---
name: ben-health
description: Ben's genotype and bloodwork master file, the single source of truth for anything touching his biology. Load this before answering any question about supplements, vitamins, minerals, nootropics, peptides, hormones, TRT, testosterone, estradiol, AI/aromatase inhibitors, pregnenolone, DHEA, retatrutide or GLP-1s, sleep or sleep drugs, caffeine, alcohol, anxiety compounds, pregabalin, statins, lipids, cardiovascular risk, cognition or Alzheimer's risk, inflammation, thyroid, or any specific drug or compound by name. Also load whenever Ben pastes a health claim, a study, a product label, a podcast clip, or a lab result and asks "is this true", "is this good for me", "should I take this", "worth it", or "does this apply to me". Use it even when he does not mention genetics or bloodwork, because the whole point is that the answer changes based on data he will not restate. Do not answer from general population defaults when this file exists.
---

# Ben: genetics and bloodwork

Single source of truth for Ben's genotype and lab data. Everything here comes from his own records, not from inference.

**Genetic source:** Geneus DNA raw export, 631,500 SNPs, analysed 17 February 2026
**Bloodwork source:** N Health (BDMS), Bangkok. Most recent full panel 25 April 2026. Lab IDs DNS904165, DNS904166, DNS904170.
**File compiled:** 27 August 2026

## How to answer with this file

When Ben pastes a claim or asks "is this good for me", split the answer in two. He has asked for this explicitly and merging them produces answers he cannot act on.

1. **Is it true in general?** State confidence. Search if it is a current fact rather than settled mechanism.
2. **Does it apply to HIM?** Given the genotype, the labs, and interactions with whatever he is currently running.

Skip "consult a doctor" boilerplate. He is deep on his own biology and it wastes the response. Give the verdict, put real uncertainty in a clause, and route anything touching the cardiac side past whoever manages it.

Answer format follows his standing preferences: short compact bullets, no long prose paragraphs, no dashes or em dashes anywhere in the text.

## Staleness, resolved 27 August 2026

**The follow up panel was never run.** On 27 August Ben supplied what he believed were his most recent tests. Every page carries collected date **25 April 2026**. The PDF had simply been exported from the portal on 4 July 2026, and the file date is what made it look newer.

So the working data is four months old, and the four markers that would settle the biggest open questions are still untested.

When a lab report arrives, **read the collected date on the report, not the file date.** That single check is what caught this.

The protocol in `references/stack.md` is the April 2026 snapshot and is known to be out of date. See "Most recent self-report" below.

**Never measured, at any point:** ApoB, homocysteine, serum Lp(a), omega 3 index, 4 point salivary cortisol with CAR, RBC magnesium, RBC zinc, cystatin C, reverse T3, p-tau217, Abeta42/40, NfL, GFAP.

**DHT and zinc were previously listed here in error.** Both were on the April panel. DHT 574.1 pg/mL, zinc 549.24 ug/L. See `references/bloodwork.md`, both carry real interpretive weight.

When one of these would settle a question, say so plainly rather than reasoning around the gap. Three unmeasured markers (ApoB, homocysteine, DHEA-S response) determine whether lanes he is already treating are actually handled. Getting those beats adding any new compound, and that comparison is usually the most useful thing to tell him.

## Most recent self-report, 27 August 2026, unconfirmed

Ben reports he has dropped most of the April stack because he felt nothing from it, and is currently running:

- Pregnenolone 100 to 200 mg AM, new since April, not in any panel
- Pregabalin 75 mg situational, rare, for stressful trading days or dates. He likes it specifically because it relaxes him without mellowing him, and has a standing wish to be "a little less tense". Watch that drift, see the escalation note in `references/stack.md`
- TRT continuing, he quoted 160 mg/week split into two doses (April file says 180 mg/week, reconcile this)
- hCG continuing
- **Aromasin STOPPED, confirmed 27 August 2026**

Treat this as current until a panel says otherwise, but flag that pregnenolone at 100 to 200 mg has never been measured against his hormones, and that it lands on an axis he is already supplementing.

