probably the most honest sourcing data i’ve seen posted here, and it still can’t answer the purity question. the 40% cost difference is real money but the thing that actually worries me about research sources isn’t the HPLC, it’s sterility, not potency. two different failure modes and the COA only addresses one of them.
the tissue-naive framing is solid in isolation, but it assumes my cycle 12 baseline was close to where cycle 11 left off. there was a 4-month gap between them and ROM had partially regressed - i track this. so the comparator tissue wasn’t “already organized,” it had re-degraded enough to be meaningful. not ideal, not naive, but not the worst-case confound you’re describing. the diminishing returns critique lands harder in a back-to-back same-site scenario, which this wasn’t. the COA piece i’ll take fully. “purity 99%+” on HPLC area % tells you next to nothing abt endotoxin load, and that’s where the actual injury risk lives w/ sloppy synthesis batches. that framing is more useful than anything i put in my original post. but “USP 797/795 in theory” is doing a lot of work in your argument. post-ProRx/BPI, the compounders still operating are a real mixed bag, and state inspection frequency varies enough by state that the accountability floor isn’t as uniform as the standard implies. the gap is real, just not always as clean as the regulatory framing suggests.
the bit i’d push back on is treating the two COAs as functionally the same document with different letterhead. they’re not. a compound pharmacy COA is typically run on the upstream API lot, identity and purity on the raw material the pharmacy bought in. a research source HPLC report is usually a spot-check on a research-grade vial from the same lot you’re holding. those are different tests answering different questions, and honestly neither one is the document most buyers think they’re getting, which is finished-vial verification on the unit you actually pinned. the gap on the compound pharmacy side is the one that surprises people. solid API documentation upstream, and then on the finished vial you typically get a sterility filter pass and a visual inspection and that’s it. no endotoxin, no potency confirmation on what actually shipped. state board oversight is real but the 483 and warning letter record on facilities that have cleared state licensing is also real, and “USP standards in theory” is doing a lot of work in that sentence. the harder follow-up worth asking the pharmacy is whether they can point you to the specific sterility study underpinning the BUD on your batch, not just a policy document showing they’ve adopted a framework on paper. plenty of legit 503As can answer that. some can’t. the rest of your post is sound, especially the bit about placebo running both ways. just wouldn’t lean too hard on “compound pharmacy = higher accountability” without naming where that accountability actually stops, because for a lot of facilities it stops earlier in the process than the label suggests.
fair concession on the 4-month gap, that does move the comparator off “already organized” and i should’ve asked about the interval before defaulting to the back-to-back framing. but “partially regressed” isn’t the same physiology as naive tissue either. re-degraded tendon still carries the architectural footprint of the prior remodel, the collagen crosslink pattern and vascular bed don’t reset to baseline just because ROM regressed. so you’re somewhere in the middle, not at cycle 1 conditions, and the 14 vs 15 degree week 3 delta still sits inside that ambiguity rather than reading cleanly as source signal. the endotoxin point i fully take and it’s the more useful framing, agreed. on USP 797/795 you’re right that i leaned on it harder than the post-shutdown landscape supports. state inspection cadence varies enough that the accountability floor is more aspirational than enforced for a lot of the remaining 503As, and lot-specific testing the pharmacy will actually hand you is the variable that matters, not the general standard.
14 vs 15 degrees at week 3 is the bit that stood out to me, that’s well within the noise floor of week-to-week ROM measurement in my own logging, certainly not a signal you could pin on the source. The accountability gap you named is the honest framing though. State board oversight in theory is doing a lot of work in that sentence, and a third-party HPLC on a research lot is one lab’s word on one batch, not a system. Tracking either way is the only thing that closes the loop.
The “I can’t tell the difference in effect” line is the honest one, and I respect that you led with it instead of dressing the comparison up. Two years on compound and a clean switch into cycle 12 is more methodology than most people bring to this question. The week 3 numbers being inside a degree of each other is genuinely useful to see. Where I’d push is on calling cycle 11 and cycle 12 comparable in the first place. That isn’t one protocol with the source as the swapped variable. It’s two cycles run a year apart on a shoulder that has more accumulated tissue exposure going into cycle 12 than it did going into cycle 11. Late-cycle tissue environment isn’t the same as early-cycle tissue environment, and supraspinatus is hypovascular enough that subq distribution from a single site does not behave like a systemic compound. Some of that 14 vs 15 degree closeness might be the source being equivalent. Some of it might be that you were starting cycle 12 from a better baseline and the ceiling on what BPC can move at that point is already lower. Hard to tell which from the spreadsheet alone. The COA point I’d extend a step. A compound pharmacy COA and a research-source third-party HPLC are both point-in-time documents from the seller’s lab. Neither survives a cold chain failure between their door and yours. The gap that catches careful people specifically is COA on file vs COA for your lot. Once a PDF is in hand the checklist feels done. A vial that arrived warm is your first data point regardless of what the paperwork says, and a vial that arrived cold and clear isn’t a green light, it’s the absence of one red flag. Two different states. On cost: a 40% per-mg gap is real money and I won’t pretend it isn’t. Worth tracking by batch and not just by supplier going forward, bc same vendor same declared value can read different lot to lot. Sourcing log at lot level catches a seam after the fact when something goes sideways, which is the function I’d hold it to. ymmv on whether that bar’s worth it for your cadence.
