Acetate vs citrate tirzepatide: what salt form actually changes (and what it doesn't)

Health & Wellness Cardiovascular
1

the compounding threads keep treating salt form as a footnote and i don’t think it is. wanted to lay out what i think is real vs what’s vibes, because the conversation usually skips the depot. the basics:

  • tirzepatide free acid is the active peptide. acetate and citrate are counterions paired with it for stability and solubility.
  • once it hits subq tissue and dissociates, you have the same peptide either way. systemic PK shouldn’t differ in any meaningful way for the molecule itself.
  • where salt actually matters is local: depot pH, dissolution kinetics, fibrillation tendency over the 4-week vial life. what plausibly differs:
  • depot pH at the injection site. acetate tends to land mildly acidic in local tissue immediately post-injection. citrate is also acidic but with different buffering capacity, and that affects how fast the depot equilibrates.
  • injection site reactions. some users report less stinging on one vs the other, and i think this is real local PD, not systemic. the depot is not the bloodstream.
  • vial stability over weeks. peptide fibrillation kinetics are pH and counterion sensitive. there’s older glargine literature (which i’ve seen cited more than i’ve read recently) where buffer choice meaningfully changed aggregation rate, and i’d be surprised if tirz behaves entirely differently. what doesn’t differ:
  • once absorbed: half-life, Cmax, AUC. mass deposited matters, salt doesn’t.
  • efficacy on appetite/glucose. if you’re getting consistent dose and the peptide isn’t degraded, salt form is upstream of the receptor binding event. the part most people get wrong is attributing systemic differences (energy, nausea trajectory week 3 vs week 4) to salt form when they probably changed dilution or vial age in the same window. a degraded vial doesn’t announce itself, and “this batch felt different” is usually batch-to-batch potency variance, not citrate vs acetate. if you’re tracking this for yourself: log dilution, mass deposited, vial week, and salt form as separate variables. otherwise you’re correlating four things and blaming whichever one you happened to notice. fwiw i’d want to see proper comparative stability data on both forms at typical compounding concentrations before updating much. haven’t seen it cleanly published. happy to be pointed at it if someone has.
2

The depot-level framing holds up, but “systemic PK shouldn’t differ in any meaningful way” is doing a lot of work when the vial-life piece isn’t settled. My concern isn’t absorption after dissociation, it’s whether aggregation rate at the depot matters before dissociation is complete. If fibrillation kinetics differ between salt forms at typical compounding concentrations, the effective delivered dose may vary across vial week 3 vs week 4 even if injection volume is identical. That’s still upstream of the receptor but it’s not invisible. The confounding variable point is well-taken though. imo My CGM caught what turned out to be prep inconsistency showing up as an 8 mg/dL drift over two weeks, and I initially attributed it to the wrong variable entirely. You need vial date, dilution, and batch logged as separate columns before any salt-form comparison means anything. The comparative stability data gap is the actual limiting factor here.

3

the aggregation-before-dissociation point is the one that actually keeps me up, because you’re right that it’s not invisible and it’s also the hardest thing to back out of self-tracking data. an 8 mg/dL CGM drift across two weeks is exactly the kind of signal that gets attributed to dose response or food noise returning, when it’s really just week-3 vial doing slightly less than week-1 vial did. the older insulin aggregation work suggests the effect isn’t linear in concentration either, which makes the missing comparative stability data even more annoying.

4

“week-3 vial doing slightly less than week-1 vial did” is exactly the signal I traced back to prep inconsistency in my own data, and the non-linearity is what made it so hard to catch: the drift wasn’t proportional to vial age, which is what you’d naively expect if degradation were simple first-order decay.