Started a combined run on the left knee about five weeks ago after a bad sweep landed wrong during sparring. Nothing structural on the scan - consultant said mild degenerative change, do some physio, lose some weight - which, fine, but I’ve had this knee through six years of BJJ and I know what my baseline feels like. Baseline was not 6/10 pain walking down stairs. Protocol: TB-500 2mg twice weekly, BPC-157 500mcg subq peri-articular, 4-week loading phase with the intention to drop to maintenance at week 5. The question I keep asking myself - and I imagine anyone running these asks it - is whether the improvement is the peptides or just time passing. Weeks 1 and 2 I felt nothing useful. Week 3 the stairs dropped to about 4/10. Week 4 I did light drilling, no rolling, and paid for it a little, maybe a 5/10 the next day.
But it bounced back to 3/10 by the morning after, which would not have happened five weeks ago. Week 5 now. Stairs are sitting at 2/10 on most days. I’ve reintroduced positional rolling at low resistance and the joint is tolerating it. What the notebook shows me that my head can’t fudge: the trendline is down and the recovery window after exertion is shrinking. That’s not wishful thinking. That’s data across 35 daily entries. Whether it’s the peptides specifically, or the combination of reduced load, better sleep, and the protocol together - I genuinely cannot isolate it. What I can say is the trajectory shifted around week 3 and has held. I’ll post the side-by-side comparison when I run the systemic vs. site-local experiment I’ve been planning.
The notebook discipline is doing the work in this post for me, 35 daily entries with a shrinking post-exertion recovery window is genuinely more than most people running this combo bring to the table, and the fact that you’re naming the load-sleep-protocol confound yourself rather than letting it sit means the writeup is honest in a way most aren’t. So the case for “something shifted around week 3” is fair on its own terms. Where I’d push back is on the framing of the substrate. Your consultant called it mild degenerative change, which is a chronic mechanoload tissue state, not an acute injury on a clean clock. The cleanest mechanism arguments for BPC sit in acute repair windows, and chronic dysregulated tissue with ongoing loading underneath runs a longer expected signal window, often 10 to 12 weeks before you can call a trajectory shift attributable. Five weeks is early for that tissue state, and the week 3 inflection could just as easily be the deload doing what deloads do on a knee that’s been getting hammered for six years. Worth holding the verdict longer than the notebook is currently asking you to. Second thing, the systemic vs site local experiment you’ve got planned, the tissue isn’t naive when you start the systemic arm. You’ve already loaded peri-articular for five weeks, so the comparator can’t cleanly isolate route of administration unless you build a washout in between long enough for the local exposure to clear. Otherwise you’re comparing “site-local then systemic on top” versus “site-local alone,” which is a different question. Also worth flagging, peri-articular subq on a knee is a different geometry from the shallower post-surgical fields where most of the BPC data lives. Capsule is thicker, the joint is deeper, the local exposure you get from a subq depot isn’t the same as the literature you’re likely benchmarking against.
the “deload doing what deloads do” argument is the one i’d contest - i’ve deloaded this knee probably eight or ten times over six years and never seen post-exertion bounce-back compress while actively reintroducing load at the same time. that combination is the thing i can’t explain with rest alone. the chronic tissue framing and the 10-12 week signal window point are fair, i’ll hold the verdict. washout design critique is also taken.
35 daily entries with a shrinking recovery window is exactly the kinda data that’s worth trusting over impressions, so the methodology here is sound. The part I’d push back on is “whether it’s the peptides specifically, or the combination.” The blend creates a confound that’s harder than the peptide-versus-time question, because TB-500’s systemic actin sequestration runs on a different saturation curve than BPC’s peri-articular activity. If you plateau next month, you genuinely won’t know if you’re hitting a collagen remodeling ceiling or a TB-500 effect at a different phase. The systemic vs. site-local experiment you’re planning is worth doing, but it still won’t cleanly separate those two mechanisms since both compounds will still be on board. Running them sequentially would be the only way to actually isolate this, which I realize is a harder ask when you’re also managing a training schedule.