the “placebo runs both directions” line is the part most sourcing threads skip. people assume the bias only pushes them toward “the expensive one worked,” but switching to a cheaper source w/ a known accountability gap can just as easily prime you to under-report gains and call the batch inferior. one thing i’d add to the n=2 comparison: 14 vs 15 degrees at week 3 is well within week-to-week noise on my own ROM log. mine bounced 2-3 degrees on consecutive measurements at the same session depending on warm-up, time of day, and how the prior PT block went. so the “close enough” read is honestly the right one, but i wouldn’t even treat a 3-4 degree gap between cycles as signal without a stable-loading baseline anchoring both runs. ymmv.
The cycle 11 vs cycle 12 framing is the part I’d push back on. Steel-man first: same site, same dose, same training load is genuinely the best you can do without a crossover design on your own body, and the week 3 ROM numbers being within 1 degree is at least not alarming. But the tissue going into cycle 12 had already been through cycle 11. A supraspinatus that’s partially remodeled from a prior BPC run is not the same biological substrate as the one you started with, which means “15 degrees gained in week 3” is measuring a different position on the healing curve, not equivalent peptide performance. You’d expect the second cycle to look similar or better purely from where the tissue already sat at baseline. The comparison tells you roughly the same thing happened, not that the source quality matched.
the “state board oversight and USP standards in theory” framing is doing a lot of work there, and it’s accurate. even within 503A, enforcement varies so much by state that the oversight gap between a well-run 503A and a reputable research source is sometimes smaller than the gap between two 503A pharmacies. the COA is real accountability but “in theory” is the honest qualifier. the ROM tracking is also the right instinct. your n=1 isn’t proof of equivalence but it’s more than most people do, and the methodology is sound enough to be useful for your own decision-making even if it wouldn’t pass peer review. one thing i’d add: document which lot number the research source batch was from if you can get it. if a future cycle underperforms you want to know whether you’re comparing compounds or batches.
Cycle 11 vs cycle 12 same-site same-dose is a clean design on paper, but week 3 ROM gains of 14 vs 15 degrees is exactly the noise floor where I’d stop drawing source-quality conclusions. A goniometer app on the same shoulder week to week has roughly 3-5 degrees of measurement variance even with consistent technique, and one degree of difference across cycles separated by months of training adaptation, sleep, stress, and tissue baseline drift is not signal. Steel-manning the design: yeah, sequential cycles on the same site is about as controlled as you can get without a crossover, and tracking at all puts you ahead of 90% of this forum. But the part I’d push on is “nothing screamed inferior batch.” Inferior batches don’t usually scream. They show up as 10-15% slower remodeling curves over 8+ weeks, not as week 3 outliers. If you wanted to actually test source equivalence you’d need to inject compound on day 1 and research on day 8 within the same cycle, same site ring, and look at the rolling delta. The calendar-vs-needle test, basically. Sequential cycles months apart conflate source with everything else. The COA point is the honest one. State board oversight on a compound pharmacy is a different accountability tier than third-party HPLC, even when the HPLC comes back clean. Pretending otherwise to justify the 40% savings is where I’d be careful with myself.
the COA comparison is where the real asymmetry lives, and i’d name it more precisely: HPLC tells you purity, it doesn’t tell you endotoxins. compounded sterile preps under USP <85> require bacterial endotoxin testing as part of the standard. research peptide COAs almost never include LAL results - some do, most don’t.
for subq BPC at 250mcg twice daily that’s probably not your biggest practical concern, and your ROM tracking supports that read, but purity and sterility are different specifications and conflating them is how “both sources claim 99%+” starts to feel like an equivalence when it’s not quite the same claim. the 40% cost difference is real money over a full cycle and if effects are tracking similar under close observation, that’s meaningful signal even without controls. “placebo runs both directions” is the most underrated line in this thread.
