Protocol Hub · follow-ups · rev 7.25 · 2026-07-23

What's waiting

Everything outstanding, in the order it's worth doing. Drafts already written.

While your folks are here
Take the regimen exactly as written.

Nothing on this page changes a dose, and nothing on it is urgent. The schedule is settled and every total has been re-verified. Come back to this list when you have room.

✅ The one thing that actually changed

ALCAR moved from 4:00 PM to 1:45 PM. It is now 6 AM (1 g) · 10 AM · 1:45 PM · 5 PM. Daily total unchanged at 2,500 mg. Everything else in the schedule is exactly where it was.

A · Under five minutes each

Do these first. Each one unblocks something else.

A1 · Bear
Confirm the ALCAR edit actually saved 1 min

Open the Master Regimen note. 1:45 PM should list ALCAR 500 mg. 4:00 PM should not. The PDF you exported on 7/23 did not have the edit — either it saved after the export, or it didn't save. If it's missing, paste in the rev 7.25 file instead of re-editing by hand.

A2 · Claude settings → Profile
Paste two corrections into your standing-context doc 3 min

Highest value on this whole page. That doc is read first, every session — so a stale line there re-imports itself into every conversation forever. Two edits, one paste each.

Replace the dual-AI bullet under 🔬 METHODOLOGY
• Dual-AI discipline — two operations, do not confuse them:
  1. CORROBORATION — to test whether a finding replicates,
     both models get the same original brief, independently,
     neither seeing the other. Only then does agreement mean
     anything.
  2. AUDIT / SYNTHESIS — handing one model's output to the
     other to stress-test its citations, find its errors, and
     integrate it against my phenotype is a DIFFERENT operation
     and is my primary workflow. It works. Do not discourage it.
  The test: am I asking the second model to AGREE, or to WORK?
  Agreement is worthless. Work is the point.
  ⟨the absolute "never cross-feed" form was Claude-origin,
   overbroad, and is retired 7-23⟩
Replace the soy line under 🚫 HARD CONSTRAINTS
⚠️ Soy — protease-fraction concern, ACTIVITY-gated, not
categorical. Kunitz/Bowman-Birk inhibit trypsin and
chymotrypsin ONLY — not lipase, not amylase. Creon is dosed
in USP lipase units, so the fraction the therapy is judged on
is mechanically untouched. EXCLUDE raw/unheated/high-TIA soy
(raw soy flour, raw soybeans, unheated soy protein powder).
CLEARED: soy lecithin, soybean oil, fermented soy, hydrolysed
soy protein/peptone, declared fermentation-medium traces.
Yardstick: NaturDAO ~96 mg/day legume protein = cleared.
Measure candidates against that, not against label presence.
⟨provenance unknown — demoted from ❌ at rev 7.21⟩
A3 · Physical check
Read the Erinamax bottle in your hand 1 min

Two questions, then write down the lot number:
  ① Does it say "as ErinaPrime®"?
  ② Does it say "Contains: Milk, Soy"?
This settles the milk question by itself. ErinaPrime launched ~June 2026 — a pre-June bottle is different material, and any tolerance you built on it does not transfer. It also decides how much weight the ND email carries.

B · Two decisions — yours, no deadline

Both are genuinely open. Neither has a wrong answer that hurts you.

B1 · Decision
Where does Erinamax live?

First, the thing that reframes it: the ~5 µg milk trace is inside the capsule. It is identical at 6 AM and at noon. Moving the slot does not reduce milk exposure at all — "no issues at 6 AM" records that the old bottle was tolerated, not that fasting protects. A3 settles that, not the clock.

  • Branch A · keep Meal 1 Fat + Creon already present. In EPI, fat solubilisation is enzyme- and bile-dependent, so this is the strongest absorption slot. Cost: adds to the Meal 1 attribution pile in B2.
  • Branch B · back to 6 AM fasted Buys a clean Meal 1. Cost: weakest absorption slot in the day — and note it never actually had the eggs, since 6:00 AM sits 30 min ahead of both the protein bite and the Creon. Weaker still once the bite goes.
  • Branch C · Meal 2 or dinner ← costs neither Fat and Creon already present, and outside Meal 1. Gets A's vehicle and B's clean window. Cost: it would be Erinamax's second move in three days — so if it goes here, it goes here and stops.

⚠️ What outranks all three: the dose is unknown within 2× (label says 0.5%, Nammex publishes 1.0%), and 2.5 mg/day has no human anchor at all — the trials ran 5.25 and 10 mg/day. Optimising absorption of a possibly sub-anchor dose is second-order. If one thing gets done here, make it C1.

B2 · Decision
How do you read the 7/21 → ~8/4 window?

It is not single-variable, and the file said it was. HistaminX started 7/21 · Erinamax moved into Meal 1 on 7/22 — same meal, next day · agmatine changed slot 7/22 · ALCAR changed slot 7/23. Nothing needs undoing. This is bookkeeping.

  • Option 1 · name them Accept the window as confounded and read any signal before ~8/4 as unattributed. Costs nothing, keeps HistaminX running.
  • Option 2 · restart clean Treat ~8/4 as the start of the real run, no further schedule changes until then.

⚠️ The specific collision to watch: Erinamax carries a declared milk trace, and the milk endpoint is the phlegm/cough response, which reads in days — the same timescale and system a probiotic readout uses.

C · Three messages — written, ready to send

Copy, paste, send. No composing required.

