diff --git a/Runs/2026-09-18.md b/Runs/2026-09-18.md
index 2e22065..734604f 100644
--- a/Runs/2026-09-18.md
+++ b/Runs/2026-09-18.md
@@ -54,7 +54,7 @@ tags: [run, morning, comeback, negative-split]
 | Cadence | 168 spm | 162 spm | 150 spm | 168 spm | **164 spm** |
 | Sleep HRV | 127ms | 123ms | 119ms | 98ms | **127ms** |
 
-**The engine is unchanged and the tissue is finally behaving.** 6'04" at 152 bpm is the same efficiency as Sep 9 (5'58"/157) and Sep 16 (6'03"/150) — VO₂Max held at 46, HR recovery 24 bpm in his usual 18–24 band, balance a clean 50/50. What's different today is the *shape*: the last three runs all faded (Sep 13: 6'52"→8'36"→9'23" → stop; Sep 16: 5'56"→6'08"), and today went the other way — 6'15" → 6'10" → 5'45". That is the warm-up pattern finally doing its job instead of the achilles announcing itself at km 2.
+**The engine is unchanged and the tissue is finally behaving.** 6'04" at 152 bpm is the same efficiency as Sep 9 (5'58"/157) and Sep 16 (6'03"/150) — VO₂Max held at 46, HR recovery 24 bpm in his usual 18–24 band, balance a clean 50/50. And this was **fasted — no breakfast, no coffee** — so 152 bpm is the *true* cost of the day, not a stimulant artefact (fuel's a non-issue at 3 km; last night's dinner covered it). What's different today is the *shape*: the last three runs all faded (Sep 13: 6'52"→8'36"→9'23" → stop; Sep 16: 5'56"→6'08"), and today went the other way — 6'15" → 6'10" → 5'45". That is the warm-up pattern finally doing its job instead of the achilles announcing itself at km 2.
 
 ## Verdict