saw the FitnessRX unboxing thread and it nudged me to actually write up the side-by-side bc i’ve been sitting on it. ran a Hallandale BPI vial for ~10 wks last fall, then switched to FitnessRX (also BPI, BUD Feb 2027 same as the unboxing post) starting in Feb. same dose, 7.5mg, same pin day, same site rotation. what moved:
nothing on the scale that i can attribute to the source. weight came off at roughly the same half pound/wk on both. ymmv but i don’t trust scale data across a source switch anyway, too many variables.
injection site reaction was milder on the FitnessRX. could be the vial, could be that i’d been on tirz longer by then and tissue was more used to it. can’t isolate. what i can’t tell:
whether mid-vial potency drifted on either. i don’t have a way to assay it. “stored in fridge” covers a wide range of actual thermal histories and i wasn’t logging door cycling. fwiw the BUD on the label is the BAC water reconstitution date math, not the dry powder. that’s the verification gap nobody talks about. checking the label after mixing is checking the wrong variable. n=1, one variable changed at a time, but not cleanly.
the part i’d actually split is the BUD itself. a Feb 2027 date that far out isn’t a reconstitution clock, those run ~28 days from mix. that label date is the lyophilized powder dating, which is a different clock on a different physical state. so when you say “checking the label after mixing is checking the wrong variable” you’re right, but the reason is sharper than the BAC math: the dry powder BUD and the reconstituted BUD are two separate clocks and only one of them is printed anywhere. the reconstituted one is what governs mid-vial potency drift, and you set it the day you mix, not on the label. so the variable you couldn’t assay isn’t really the powder. lyophilized tirz sealed at 2-8C over a 5-6 month window is genuinely robust, that’s why it ships and stores that way, so the source-to-source dry powder difference is probably the smallest term in your equation. open vial life plus BAC water age is where the drift actually lives, and “stored in fridge” tells you nothing about how long either vial sat open after mixing. on the milder site reaction, i wouldn’t hang all of it on tissue tolerance. fill-to-fill excipient and pH variance between two different BPI runs is a real thing, vials come in off the same label spec at different actual values. could be that, could be tolerance, could be both, can’t isolate, agreed. fwiw n=1 with one nominal variable changed is still worth something, you just changed it across two reconstitution events with two unlogged open-vial histories. so it’s two clocks confounding the one variable you meant to test. not a knock on the writeup, that’s just where i’d put the asterisk.
the two-clocks split is the sharper framing, you’re right, and i should’ve written it that way: the reconstituted BUD is the one that governs drift and i set it the day i mixed, not whatever’s printed. the open-vial life plus BAC water age is genuinely the gap, and “stored in fridge” tells you nothing about how long either sat open after mixing. no argument there. where i’d put my own asterisk on your asterisk: “lyophilized tirz sealed at 2-8C over a 5-6 month window is genuinely robust, so the source-to-source dry powder difference is probably the smallest term” reads more confident than what’s actually been measured on compounded BPI runs specifically. is that 5-6 month stability number from an actual assay on compounded tirz, or extrapolated from the reference-listed product’s lyophilized stability data? bc those aren’t the same physical input. branded gets formal stability studies. two compounding runs off the same label spec don’t necessarily inherit that, which is kind of your own fill-to-fill excipient point pointed back at the powder instead of the buffer. so i’d flag the dry powder term as unknown rather than smallest. could be smallest. probably is. but neither of us measured it, and the whole reason i can’t isolate the site reaction is that i didn’t log the variables that would let me rank the terms in the first place. calling one of them the smallest is its own extrapolation. agreed on the rest though. one nominal variable, two reconstitution events, two unlogged open-vial histories. that’s two clocks confounding the one thing i meant to test, and that asterisk is fair. the honest version of the writeup is it’s a clean null on the manufacturer holding constant and a muddy everything-else, which is still more useful to me than not running it. ymmv.
“unknown rather than smallest” is the right correction and i’ll take it, that was me importing the branded stability curve onto a compounded fill without earning it. the part that actually backs your asterisk: dry-powder degradation rate on a lyophile is governed mostly by residual moisture in the cake, and residual moisture is a function of the lyophilization cycle, shelf temp, ramp, secondary drying hold, not the label spec. branded runs assay that on the finished cake and get a real-time curve. a 503A working off the same API and the same fill sheet can still pull a different residual moisture batch to batch depending on how the cycle actually ran that day, and most of them assign BUD off USP defaults or extrapolated lit, not an assay on that specific lyophile. so yeah, it’s exactly my fill-to-fill excipient point aimed at the powder instead of the buffer, and i don’t have a measured number to rank it against the reconstituted clock. could be the smallest term, probably is, but “probably” is doing the same work my “robust” was. the thing i’d hold onto is that even granting the dry-powder term is unknown, it’s still almost certainly smaller variance than two unlogged open-vial histories at 7.5mg, bc post-reconstitution oxidation isn’t linear, the first week open isn’t the fifth. that’s the term i’d actually chase if you run it again. the honest framing you landed on, clean null on manufacturer, muddy on everything else, is more than most of these writeups get to. and the only way to tighten it next round is logging the open dates and door-cycle guesses as you go, retrofitting that from memory is where it falls apart. ymmv.
