Tretinoin — the dose/site question

Is a tiny facial amount a different animal than the massive nightly back application? (Nick asked, Jul 9 2026)

What actually happened (the avoid)

Nick used tretinoin heavily for BACK acne → an April–May 2026 depression cascade (compounded by grief load) → stopped. Documented as a genuine SAFETY avoid, not a non-responder reaction. So the avoid is real and specific: aggressive, high-surface-area tretinoin. The open question is whether the exposure — not the molecule — is what caused it.

The mechanism — dose × surface area × frequency

Topical tretinoin's systemic exposure scales with:

Surface area. The back is ~15–20× the area of the peri-orbital/face zones. Absorption is roughly proportional to treated area.
Amount/concentration. A heavy nightly back slather is far more product than a pea-sized 3×/week facial dab.
Frequency. Nightly vs 3×/week ≈ a further ~2× in cumulative dose.
Skin integrity + occlusion. Inflamed/broken back skin + a shirt/lying-on-it both increase absorption vs intact facial skin left open to air.

Net: nightly heavy back use could plausibly be one-to-two orders of magnitude more systemic retinoid exposure than a tiny intermittent facial dose — a different exposure class, not a rounding difference.

The retinoid→mood link (the crux)

The strong, well-documented retinoid/depression signal is with oral isotretinoin — i.e. systemic exposure acting on CNS retinoic-acid signaling. Topical tretinoin normally produces minimal systemic levels, which is why it's generally mood-safe at typical facial doses. Nick's cascade is the tail case: a high-surface-area, high-dose, occluded application likely pushed systemic exposure toward the range where the isotretinoin-type mechanism can bite — especially layered on an already-loaded nervous system (grief + collapse recovery + HPA suppression).

The hypothesis: the crash was driven by systemic dose, not the molecule per se. A tiny, intermittent, open-air facial dose is a fundamentally smaller exposure and may be tolerable where the back slather was not.

🔴 The honest counter-weights (both/and)

N=1 sensitivity is the wildcard. He had a real cascade; "lower dose is lower risk" is true but not "zero risk."
Context matters as much as dose. The crash coincided with peak grief + suppressed cortisol. In a stable nervous-system window the same small dose is a very different bet.
His skin isn't stable yet (active back acne + Malassezia) — any trial is gated behind stable skin regardless.
Cheaper/safer first-line exists (GHK-Cu, fascia + mewing, sun protection for pigment). Tret is the escalation, not the opener.

The cautious path IF he ever trials it (for his derm, not self-start):

Skin stable first (no active acne/Malassezia).
Stable nervous-system window (not during a grief/stress spike; ideally clear of kratom-taper turbulence).
Tiny, intermittent, FACIAL only — lowest strength, pea-size or less, 1–2×/week, open-air (no occlusion), one small zone.
Explicit mood tripwire — pre-agree with Chantelle/derm on early-warning signs and a hard-stop at the first hint.
Never the back-slather pattern again — high surface area + occlusion is the exposure that burned him; that stays permanently off.
Loop the derm — a supervised trial decision, not a casual re-add.

Verdict (for review): the mechanism supports Nick's intuition — the back cascade was most likely a systemic-dose phenomenon (area × amount × frequency × occlusion × a fragile nervous system), not proof the molecule is off-limits at any dose. A tiny intermittent facial dose is a plausibly-different-and-safer exposure — worth holding as a "maybe / future supervised trial," gated on stable skin + a stable NS window + a mood tripwire, with the back-slather pattern permanently retired. Not adopted; his call + his derm's.

Source: projects/personal/health/tretinoin-dose-mechanism-2026-07-09.md · memory/health-full.md (avoid list + appearance threads) · back-scars/plan.md. A fresh web/deep-research citation pass is still owed. Grounded in established retinoid pharmacology + Nick's files.