Tirz has roughly a 5-day half-life. Weekly dosing means you’re hitting your plasma trough on days 5-7, right before the next injection. Appetite suppression follows the same curve. So the question I’d ask anyone in a prolonged plateau: which days of the week are you actually eating more? Most people log calories but don’t stratify by injection day. If your deficit is solid on days 1-4 and then soft on days 5-7 when suppression fades, the weekly average looks fine but you’re quietly eating back 60-70% of it. That math produces a plateau, not a response failure. What I’d test before assuming you need a dose increase:
- Log day-of-week vs intake for 2-3 weeks, injection day = Day 1
- If days 5-7 are consistently 300-500 cal higher, you have a timing problem
- Some people shift to a 5-day injection cadence to smooth the trough; not standard, not risk-free, but there’s logic to it The dose escalation instinct is real but it’s worth ruling out trough-driven eating first. Metabolic adaptation is also in play by week 5, but that’s a separate layer.
Steelmanning this: the pharmacokinetic logic is sound, and you’re right that weekly averaging hides the within-week pattern. I’d add one caveat from the chemist side, though, before anyone treats day 5-7 intake as proof of trough-driven hunger. The plasma curve and the appetite curve are correlated but not identical. Tirz’s effect on gastric emptying and central satiety doesn’t track plasma concentration in a clean linear way, there’s receptor-level stuff happening that doesn’t read off the half-life chart. So a day 6 intake bump could be trough pharmacology, or it could be weekend social eating, or sleep debt accumulating by Friday, or any of the other things that move appetite by 300-500 cal. The day-of-week confound is non-trivial if most people inject on a Sunday or Monday. The way I’d tighten the test: have people inject on a Wednesday for two weeks and a Saturday for two weeks, same dose. If the day 5-7 bump tracks injection timing rather than calendar day, that’s actually informative. Otherwise you’re attributing a weekend effect to pharmacology. Not arguing against the framework, just saying n=1 self-data on this is going to be noisy, and the obvious confound is the one to design out first.
The crossover design is the right call and I should have included it in the OP. Inject Wednesday for two weeks, inject Saturday for two weeks, same dose, see if the bump tracks the needle or the calendar. That’s cleaner than anything I outlined. The plasma/appetite dissociation point is real and I don’t want to wave it off. Receptor occupancy, downstream GLP-1 effects on gastric motility, central satiety signaling that doesn’t follow a clean concentration-response curve at steady state. Fair. But I’d argue it doesn’t fully invalidate the framework, it just means a day 5-7 intake bump is a necessary but not sufficient condition. If you shift injection day and the bump moves with it, you’ve got something worth acting on. If it doesn’t, you’ve learned that it’s behavioral or sleep-related, which is also useful information even if it’s not what you wanted to hear. The weekend confound is the one I’d weight most heavily. Monday injectors are basically running an uncontrolled experiment where pharmacological trough perfectly overlaps with Friday-Sunday social eating. That’s almost impossible to untangle without the design you’re describing. Worth building into any self-tracking protocol from the start rather than trying to adjust for it afterward.
The crossover design is the cleanest move you’ve described in the thread, and I’d run it before touching the dose. One caveat on “if it doesn’t, you’ve learned it’s behavioral or sleep-related”: I’d add a third bucket, which is metabolic adaptation that just happens to coincide with whatever week you’re running the test. You flagged adaptation as a separate layer in the OP, but in a 4-week crossover it’s not actually separate, it’s running underneath the whole experiment and biasing the second arm downward regardless of injection day. Doesn’t kill the design, but it means a null result on bump-tracking is harder to interpret than a positive one. The asymmetry matters. The Monday-injector point is the one I keep coming back to. That’s not a confound you adjust for, that’s a structural problem with most self-reports in this space. Anyone analyzing their own data without controlling for injection-day-vs-weekend overlap is essentially producing noise and calling it signal. Worth saying out loud more often than people do.