### The estradiol situation, currently the largest open risk in this file

Confirmed 27 August 2026: he restarted Aromasin around 11 April 2026, took it for roughly **2 to 3 weeks**, and stopped. **Estradiol has been unmanaged since about early May, roughly four months.**

Two things are stacking in the same direction and neither is being watched:

1. Supraphysiological TRT, total T 1095 at the last measurement
2. **hCG, which preferentially raises estradiol relative to testosterone.** Forest 1975 at stimulation doses: E2 rose 2.3–4.0x while T rose only 1.4–1.9x. Honest caveat: that magnitude is established at 2,000–5,000 IU. At the 250–500 IU used alongside TRT the two published series are underpowered (Hsieh 2013, n=26, ~68% E2 rise, p=0.11) or null (n=31, retrospective). Genuinely unresolved rather than settled either way.

**Pregnenolone is NOT a third contributor — that claim is withdrawn.** No human study shows oral pregnenolone raising testosterone or estradiol in men.

His off-AI history is the receipt: **62, then 86, then 89 pg/mL** across 2024 and 2025. The single measurement of 29 was taken 14 days into the Aromasin restart and is not representative of his unmedicated state.

**The counter-argument he will make is "I feel fine", and the answer is that he felt fine at 89 too.** What happened during that window was invisible to feel: **PSA doubled from 0.505 to 1.146**, then fell back to 0.864 once dose and E2 were controlled. That is the strongest single argument in this file, because it is his own data and it maps to his own blind spot. Anything acting on markers rather than receptors will not be felt, which is exactly why he stops taking it.

**Push this on prostate and cardiovascular grounds ONLY. Do not claim high E2 is causing his anxiety or mood — that link does not exist in the literature.** See `references/anxiety.md`: the headline citation for "high E2 and mood in men" is a 3 pg/mL difference with no anxiety measure, trans women at 100–200 pg/mL show no anxiety worsening, and nobody has ever tested E2 at 80–90 pg/mL against anxiety in men. Using a wrong argument for a right action costs credibility he will not give back.

Do not moralise about this. State the mechanism, produce the PSA numbers, and push for the draw.

**The pregnenolone interaction to raise:** he was started on DHEA 25 mg to correct a low DHEA-S (103.4, range floor 139.7). Pregnenolone converts upstream of DHEA. He is now dosing the same lane twice from two directions, which both confounds the DHEA-S recheck and makes the ceiling unpredictable. **With Aromasin stopped there is no aromatase management at all**, though note pregnenolone itself does not appear to feed that cascade in humans.

**Why pregnenolone lands on him the way it does — EARLIER EXPLANATION RETRACTED 27 Aug 2026.** A previous session argued that weak 5-alpha-reductase left the calming allopregnanolone arm quiet. Measured human PK contradicts this: Sripada 2013 (n=31 men, 400 mg oral) found allopregnanolone rises **7.1-fold**, pregnenolone 3.2-fold, **DHEA-S falls ~5%**, DHEA and cortisol unchanged. Testosterone and estradiol were not raised in any human study. Two claims are therefore withdrawn: that pregnenolone suppresses his allopregnanolone, and that it feeds the aromatisation cascade.

**What stands:** the ~5% DHEA-S reduction works mildly against the DHEA 25 mg he was given for a DHEA-S of 103.4. And the real mechanistic caveat is the **inverted-U**: partial elevation of allopregnanolone is *anxiogenic* in a susceptible minority (severe reactions 3–8%, moderate up to 25%, Backstrom's group), with low-to-moderate levels increasing amygdala reactivity and higher levels decreasing it. Whether pregnenolone helps or hurts him depends where he sits on that curve, which is not predictable in advance.

**Follows from that:** do not casually suggest a 5-alpha-reductase inhibitor (finasteride, dutasteride). The mechanism for anxiogenesis is coherent and the depression epidemiology is real but disputed (register HRs 1.3–1.9; the 2.2-million-patient meta-analysis is null).

**Pregnenolone did NOT cause his baseline tension.** He falsified this himself on 27 August: tense on days he takes none, and tense for a decade before he started it.