C1 · Email
Nootropics Depot — erinacine spec + allergen data

The highest-value message of the three. Question 4 is the one that matters: it decides whether you're taking 2.5 or 5 mg/day.

support@nootropicsdepot.com
Subject: Erinamax — erinacine A assay and batch allergen data

Hello,

I take Erinamax daily and I'm trying to pin down two things
the label leaves open. Four questions:

1. Batch allergen testing — what method, what LOD/LOQ, and
   what ppm threshold? "Non-detectable" against a 2.5 ppm
   assay and against a 20 ppm assay are different numbers,
   and I need to know which one I'm reading.

2. What form is the soy in the liquid-culture medium —
   intact protein, hydrolysate, or peptone?

3. What was the lot-transition date to ErinaPrime material?
   I want to know whether my bottle predates it.

4. Is the label's 0.5% erinacine A a guaranteed minimum or
   the actual assay value? Nammex publishes ErinaPrime at
   1.0%, and your own wording is "at least 0.5%" — so my
   actual daily intake is either 2.5 mg or 5 mg and I can't
   tell which. If you have a COA for a recent lot, that
   would answer it directly.

Thank you,
Nathan Distelhorst
C2 · Portal message — Virginia Mason
Records pull — histology and PT/INR

A records request, not a clinical question. Both may already exist; nobody has to order anything.

VM patient portal · re: 6/29–7/3 admission
Two records questions from my 6/29–7/3 admission — I'm not
asking for anything new to be ordered, just for what's
already in the chart.

1. During the 6/29–7/3 enteroscopy, was gastric HISTOLOGY
   sent to pathology, or was H. pylori testing done by
   CLOtest / rapid urease only? If histology was sent, I'd
   like that result.

2. Could I have the PT / INR drawn during that same
   admission?

Context for #2 if it's useful: I have EPI, which causes
fat-soluble vitamin malabsorption including vitamin K, and
I've never had vitamin K status checked. If PT was
prolonged at the time of the bleed, that's worth knowing.

Thank you,
Nathan Distelhorst
C3 · Message — Dr. Kuhlman
Bring him the BP re-baseline and ask the 80 mg question

This is a prescriber question, genuinely his. The blood-pressure case for stepping up has evaporated — he should see the numbers that made it evaporate.

Naturopathic Health Center · patient portal
I re-analysed my full blood pressure export — 59 readings,
Jun 24 through Jul 22 — and the picture has changed enough
that I think it changes the telmisartan question.

Fasted morning, on 40 mg:
  Jul 4–8    136.6 / 66.9   (pulse pressure 69.7)
  Jul 9–22   120.8 / 66.5   (pulse pressure 54.3)

That's -15.8 systolic and -15.4 pulse pressure with no
antihypertensive change — and it happened despite Adderall
going back to full dose on 7/13 and citrulline coming out
on 7/10, both of which should have pushed the other way.

My question: at a fasted morning of 120.8/66.5, is there
any remaining indication for stepping to 80 mg? My
understanding is that if a case remains it would be renal
rather than hypertensive, which would turn on the UACR on
the ~8/1 draw.

One related item — both VM and Providence still have me
listed at telmisartan 80 mg daily. Could that be corrected
to 40 mg, along with the Adderall resumption to the full
70 mg/day?

Thank you,
Nathan Distelhorst

D · Daily, while you're busy

Two things. Nothing else needs your attention.

D1 · 5:55 AM
Morning BP — three readings, properly

Sit 5 min · feet flat · back supported · bare arm at heart level · no talking. Three readings one minute apart. Discard #1. Average #2 and #3. Log to Apple Health.
Why it's on this list: 9 of the last 16 days had any morning session and only one had all three readings. Every gate in the file rests on this number, and it's the sparsest series you have. If you do one thing daily this week, do this.

D2 · Everything else
Run the schedule exactly as written — no changes pending

Only ALCAR moved. Do not change anything else while you're away, including the two agmatine items flagged in the regimen — those are logged questions, not instructions. Adding a fourth change to the attribution window is the one thing that would actually cost you something.

E · When you're properly back

None of this is time-sensitive before ~8/1.

E1 · Research
Run the agmatine question through both models

Files are written and ready: the Perplexity brief and the stripped context file. This one is a corroboration pass — same prompt and context to both, independently, neither seeing the other. Then hand Claude both outputs and ask it to audit them against each other. That second step is your normal workflow and it's correct.
Why it matters: agmatine is 1,000 mg/day, it's the largest amine load in the stack, and no rationale for it has ever been written down in any revision. Re-derive it or retire it.

E2 · Labs
Chase the iron panel result

Requested 7/22–26. Ret-He (CHr) and TSAT are the ones that matter — ferritin alone will falsely reassure at 5 weeks post-op. This result is the only remaining measurement gate on R-ALA.

E3 · Free
Look up rs10830963 in your AncestryDNA raw file 10 min

Settles melatonin 1 mg vs 3 mg. Costs nothing, may make the question moot.

E4 · Measurement
Three days weighed protein, three days logged fluid

Protein is named in the file as the single largest untouched lever — the post-surgical target is roughly 112–150 g/day and you've never measured against it. Amino-9 and HMB are not food. Fluid intake has never been quantified either, and it sits on the hyponatremia differential.

E5 · ~8/1 gate
Prep the Wancata check-in and the full draw

Draw list, telmisartan reassessment, and the standing-constraints provenance sweep — every ❌ and ⚠️ in that section, one at a time, asked "point at the source." The soy rule survived five adversarial revisions by looking settled; assume at least one more does.

⚕️ Unchanged — call the team, not a supplement question

Dark or tarry stool · visible blood · new dizziness · resting heart rate climbing · fever or rigors · any BP reading ≥180/120 → that's a call, not a message and not a wait.

← Protocol Hub