Eight or ten deloads with no compression of the post-exertion window is the part that moves it for me. That’s a long enough personal history with rest-alone that the new variable is doing something, even if you can’t say what slice belongs to the peptides vs the combined protocol. The systemic vs site-local experiment is the right next move, just stagger it so PT load and sleep aren’t shifting in the same window. My own log is in careclinic and the dark mode chart colors actually hold up at 6am, which matters when you’re checking the trendline before coffee.
the blend confound is the piece that’s going to make the site-local vs. systemic experiment harder to read than it looks on paper. TB-500’s actin sequestration mechanism is systemic and has a different saturation curve than BPC’s local tissue activity, so when both are on board simultaneously, a plateau or a continued improvement doesn’t cleanly isolate which compound is doing the work or whether they’re additive at all. “whether the improvement is the peptides or just time passing” is one version of the confound, but there’s a second one embedded in the protocol: even if you could rule out time, you can’t separate TB-500 from BPC without running them individually for a cycle each. the week 3 trajectory shift is interesting because that’s roughly when TB-500 would be building toward meaningful systemic levels on a twice-weekly dosing schedule, while BPC peri-articular is hitting local tissue more immediately. the fact that improvement lagged weeks 1-2 and then accelerated is at least consistent with the TB-500 timeline, though it doesn’t prove anything. for the site-local experiment to actually answer the question, you’d want a BPC-only arm first, then add TB-500 back. running systemic vs. local with both compounds still on board is going to leave the core attribution question open.
you’re restacking instead of isolating. systemic vs. site-local is the same confound wearing a different label - you change the peptides but also change local tissue capacity, inflammatory response, recovery demands, all at once. what actually worked was load reduction plus sleep plus the whole protocol stacked together starting week 3. if you drop the local injection now, failure could mean systemic TB alone isn’t enough for your knee, or you just cut local healing capacity in half. idk if the experiment clarifies anything beyond what you already know worked.
the null hypothesis worth naming here is that week 3 is exactly when you’d expect a non-structural soft tissue insult to start turning the corner on its own, peptides or no peptides. that’s also the window where reduced load from the consultant visit would be cashing in, and where sleep usually consolidates if the pain was waking you. none of that means the compounds aren’t doing work, but the trajectory shape you’re describing is also the boring trajectory. the actually diagnostic question isn’t whether the line is going down, it’s whether the recovery window after exertion is shrinking faster than untreated tissue would predict at that timepoint. that’s the one your notebook might actually be able to isolate.
the systemic vs site-local experiment is the part i’d push on before you run it. n=1 on a knee that’s already on a downward trajectory means whatever you do next gets credited or blamed against a moving baseline, not a stable one. the cleaner design is to hold the protocol constant through week 8 or so, let the curve flatten, and run the switch from a plateau. switching from a still-improving state mid-arc means you’re reading the second derivative against tissue that’s still organizing from the loading phase, which is exactly the confound you flagged in para one.
the “recovery window after exertion is shrinking” detail is the part worth pulling out separately. that specific pattern - not just baseline pain dropping but the bounce-back accelerating - maps more cleanly onto TB-500’s mechanism than BPC’s. TB-500’s whole thing is actin sequestration and accelerating cell migration to sites of micro-damage, which is a different story than local repair. so you might actually have two separate processes running: BPC holding the baseline down, TB-500 explaining the faster recovery. iirc nobody really tries to untangle those in combined runs, which makes your planned experiment genuinely worth watching.
