The wearable record.
Cortexin + Bemitil: a maritime N-of-1 observation, with the daily measurements left visible.
Three phases. One person.
Before reported start: September 9–11 (3 records). Combined exposure: September 12–21 (9 records). After Cortexin stopped, with Bemitil continuing: September 22–27 (6 records).
41.6% → 50.8% → 52.5%
Mean daily REM + SWS share across the three phases. Main sleep averaged 249.7 → 290.2 → 248.5 minutes. The higher share does not mean the sleep obtained was sufficient.
*Working exposure dates reported by the participant. Dots show available cycles; no interpolation across missing records. The September 27 cycle was still open, so day strain is missing.
| Metric | Before reported start 3 sleep records | Cortexin + Bemitil 9 sleep records | Bemitil continued / Cortexin stopped 6 sleep records |
|---|---|---|---|
| REM + SWS share (%) | 41.6n = 3 | 50.8n = 9 | 52.5n = 6 |
| REM + SWS (min) | 103.7n = 3 | 148.1n = 9 | 128.7n = 6 |
| Main sleep duration (min) | 249.7n = 3 | 290.2n = 9 | 248.5n = 6 |
| Deep / SWS (min) | 56.7n = 3 | 69.4n = 9 | 56.7n = 6 |
| REM (min) | 47.0n = 3 | 78.7n = 9 | 72.0n = 6 |
| Sleep efficiency (%) | 89.0n = 3 | 93.1n = 9 | 94.5n = 6 |
| WHOOP recovery (%) | 31.0n = 3 | 54.1n = 9 | 61.5n = 6 |
| HRV (ms) | 29.3n = 3 | 33.4n = 9 | 33.7n = 6 |
| Resting heart rate (bpm) | 68.7n = 3 | 67.2n = 9 | 63.8n = 6 |
| Day strain | 14.8n = 3 | 9.6n = 9 | 9.2n = 5 |
| Sleep debt (min) | 81.7n = 3 | 92.0n = 9 | 94.7n = 6 |
| Sleep need (min) | 478.0n = 3 | 445.2n = 9 | 413.8n = 6 |
Inspect every daily value for REM + SWS share
| Cycle date | Reported phase | REM + SWS share (%) |
|---|---|---|
| 2026-09-08 | Before reported start | Missing |
| 2026-09-09 | Before reported start | 33.85 |
| 2026-09-10 | Before reported start | 42.37 |
| 2026-09-11 | Before reported start | 48.62 |
| 2026-09-12 | Cortexin + Bemitil | 43.25 |
| 2026-09-13 | Cortexin + Bemitil | 57.86 |
| 2026-09-14 | Cortexin + Bemitil | 46.51 |
| 2026-09-15 | Cortexin + Bemitil | 52.63 |
| 2026-09-17 | Cortexin + Bemitil | 46.76 |
| 2026-09-18 | Cortexin + Bemitil | 70.85 |
| 2026-09-19 | Cortexin + Bemitil | 46.59 |
| 2026-09-20 | Cortexin + Bemitil | 44.24 |
| 2026-09-21 | Cortexin + Bemitil | 48.51 |
| 2026-09-22 | Bemitil continued / Cortexin stopped | 68.87 |
| 2026-09-23 | Bemitil continued / Cortexin stopped | 61.13 |
| 2026-09-24 | Bemitil continued / Cortexin stopped | 47.76 |
| 2026-09-25 | Bemitil continued / Cortexin stopped | 59.52 |
| 2026-09-26 | Bemitil continued / Cortexin stopped | 32.68 |
| 2026-09-27 | Bemitil continued / Cortexin stopped | 45.23 |
How this was calculated.
The participant supplied a WHOOP ZIP containing 19 physiological cycles and 34 sleep/nap records, plus workouts and journal entries. This analysis uses the main-sleep fields in physiological_cycles.csv, avoiding duplicate counting of sleeps from the second file. It is not a 24-hour total including naps.
Dates are the cycle-start dates as written in the export, without timezone conversion. September 8 has no main-sleep fields and is excluded from every phase mean; September 16 has no separate cycle-start record. September 15’s sleep crosses midnight. The exposure markers are approximate calendar boundaries, not verified dose-before-sleep timestamps.
Restorative share = 100 × (deep/SWS minutes + REM minutes) ÷ asleep minutes, calculated per record and then averaged. Other values are arithmetic means of available measurements within sleep-bearing cycles. Missing values are never zero-filled. This reproduces the preliminary chat averages while making the 3 / 9 / 6 denominators explicit.
What remains unresolved.
Concurrent exposure
Bemitil began around the same time and continued after Cortexin. “Post-Cortexin” is not an untreated follow-up or a clean Bemitil experiment.
Changing workload
Mean day strain fell from 14.8 to 9.6 to 9.2. Watch schedules, sleep pressure, naps, other substances and normal variation may explain some or all changes. No adjustment or causal model was fitted.
Small baseline
Three pre-period sleep records cannot establish a stable baseline. WHOOP sleep stages are device estimates, not polysomnography; observations are serially related, not independent trial participants.
CHRONAS, a separate clinical report in patients with cerebral ischemia, is included in the research archive. It does not validate these personal wearable findings.
A record that can grow.
Updated through the September 27 export. Weekly participant-supplied exports can extend the record using the same calculation rules. No live WHOOP connection or automatic weekly upload is configured. Journal answers, workout details, raw timestamps and the original ZIP are not published here.