Sources: health-report-2026-08-26-to-2026-08-27.json (Apple Watch sleep stages, exported 05:45 on 08-27) + OSCAR_Greg_by_Day_2026-08-26.csv (ResMed AirSense/AirFlow 11 via OSCAR). Session start 22:53 on 08-26 — 7 min later than 08-25's 22:46 (bedtime trajectory reversed slightly). CPAP hours 6.77 vs. Watch TIB 6.68 (Δ ≈ 5 min, CPAP longer — normal alignment).
DEEP-SLEEP COLLAPSE RESOLVED — YESTERDAY'S 6.63% WAS A ONE-OFF. Deep bounced from 6.63% on 08-25 → 17.98% tonight (+11.35 pts). Rank 44/52 (top quintile — high deep %), and comfortably above 30-day mean of 13.11% and 7-day mean of 11.51%. This decisively resolves the 08-25 report's binary question: deep did not stay below 12% for a 2nd consecutive night, so reading (b) — one-night Apple Watch stage-detection artifact — is now the favored explanation over reading (a) — sustained unmeasured stressor. The physiological alternative (alcohol, illness, meal timing) would have propagated at least partially into tonight; the actual 08-25 → 08-26 delta (+11.35 pts) is larger than the 08-24 → 08-25 delta (−5.31 pts), which is consistent with a measurement flip on 08-25 fully reversing rather than a stressor half-clearing. Case closed on the deep-sleep question unless deep drops again below 10% inside the next 5 nights.
NEAR-RECORD ARCHITECTURE NIGHT — REM AT #3/52, AWAKE AT #4/52, EFFICIENCY AT #2/52, DEEP AT #9/52. Tonight is the strongest single architecture night in the matched history on multiple axes simultaneously. REM 31.27% is the 3rd-highest ever recorded (only 07-25 at 32.89% and 07-22 at 31.61% beat it), awake time 4.52 min is the 4th-lowest ever, efficiency 98.87% is the 2nd-highest ever (behind 08-21's 99.45%), and deep 17.98% ranks 9/52 (top quintile). The composite is unusual: nights that peak on REM% typically pay some cost on deep or awake, but tonight is elevated on all three architecture channels with no offset. The efficiency streak (≥ 95%) is now 11 consecutive nights (08-16 → 08-26) — the longest such streak in the matched window (previous record was the 7-night 08-19 → 08-25 streak that closed yesterday). This is a genuinely excellent sleep night — one of the best on record.
PRESSURE CONFIRMED AT NEW OPERATING POINT — 08-25's PARTIAL REINFLATION IS THE NEW BASELINE. P_avg 7.05 → 7.09 (+0.04); P_90 9.10 → 9.42 (+0.32). The 08-25 report's central thesis — "the new efficient operating point is landing in the 7.0–7.5 P_avg band" — is now confirmed on night 2 of the new regime. P_avg 7.09 is rank 12/52 (bottom quintile, still ~0.35 cmH₂O below the pre-cluster 30-day baseline of 7.4), and P_90 9.42 is rank 9/52 (bottom decile). The titrator held tonight rather than continuing to reinflate toward the 7.5+ baseline, and event burden dropped further as a result (AHI 1.81 → 1.61 → 1.33; H 3 → 2 → 1). The auto-titrator has settled: the pre-cluster 7.4–7.5 pressure baseline is no longer the operating point, and the new baseline appears to be 7.0–7.2 P_avg / 9.1–9.5 P_90. Either the underlying airway condition genuinely improved over the cluster, or the leak-inflated pre-cluster machine-resistance estimate is now unwound. Either way, event burden at the new lower pressure is better than it was at the higher baseline, which is the definition of a superior operating point.
CA CHANNEL: 08-25's CA=2 WAS NOISE, NOT A TREND. CA count 2 → 1 (rank 21/52 — at median). The 08-25 report's low-pressure CA sub-pattern watch trigger ("CA count ≥ 3 for 2+ nights with pressure ≤ 7.5") is decisively not tripped. Tonight's CA=1 at P_avg 7.09 fits the normal baseline distribution (median CA = 1). Off active watch for CA. The observation that 5 of the last 7 CA≥2 nights have been at P_avg ≤ 7.6 remains a valid background pattern, but the current pressure regime (7.0–7.2) is not producing sustained CA elevation.