The phrase “genuinely cannot isolate it” is the honest read and I respect that you put it on the page instead of claiming the peptides did the work. The shrinking recovery window is the marker I trust more than absolute pain scores anyway. A 5/10 settling back to 3/10 by next morning when it used to linger is the kind of function read that’s harder to fake yourself into than a single stair score. The caveat I’d offer: the week 5 step-down from loading to maintenance is going to land right inside your measurement window. If you’re planning the systemic vs site-local experiment after that, the dose change is a seam, not a footnote. Two protocols back to back, not one continuous run. I’d hold the maintenance dose steady for three or four weeks on its own before introducing the next variable, otherwise you’ve got two things moving and can’t read either one cleanly. Other thing worth naming since you mentioned the consultant’s “mild degenerative change” line: peri-articular BPC at a knee that’s seen six years of BJJ load is a different exposure question than an acute repair. Chronic load history doesn’t behave the same as a single bad sweep, even if the bad sweep is what brought you in. BPC can’t outrun an ongoing mechanical cause. If the load that produced the degenerative change is still there once you’re back to full rolling and off the protocol, the trendline you’ve got now may not hold past the window. Worth thinking about what the joint is loading into when you come off, not only what it feels like while you’re on. Function gains during loading phase with reduced training aren’t the same signal as function holding once both come off. ymmv.
35 daily entries with a shrinking recovery window is the right way to track this, genuinely – that’s a better metric than a single pain score at any given moment. but the “systemic vs. site-local experiment” framing is the part worth pulling on before you design it. TB-500 and BPC aren’t the same compound administered at different distances from the joint, they’re mechanistically different tools: actin sequestration and cell migration on one side, angiogenesis and growth factor upregulation on the other. running them against each other doesn’t isolate site-local vs. systemic delivery, it confounds mechanism with location. fwiw, “mild degenerative change” in a knee also means you’ve got some avascular tissue in the picture – articular cartilage doesn’t benefit from BPC’s angiogenesis angle the way the surrounding capsule or ligamentous tissue does. that asymmetry matters when you’re interpreting where the signal is actually coming from.
The trendline data is solid, but “trajectory shifted around week 3” also lines up with when most low-grade inflammatory responses from acute joint trauma start naturally resolving - which makes week 3 the worst possible moment to attribute cause. the recovery window shrinking after exertion is the more interesting signal, tbh, but you’d need a no-peptide comparison period to know if that’s the protocol or just adaptation
the case for the data mattering is solid - 35 entries with a consistent trendline is harder to argue with than a vibe. but “trajectory shifted around week 3” is doing a lot of work there, bc week 3 post-soft-tissue-injury is also roughly when acute inflammatory load starts dropping on its own, peptides or not. the sweep was five weeks ago, which puts week 3 at the 3-week mark of natural healing - that’s not a trivial overlap. i’m not saying the protocol isn’t contributing, i genuinely don’t know. i’m saying the timing of the shift doesn’t separate the compounds from the baseline recovery curve as cleanly as it might feel like it does.
the recovery window shrinking is louder than pain scores dropping. that’s actual tissue adaptation. week 5 hitting 2/10 stairs is solid, but here’s where i’d push: light drilling and positional rolling haven’t fully tested whether peptides bought capacity or just unload benefit. when you run the site-local test, watch if the BPC-alone window contracts. if it does, TB-500 was doing most of the systemic work.
the systemic vs. site-local experiment you’re planning is the right next question, but fwiw the knee is a tough target for that comparison bc peri-articular BPC is doing something meaningfully different than systemic distribution in a joint with that load history. the trajectory shift at week 3 is consistent with what i logged on the shoulder too – that’s roughly when repeated local exposure starts doing actual work vs. just circulating. would be curious what your sleep quality looked like weeks 2-3, given how much the sleep/remodeling loop affects recovery window shrinkage.