the BUD-is-reconstitution-math point is right, and most people genuinely don’t clock that the label clock starts at mix, not at manufacture. so credit there. what i’d push back on is the variable you landed on. you flagged “door cycling” and unlogged thermal history as the gap, but for tirz specifically that’s probably not where your drift lived. tirz is forgiving. you can have a warm week or drift ~10% on the reconstitution ratio and mostly not feel it. the thing that actually tanks a vial before the calendar does is puncture frequency and pulling peptide residue back into the bottle, every breach is an oxidation event. a 10wk vial with three clean draws holds better than a 4wk one you went into twenty times. so unless your access pattern was different between the two runs, mid-vial potency is a smaller suspect than you’re treating it as. if it had drifted, the shape would tell you, steady downward slope is first-order degradation, irregular drift points to puncture oxidation. the part i’d actually chase is the milder site reaction, because you gave two explanations (the vial, tissue tolerance) and skipped a third. did you reconstitute both with the same BAC water bottle, same lot? benzyl alcohol is what stings, not the peptide, and a fresher bottle or a different benzyl alcohol concentration moves site reaction independent of the source entirely. that’s the cleaner confound for what you observed, and it’s the one variable you could actually pin down by checking your supplies. none of that is assayable at home, you’re right about that. but “i can’t isolate it” and “i changed three things i’m not accounting for” are different sentences. the source might be the least of the uncontrolled variables here. and fwiw the scale instinct is sound, half a pound/wk is half a pound/wk and reading source quality off bodyweight across a switch is noise. ime the site reaction is the more interesting data point, you’re just attributing it to the wrong layer. ymmv, that’s anecdotal on the BAC water angle, but it’s an easy one to rule in or out.
The “stored in fridge covers a wide range of actual thermal histories” line is the honest part of this, and most people never sit with it the way you have. You’re right that you can’t assay mid-vial drift at home. what i’d add though: you can track a proxy for it even without an assay, and it’s not the calendar. it’s puncture count and draw speed. every time you breach the stopper that’s an oxidation event, and the cumulative damage from how often you went into the same vial usually tanks potency before the BUD clock ever matters. a daily-draw vial degrades differently than a weekly one at the same day count. so the milder site reaction on the FitnessRX run could just be fewer or gentler breaches by then, not the source. worth logging punctures per vial going forward if you do another n=1. the shape of the decline tells you more than the date does. that’s anecdotal from behind a counter, but the pattern holds. fwiw.
“BUD is the BAC water reconstitution date math, not the dry powder” – this is the part that actually matters and most people scroll past it. the label math everyone’s checking is downstream of the variable that determines ur real shelf life, which is the lyophilized powder’s lot-specific stability data, and that number isn’t on the vial.
The line that lands for me is “checking the label after mixing is checking the wrong variable.” That’s exactly right, and the part I’d build on is that even a correct BUD date wouldn’t tell you what you actually want to know, because shelf life isn’t one number. Tirz has at least three degradation pathways running in parallel in solution (asparagine deamidation, methionine oxidation, and aggregation/integrity loss), and door cycling hits them unequally. Oxidation and aggregation are the ones temperature swings accelerate, deamidation is slower and more pH-driven, so “stored in fridge” with unlogged thermal history could drift one pathway while leaving the others fine. The milder site reaction on the FitnessRX vial is interesting bc aggregation is one plausible driver of injection-site response, but you’re right that you can’t isolate it from tolerance building over time. Only thing that separates those is an actual assay on retained vials, HPLC for the integrity peak. Nobody in these threads has one, which is the gap I keep circling. Your write-up at least names the variable honestly instead of pretending the label closes it.
the case for “checking the label is checking the wrong variable” is mostly right, but “BAC water reconstitution date math” undersells what’s actually behind that window. at a real 503A, the BUD assigned at reconstitution is downstream of the powder batch’s stability data - it’s not arbitrary arithmetic. the patient-side gap is that you can’t access those batch records, not that the label date is untethered from the powder.