08-25 PREDICTIONS: 4 OF 5 CLEAN HITS, 1 MISSED. The 08-25 report set 5 explicit "what to watch on 08-26" checkpoints: (1) Deep ≥ 12% = tonight was a one-off → Actual 17.98%, CLEAN HIT — one-off confirmed; (2) CA ≤ 1 = tonight's CA=2 was noise → Actual CA=1, CLEAN HIT; (3) P_avg 7.0–7.5 confirms new operating point → Actual 7.09, CLEAN HIT — center of predicted band; (4) OA count 3–8 confirms cluster fully done → Actual OA=7, CLEAN HIT — center of band; (5) Bedtime 22:30 or earlier confirms improvement trajectory has capacity → Actual 22:53, MISSED — reversed 7 min later. Verdict: 4 of 5 clean, 1 informative miss (bedtime capacity did not materialize — 22:46 → 22:53 is directional regression). The 08-25 report's framing on the deep-sleep question and the pressure question both scored decisive, correct calls.
BEDTIME TRAJECTORY REVERSED — 4-NIGHT IMPROVEMENT ARC ENDED. Trajectory across 5 nights: 23:35 → 23:04 → 22:53 → 22:46 → 22:53. The 4-night improvement direction (−49 min total, decelerating) that peaked at 22:46 on 08-25 has ended; tonight is +7 min back the wrong way. The 22:46 low is now looking like a local trough, not a new floor. Current operating band is now 22:46–22:53, 16–23 min past the 22:15–22:30 target. Not clinically important — TIB 6.68 h and efficiency 98.87% are both excellent — but the "trajectory has capacity to reach 22:30" thesis from the 08-25 report is now weaker. Behavior is settling around 22:45–22:55 as a soft floor.
HYDEPNEA COUNT AT NEAR-FLOOR: H=1 IS RANK 5/52. Hypopneas dropped from 2 → 1 tonight — only 4 nights in the matched history had 0 hypopneas. Combined with RERA=0 (rank 1/52, at floor) and Leak_90 = 0 (rank 1/52, at floor), three of the CPAP event/leak channels are simultaneously at or near their historical floor tonight. This is an unusually clean CPAP night.
| Metric | Value |
|---|---|
| Bedtime (session start) | 22:53 on 08-26 (+7 min vs. 08-25; still +23–38 min past 22:15–22:30 target; trajectory reversed) |
| Time in bed | 6.68 h (400.9 min) — rank 32/52; −2.5 min vs. 08-25 (403.4); +0.8 min vs. 30-day (400.2) |
| Total sleep duration | 6.61 h (396.4 min) — rank 34/52; +3.0 min vs. 08-25 (393.4) |
| Sleep efficiency (TST/TIB) | 98.87% — rank 50/52 (2ND-HIGHEST EVER); +1.36 pts vs. 08-25 (97.51%); +2.92 pts vs. 30-day (95.95%) |
| Awake time | 4.52 min — rank 4/52 (4TH-LOWEST EVER); −5.5 min vs. 08-25 (10.04); −12.3 min vs. 30-day (16.82) |
| Core sleep | 3.35 h (201.2 min) — 50.76% — rank 5/52 (bottom quintile — very low core %); −18.75 pts vs. 08-25 |
| Deep sleep | 1.19 h (71.3 min) — 17.98% — rank 44/52 (top quintile); +11.35 pts vs. 08-25 (6.63%); +4.87 pts vs. 30-day (13.11%) |
| REM sleep | 2.07 h (123.9 min) — 31.27% — rank 50/52 (3RD-HIGHEST EVER); +7.42 pts vs. 08-25 (23.85%); +7.81 pts vs. 30-day (23.46%) |
