FDA pulled tirzepatide off the shortage list, which triggered the grace period wind-down for compounding pharmacies. Most of the discussion I’ve seen is about access and cost, which makes sense. But the piece nobody seems to be tracking carefully: what happens to the metabolic numbers when someone switches from compounded to branded, or has to pause entirely while sorting out insurance. The GLP-1 mechanism doesn’t care which vial it came from in principle, but dosing consistency does matter - and people on compounded often had non-standard concentrations or titration schedules. If you’re tracking fasting insulin or A1c and you switch formulations, you’d want a fresh baseline draw within 4-6 weeks, not assume your previous trend just continues. I’m on Zepbound through insurance so this doesn’t hit me directly, but I’ve been watching the threads. The people who are going to lose the most ground are the ones who weren’t tracking labs to begin with - they won’t know if the switch changed anything metabolically. If you’re in that situation, get labs before you switch, not after.
the 4-6 week window is right for A1c but fasting insulin is noisier than people give it credit for, so a single draw at the switch point can easily mislead you; ideally you want two pre-switch draws a couple weeks apart to know what your actual baseline variance looks like before attributing any post-switch shift to the formulation change.
Two draws is exactly what I landed on after my week 6 baseline surprised me: single-point fasting insulin of 11, then 14 two wks later, no formulation change, nothing different. If I’d switched something b/w those draws I’d have blamed the switch. The variance across draws on the same protocol is real enough that “baseline” should probably mean an average of two, not a snapshot. Tracking both in the same place where you log symptoms makes the scatter visible, otherwise the numbers just sit in separate lab PDFs with no context.
Two draws as the actual baseline is the right instinct, and I’d go further: for fasting insulin specifically, even two might undersell the noise. The intra-individual coefficient of variation on fasting insulin sits somewhere in the 20 to 30 percent range in most of the methods literature I’ve seen, which means an 11 followed by a 14 isn’t a signal at all, it’s just what the assay does on a Tuesday versus a Thursday. HOMA-IR inherits that same scatter and then some, because you’re multiplying two noisy numbers together. So the “average of two” rule is probably still optimistic if anyone is trying to detect a real shift from the formulation switch. Three draws over four to six weeks would be closer to honest, though I appreciate that’s a lot of phlebotomy for a question that might not even have a clean answer. Where I’d partly push back is the framing that the scatter only becomes visible when you log labs alongside symptoms. That helps with attribution, no question, but the deeper issue is that most people (and a fair number of GPs, in my experience) treat a single fasting insulin like it’s a thermometer reading. It isn’t. It’s a draw from a distribution, and the distribution is wider than the clinical reference range suggests, particularly in people who are mid-titration on a GLP-1 where appetite, meal timing, and sleep the night before all bleed into the result. A1c is the more forgiving variable here precisely because it integrates over weeks; fasting insulin is the one that gives people whiplash. If anyone’s switching formulations and only has the budget for one extra draw, I’d put it on A1c at the eight week mark and treat the fasting insulin as directional only. Not perfect, but it’s the variable that actually moves slowly enough to mean something.
the “fresh baseline draw within 4-6 weeks” point is solid, but i’d push back on framing the compounded-to-branded switch as the main risk here. the bigger confound for most people isn’t formulation, it’s the dose gap during the insurance fight. i’ve seen threads where people were off tirz for 6-10 weeks sorting out prior auth, and that’s where the metabolic numbers actually move. a 4-week rebaseline assumes you transitioned cleanly and stayed on dose, which isn’t what’s happening for most people losing compounded access rn. if anything, the lab timing recommendation should probably be “draw at the start of the gap, not after you’ve restarted” so you capture the actual interruption effect, not just the formulation difference.