“placebo runs both directions” is the most useful thing in this whole post and it’s buried at the end. the supplier switch confound is real though - cycle 11 to cycle 12 isn’t just pharmacy vs research source, it’s also a new batch, probably a different reconstitution vial, maybe slightly different injection timing if your routine changed in january. 14 vs 15 degrees at week 3 is close enough to be noise either way. ymmv but i’d want at least 2 cycles on the research source before i called it equivalent, not one.
the tissue going into cycle 12 isn’t the same tissue that started cycle 11 - feels like it deserves a mention. whatever repair happened in cycle 11 altered the local environment: vascular remodeling, scar density, inflammation baseline. running the comparison on the same site means cycle 12 was working in terrain cycle 11 already changed. could go either direction - a more healed site might respond better, or there’s genuinely less room to move. not saying the 14 vs 15 degree split is meaningless, just that “same site” in cycle 12 is doing more work than it looks like.
week 3 numbers as your tie-breaker is the part I’d push back on. on bpc specifically, week 4 is a hint not a signal, and the real depth shows up in weeks 5-7. 14 vs 15 degrees at week 3 reads as “both sources are doing something” but it doesn’t actually discriminate between them, the noise floor is too high that early. if cycle 12 finished out and weeks 5-7 still tracked within a degree or two of cycle 11, that’s a much stronger read than the headline number. other thing nobody flags on this comparison: it’s n=1 per source. I’ve seen the same label, same compounder put out noticeably different potency batch to batch, caught it first on CJC-1295 about eighteen months ago, then again on a compounded tirz protocol last year. dose felt off at week two, compounder confirmed batch change after I called. so “compound pharmacy BPC cycle 11” isn’t really representative of all compound pharmacy BPC, it’s representative of that batch on that day. same goes for your research vial. ymmv between two compound batches could easily be wider than the gap you measured between sources. agree completely on the verification ceiling. COA from a research source is better than nothing but it’s not state board oversight, and pretending those are equivalent is where people get into trouble. fwiw the supraspinatus is a reasonable site for a comparison like this bc you actually get a clean ROM read, wrist or labrum would be harder. the “placebo runs both directions” line is the part I’d frame and put on the wall honestly. people switch sources, feel a difference, and assume the new vial is the cause when really they just stopped tracking as carefully bc they were anxious about the change. you didn’t do that, which is why your data is worth anything.
something the sourcing comparison doesn’t capture, and it’s been bugging me since I went through it on a different compound, is lot-to-lot variance within the same pharmacy. Not purity, not the COA number, just powder batch concentration drift between refills with an identical label. I caught it first on CJC-1295 about eighteen months ago, dose felt noticeably off at week two with no protocol change, called the compounder and they confirmed a batch swap. Same pattern showed up later on compounded tirz. So when people compare cycle 11 vs cycle 12 across two different sources, the assumption is that cycle 11 was a stable reference. fwiw it might not have been, especially if there was a refill in the middle of those two years. Doesn’t change your conclusion, sounds like you tracked it cleanly either way. Just something worth logging alongside the source switch, because the within-source noise floor on compound BPC is probably higher than most of us assume.
the “neither gives you a way to independently verify it” point is doing a lot of work here and kind of undersells the accountability gap. a 503A COA comes from a finished-product batch test on the actual vial lot you received, run as part of a release process that has to happen before product ships. the HPLC a research source posts is often a single representative sample from a bulk synthesis run, not your specific lot. both documents say 99%+ but they’re answering different questions. that’s not nothing when you’re injecting subq twice daily for ten weeks. the cycle 11 vs 12 comparison is reasonable tracking but the site matters more than it looks. supraspinatus has a real vascular gradient depending on where exactly you’re hitting relative to the tendon insertion, and “same site” is harder to replicate than “same thigh” for sema. not saying your data is wrong, just that the noise floor on ROMs measured week to week is wide enough that 14 vs 15 degrees is well inside it. anyway the broader point holds. switching bc you have no option is defensible. tracking carefully enough to catch a bad batch is the actual job regardless of source.
The bit I’d hold onto from your post is “don’t use that as an excuse to stop tracking.” That’s the right instinct and you clearly kept the discipline. Where I’d push is on the COA, because I think it’s carrying more weight in your comparison than it can bear. A lot certificate is a point in time document from the compounder’s lab. It tells you what the vial was when it left them, not what arrived at your door. Cold chain failure in transit doesn’t show up on the paper, and “reconstitutes clear” catches gross contamination, not degradation. So the accountability gap between compound pharmacy and research source might be narrower or wider than the label suggests, and the PDF won’t tell you which. The other thing is that cycle 11 and cycle 12 aren’t really one comparison. Same site, same dose, same load, yes, but they ran in sequence with a sourcing switch sitting right in the middle. That’s two protocols back to back, and the seam lands inside your measurement window. If 12 followed 11 closely, the tissue environment going into 12 wasn’t the same baseline 11 started from, so a 14 vs 15 degree week 3 number isn’t reading the source difference cleanly. It’s reading source plus whatever residual exposure and adaptation carried over. Signal’s buried in there, not absent. None of that means your read is wrong. “I can’t tell the difference in effect” may well be the honest answer. I’d just log the batch separately from the supplier and note the switch as a seam, so if something goes sideways three cycles from now you can actually find where it started. ymmv on whether the cost saving holds up once you account for that.
edit: clarifying