| CPAP usage | 6.77 h — rank 30/52; −0.08 h vs. 08-25 (6.85); +0.00 h vs. 30-day (6.77) |
| AHI | 1.33 — rank 12/52 (bottom quintile); −0.28 vs. 08-25 (1.61); −0.45 vs. 30-day (1.78) |
| RDI | 1.33 — RERA contributes 0 events (rank 1/52, at floor) |
| OA / H / CA / UA / RERA | 7 / 1 / 1 / 0 / 0 — OA rank 15/52; H rank 5/52 (near floor); CA rank 21/52 (at median); RERA rank 1/52 (at floor) |
| OA per hour | 1.03 /h — rank 20/52; +0.01 vs. 08-25 (1.02); −0.16 vs. 30-day (1.19) |
| Pressure avg | 7.09 cmH₂O — rank 12/52; +0.04 vs. 08-25 (7.05); −0.32 vs. 30-day (7.41) — held at new operating point |
| Pressure 90th %ile | 9.42 cmH₂O — rank 9/52 (bottom decile); +0.32 vs. 08-25 (9.10); −0.52 vs. 30-day (9.94) |
| Pressure max | 10.54 cmH₂O — vs. 11.12 last night (slight drop) |
| Leak avg | 0.72 L/min — rank 24/52; −0.01 vs. 08-25 (0.73); −0.10 vs. 30-day (0.82) |
| Leak 90th %ile | 0.0 L/min — rank 1/52 (at floor); −2.4 vs. 08-25 (2.4); −2.53 vs. 30-day (2.53) |
| Leak max | 16.8 L/min — vs. 12.0 last night (elevated in the single-event tail but not sustained) |
| Compliance flag | 1 (compliant) |
| Date | AHI | OA | H | CA | RERA | Leak avg | Leak 90 | P_avg | P_90 | Hours | Deep % | REM % | Awake (min) | Eff |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 08-11 | 2.57 | 12 | 4 | 0 | 0 | 0.42 | 1.2 | 7.28 | 10.02 | 6.62 | 13.57 | 25.24 | 21.07 | 94.61% |
| 08-12 | 2.68 | 13 | 3 | 2 | 1 | 0.28 | 1.2 | 7.99 | 10.00 | 6.72 | 20.78 | 19.11 | 28.10 | 92.80% |
| 08-13 | 1.29 | 8 | 0 | 2 | 0 | 0.20 | 1.2 | 7.01 | 9.54 | 7.75 | 10.93 | 26.70 | 20.07 | 95.60% |
| 08-14 | 1.63 | 11 | 1 | 0 | 1 | 0.22 | 0.0 | 7.94 | 10.86 | 7.38 | 9.73 | 29.20 | 33.62 | 92.19% |
| 08-16 | 1.13 | 3 | 3 | 1 | 2 | 0.91 | 4.8 | 7.92 | 10.94 | 6.22 | 14.73 | 19.78 | 6.02 | 98.34% |
| 08-17 | 1.11 | 5 | 2 | 0 | 0 | 1.14 | 1.2 | 7.57 | 9.54 | 6.28 | 14.21 | 25.35 | 7.53 | 97.95% |
| 08-18 | 1.63 | 7 | 2 | 1 | 1 | 1.19 | 3.6 | 7.55 | 11.11 | 6.12 | 8.86 | 18.71 | 13.55 | 96.34% |
| 08-19 | 2.20 | 10 | 5 | 0 | 0 | 1.92 | 4.8 | 8.18 | 10.34 | 6.83 | 16.50 | 24.11 | 7.02 | 98.25% |
| 08-20 | 2.55 | 14 | 2 | 0 | 0 | 0.69 | 1.2 | 7.80 | 11.16 | 6.27 | 7.68 | 20.91 | 13.05 | 96.43% |
| 08-21 | 2.82 | 15 | 6 | 1 | 0 | 0.20 | 0.0 | 7.68 | 10.14 | 7.80 | 14.18 | 25.49 | 2.51 | 99.45% |
| 08-22 | 1.54 | 8 | 2 | 0 | 3 | 0.74 | 0.0 | 8.00 | 10.72 | 6.48 | 13.82 | 17.51 | 18.06 | 95.31% |
| 08-23 | 0.77 | 3 | 2 | 0 | 2 | 0.94 | 3.6 | 7.53 | 10.66 | 6.48 | 14.78 | 30.09 | 7.53 | 98.04% |
| 08-24 | 1.81 | 9 | 3 | 0 | 1 | 0.37 | 0.0 | 6.64 | 8.68 | 6.63 | 11.94 | 25.33 | 10.54 | 97.32% |
| 08-25 | 1.61 | 7 | 2 | 2 | 0 | 0.73 | 2.4 | 7.05 | 9.10 | 6.85 | 6.63 | 23.85 | 10.04 | 97.51% |
| 08-26 | 1.33 | 7 | 1 | 1 | 0 | 0.72 | 0.0 | 7.09 | 9.42 | 6.77 | 17.98 | 31.27 | 4.52 | 98.87% |
(08-15 not in table: no matched Watch sleep-stage session.)