## Headline findings, the four that drive most answers

1. **Cardiovascular outranks cognitive.** Triple homozygous risk at 9p21 (rs10757274 AA, rs4977574 AA, rs1333049 GG), the strongest known common genetic risk factor for coronary artery disease, stacked on APOE e4 and a pro inflammatory cytokine profile. LDL-C looks excellent (42) and understates the risk. **ApoB is the correct instrument and has never been measured.**
2. **The lipid problem is triglycerides, not LDL.** TG 177, VLDL 35, TG/HDL ratio 4.32, HDL 41. Small dense LDL pattern. EPA is the lever. Any advice framed around "lower his LDL" is working on a problem he does not have.
3. **APOE e4 carrier, probably e3/e4, inferred not directly typed.** Three proxy SNPs heterozygous plus rs7412 CC (no e2). BDNF Val/Val and FKBP5 CC are genuine counterweights, not consolation.
4. **Pro inflammatory baseline.** IL-6 GG (high production) plus IL-10 CC (low braking). hsCRP 2.44, above the under 1.0 that matters for an e4 carrier. Omega 3, exercise and sleep have outsized returns on this specific combination, which is why they keep being the answer.

## Corrections to earlier profiles, read before trusting any prior summary

Older summaries in prior chats contain errors that get repeated. The raw Geneus data supersedes them.

| Claim floating in old profiles | Raw data truth |
|---|---|
| "MTHFR C677T heterozygous" | **Wrong.** rs1801133 is **GG, wild type.** Only A1298C (rs1801131) is heterozygous, a mild reduction. Methylation urgency is far lower than previously assumed, and TMG plus methylfolate are not genotype mandated. Homocysteine settles it and has never been run. |
| "APOE e4 confirmed" | Inferred from three proxy SNPs, not direct rs429358 genotyping. Confidence is high but direct typing is still outstanding. |
| DRD2 rs1079597 "GG" | Raw data shows CC. Strand orientation may explain it. |
| "Atherogenic high LDL e4 case" | LDL direct is 42. The problem is TG and VLDL. |
| "Needs more vitamin D" | He is at **92 ng/mL, saturated.** Do not increase. |

## Standing interpretation rules

- **Never infer methylation urgency from the word MTHFR.** C677T is wild type.
- **Caffeine is a clearance problem, not a sensitivity problem.** CYP1A2 AC means slow clearance, but ADORA2A CC means normal receptor sensitivity. So the issue is duration in system, not jitteriness. Combined with CLOCK AA, afternoon caffeine degrades sleep architecture. Hard cutoff at noon is the genotype indicated rule. This also explains why a single morning espresso works well for him and daily creeping doses do not.
- **Serotonergic agents work uphill.** TPH2 TT (low synthesis) plus HTR1A CC (increased autoreceptor braking) is a double hit on serotonin capacity. Sleep repair and GABA levers move his mood more efficiently than SSRIs or serotonergic supplements.
- **Anxiety is glutamate driven and GABA responsive, not serotonergic.** CACNA2D1 CC confirms the clean pregabalin fit at 75 mg. This is a genotype match, not a habit.
- **He is constitutionally tense and describes it as lifelong.** MAOA TT (reduced monoamine oxidase A, elevated noradrenaline), ADRB2 Arg/Arg (enhanced beta-2 agonist response), GABRA2 TT (anxiety linked GABA-A subunit), TPH2 TT plus HTR1A CC (low serotonin capacity), CLOCK AA. That combination is a high noradrenergic, low serotonin, GABA sensitive phenotype. **Treat baseline tension as a trait to be managed, not a symptom to be cured**, and do not chase it with cortisol or adrenal interventions.
- **Cortisol and adrenal support are a dead end for him.** FKBP5 CC means efficient HPA recovery and his AM cortisol was 10.3, mid range. Adaptogens aimed at cortisol are low yield. This rules out a whole category of "stress supplements" before he spends money on them.
- **His chronic tension is untreated GAD, not a supplement deficiency.** Lifelong, background, functional. He has never had one adequate trial of any first-line treatment. **Read `references/anxiety.md` before answering anything about his anxiety** — it carries the full evidence review and the corrections to several confident claims made in earlier sessions.
- **He is taking 75 mg pregabalin, which is HALF the lowest dose ever tested for GAD.** Licensed range 150–600 mg. 150 mg failed to separate from placebo in one trial. There is no efficacy data at 75 mg at all.
- **Beta blockade for his tension: withdrawn.** Steenen 2016, 8 RCTs, 202 participants, zero in GAD, "insufficient to support routine use in any anxiety disorder". The somatic-only teaching rests on n=6 from 1974.
- **Dopaminergic tone is already high.** MAOA TT (reduced clearance) plus DRD2 rs6277 AA (reduced D2 expression) means elevated drive with reduced postsynaptic feedback. Relevant to stimulants, to trading behaviour, and to why compounds that mellow him cost him drive.
- **Statins are fully available if ever needed.** SLCO1B1 TT, no myopathy barrier. Worth remembering given the 9p21 picture.
- **Retatrutide dose response may be non standard.** GLP1R rs2268657 TC. Titrate empirically.
- **Vitamin D is saturated at 92 ng/mL.** Do not add. VDR is intermediate across all three SNPs, which argues for keeping the level up, not pushing past 100.
- **B12 is 1482, well above range, supplement driven.** Standalone B12 already dropped. Do not add more.
- **Zinc was 549.24 against a floor of 551, low while he was taking 25 to 30 mg daily.** He was not overshooting, he was at the floor despite supplementing. Plasma zinc is insensitive and his IL-6 GG plus hsCRP 2.44 depresses it further, so treat it as suggestive rather than definitive. This is the one dropped item with a measured abnormality behind reinstating it.
- **Modest 5-alpha-reductase.** DHT:T ratio roughly 5.2% where typical is around 10%. Bears on hair loss drugs, prostate and sebum. Does NOT explain his pregnenolone response — see the retraction above.
- **PSA velocity matters more than absolute value.** It doubled during the high T and high E2 window (0.505 to 1.146) then fell to 0.864 once dose and E2 were controlled. Keep it on every panel.
- **Haematocrit on TRT:** 47.7 currently, was 53.0. Donate at 50 to 52.
- **Stack pressure exceeds evidence in places.** When he asks about adding something, the honest comparison is usually against measuring one of the never tested markers instead.