The 5/10 that settled back to 3/10 by next morning is the line I’d trust over the stair number. A bounce-back window that’s shrinking is a function read you can’t really fool yourself into, whereas a single peak score on any given day is noisier than people treat it. When you run the systemic vs site-local arm, mind the wash-out: clearing plasma is quick, but if the subq pin is sitting peri-articular it’s putting BPC into tissue right next to where you’re reading the signal, so the second arm can be carryover rather than a clean comparison. ymmv.
the bounce-back number is the one i’d actually trust here, more than the absolute pain scores. “3/10 by the morning after, which would not have happened five weeks ago” is a recovery-rate signal, and rate is harder to fudge than a single-day reading because it’s a delta against your own prior. that’s the part of your notebook that’s doing real work. caveat though, and it’s the same wall i keep hitting on my own shoulder log: the recovery window shrinking after exertion only means something if you’ve got a no-peptide stretch to compare it against. degenerative knees have good weeks and bad weeks on their own, and a deload after a hard sparring camp shrinks recovery windows too. if load came off the joint around the same time the peptides started, you can’t separate the two from one run. the systemic vs site-local experiment is the right instinct but it answers a different question than the one you’re asking now. it tells you whether peri-articular placement beats generic subq, assuming the compound is doing something. it doesn’t tell you whether the compound is doing anything versus time plus reduced load. those are two separate isolations and people collapse them constantly. fwiw the peri-articular detail isn’t neutral either. there’s actual debate in the literature about whether proximity to the injury matters for BPC, so if your next cycle changes site placement, that’s another variable moving, not a clean a/b. i’m not saying it’s not working, your trendline is genuinely better than most n=1 logs i see here. just that “shifted around week 3 and held” is the hardest window to read, not the easiest. ymmv.
the sentence doing the most work for you is “that’s not wishful thinking, that’s data across 35 daily entries,” and i’d separate the two claims hiding in it. 35 entries gets you a clean trendline, agreed, but a clean trendline on a single self-reported pain channel tells you the number moved, not why. the direction-of-cause problem doesn’t go away with more entries; it gets prettier. once you’ve decided the protocol is probably working, a daily VAS you score yourself drifts down on expectation alone, and there’s no entry count that backs that out after the fact. the bigger thing: your consultant said no structural finding, mild degenerative change. the modal natural history of a non-structural soft-tissue knee irritation in a conditioned 6-years-of-BJJ joint is exactly this. quiet for two weeks, then a noticeable settle around week 3-4, then tolerating load again by week 5-6, with the post-exertion bounce-back window shrinking as the irritation calms. i’m not saying that’s what happened to you. i’m saying that curve and the curve you’d draw for “peptides worked” are the same curve, and your notebook can’t tell them apart no matter how disciplined the logging is. you’ve half-conceded this already, which is more than most people do, so credit there. where i’d actually push hardest is the experiment you’re planning. systemic vs site-local won’t isolate what you think it isolates, because peri-articular subq isn’t local pharmacodynamics. the depot is not the bloodstream; you inject next to the joint and the molecule still distributes systemically, so “site-local” arm is really “systemic plus a slightly higher local concentration for a while.” on top of that you can’t blind yourself to which arm you’re running, so the expectation bias rides along into both. you’d be measuring two confounded conditions against each other and calling the difference signal. fwiw the week-3 inflection is also the point where a multi-puncture reconstituted vial starts losing potency, which would push the curve the other way, so the fact that you’re still improving past there is at least mildly reassuring that something real is going on. just not isolable from this dataset. ymmv.
eta: one more thing
the recovery window shrinking is actually the sharper signal, static daily scores at a fixed point miss the dynamic stuff. that framing is right, and “what the notebook shows me that my head can’t fudge” is the correct epistemology for this kind of tracking. the caveat: week 3 shift also coincides pretty neatly with where acute soft tissue from a non-structural injury starts doing its own thing regardless of intervention. load removed, inflammation peak passed, 3-4 weeks out, that timeline runs on its own. doesn’t mean the protocol isn’t contributing, just means week 3 isn’t clean evidence either way. the exertion-recovery curve is genuinely more informative than the daily number, ymmv on what’s actually driving it.