## Behavioural notes that change what to say

Ben optimises inputs relentlessly and rarely reviews outputs. This shows up in this domain constantly: compounds get added, the marker they were meant to move never gets rechecked. When he asks about adding something, **first ask what the last thing he added actually did to its marker.**

The 27 August 2026 episode is the clean example. Asked for his most recent bloodwork, he supplied a July-exported PDF of the April panel, believing it was new. He had not run the follow up. He also had two markers (DHT, zinc) sitting in his own records that his master file listed as never measured. **The data he already owns is under-read, which is the same pattern as the un-reviewed outputs, one layer down.** When he says "here are my latest tests", check the collected date.

Specifically outstanding:

1. **Book the panel.** ApoB, homocysteine, DHEA-S, progesterone, omega 3 index, RBC zinc, repeat DHT. It has been outstanding since roughly July.
2. Did DHEA 25 mg move DHEA-S from 103.4 toward 250 to 400? Never checked, and pregnenolone now confounds it.
3. **Did E2 normalisation actually fix the 3 to 5am waking?** That was the entire purpose of restarting Aromasin. Ask for the result, not the plan.
4. Did the Dayvigo taper trial happen, and what happened?
5. Is he still taking Aromasin, or did it go with the rest of the dropped stack? Unmanaged E2 ran him to 89 across 2024 and 2025.

## Reference files

Read the one that matches the question. Do not load all three by default.

- **`references/genetics.md`** — full SNP tables by system: APOE and Alzheimer's loci, dopamine, serotonin, GABA and glutamate, adrenergic and stress, neuroplasticity, CYP pharmacogenomics, drug transport, cardiovascular and lipid, weight and GLP-1, methylation and VDR, inflammation, sleep and circadian.
- **`references/bloodwork.md`** — the full 25 April 2026 panel with draw context, plus multi point historical trends for estradiol, testosterone, SHBG, haematocrit, PSA and liver markers.
- **`references/anxiety.md`** — his chronic anxiety: the full treatment evidence ranking, why pregabalin tops it for his genotype, the dose problem, the honest costs, and a table of retracted claims. **Read this for any anxiety, tension, stress or mood question.**
- **`references/stack.md`** — the April 2026 protocol, injectables, daily timing, and the interaction notes to carry forward.

Draw context matters for interpreting the panel and lives at the top of `bloodwork.md`. Do not read hormone values without it, particularly the estradiol number, which was run on standard immunoassay rather than LC-MS and was taken 14 days into an Aromasin restart.

## What this file is not

Not a physician and not clinical advice. Route anything touching the cardiac side past whoever manages it. The value here is that it stops general population defaults being applied to someone whose genotype and labs make several of those defaults wrong.