Cluster arc + full release + new operating point (08-18 → 08-26):
| Date | AHI | OA | H | OA+H | OA/h | Leak_avg | P_90 | RERA | CA | Deep % | REM % | Interpretation |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 08-18 | 1.63 | 7 | 2 | 9 | 1.14 | 1.19 | 11.11 | 1 | 1 | 8.86 | 18.71 | Leaky, moderate events |
| 08-19 | 2.20 | 10 | 5 | 15 | 1.46 | 1.92 | 10.34 | 0 | 0 | 16.50 | 24.11 | Peak leak, rising events |
| 08-20 | 2.55 | 14 | 2 | 16 | 2.23 | 0.69 | 11.16 | 0 | 0 | 7.68 | 20.91 | Leak-fixed, events still rising |
| 08-21 | 2.82 | 15 | 6 | 21 | 1.92 | 0.20 | 10.14 | 0 | 1 | 14.18 | 25.49 | Leak floor, event peak |
| 08-22 | 1.54 | 8 | 2 | 10 | 1.23 | 0.74 | 10.72 | 0 | 3 | 13.82 | 17.51 | Reversion + RERA emergence |
| 08-23 | 0.77 | 3 | 2 | 5 | 0.46 | 0.94 | 10.66 | 0 | 2 | 14.78 | 30.09 | Full release + RERA persists |
| 08-24 | 1.81 | 9 | 3 | 12 | 1.36 | 0.37 | 8.68 | 0 | 1 | 11.94 | 25.33 | Titrator experiments low + RERA resolves |
| 08-25 | 1.61 | 7 | 2 | 9 | 1.02 | 0.73 | 9.10 | 0 | 2 | 6.63 | 23.85 | Partial reinflation + CA=2 + deep collapse (WATCH ARTIFACT) |
| 08-26 | 1.33 | 7 | 1 | 8 | 1.03 | 0.72 | 9.42 | 0 | 1 | 17.98 | 31.27 | New operating point confirmed + architecture near-record + all channels quiet |
Interpretation. The cluster arc is not just released — the machine has settled at a new, lower, better operating point. AHI trended down across the last three nights (1.81 → 1.61 → 1.33), event burden is at or below the 30-day baseline on every channel, and pressure is stable at 7.05–7.09 (a real, sustained ~0.35 cmH₂O downshift from the pre-cluster 30-day baseline of 7.4). The 08-24 pressure-drop experiment (P_avg 6.64) is now understood in retrospect as the machine's overshoot — it correctly identified that the airway supported lower pressure, undershot by ~0.4 cmH₂O, and has hedged back up to 7.05–7.09 as the actual sustainable point. Event burden at the new lower pressure is meaningfully better than at the pre-cluster baseline (tonight's AHI 1.33 vs. 30-day 1.78; H=1 vs. mean 3.03), which is the strongest possible validation that the new operating point is superior, not just different. On the architecture side, tonight is a near-record composite (REM #3, awake #4, efficiency #2, deep top quintile, 11-night efficiency streak), and the 08-25 deep-sleep collapse is now diagnosable as a Watch measurement artifact rather than a physiological event. Both the CPAP and architecture stories are simultaneously excellent tonight — an unusual conjunction the summer's decoupling analysis has never seen at this magnitude.
| 08-25 report prediction for 08-26 | Actual 08-26 outcome | Verdict |
|---|---|---|
| Deep ≥ 12% = one-off (Watch artifact or single-night stressor) | Deep = 17.98% | CLEAN HIT — one-off confirmed, reading (b) validated |
| Deep < 10% for 2nd consecutive night = unmeasured stressor real | Deep = 17.98% | NEGATIVE — trigger not tripped |
| CA ≤ 1 = tonight's CA=2 was noise | CA = 1 | CLEAN HIT |
| CA ≥ 3 with pressure ≤ 7.5 = trading OA for CA | CA = 1, P_avg = 7.09 | NEGATIVE — trigger not tripped |
| P_avg 7.0–7.5 confirms new efficient operating point | P_avg = 7.09 | CLEAN HIT — center of band |
| P_avg > 7.6 = titrator still re-inflating | P_avg = 7.09 | NEGATIVE |
| P_avg < 6.8 = titrator over-corrected the correction | P_avg = 7.09 | NEGATIVE |
| OA count 3–8 confirms cluster fully done | OA = 7 | CLEAN HIT |
| OA count ≥ 10 for 2+ nights = OA question re-opened | OA = 7 | NEGATIVE |
| Bedtime 22:30 or earlier confirms improvement capacity | Bedtime = 22:53 | MISSED — reversed 7 min |
| Bedtime 22:45+ confirms current floor | Bedtime = 22:53 | CONFIRMED — floor holds at ~22:45–22:55 |
Net verdict: 5 clean hits, 4 negative-triggers-not-tripped, 1 miss (bedtime capacity did not materialize). The 08-25 report's dominant call — that the deep-sleep collapse was a Watch artifact rather than a physiological signal — has now been settled correctly on the artifact side. The pressure-baseline call — that the new operating point is 7.0–7.5 — is validated at night 2. The CA noise call is validated. The only miss is on bedtime capacity, which was a soft prediction anyway.
Buckets on nightly leak average:
the user correlations (n=52, tonight included):
| Pair | r | Reading |
|---|---|---|
| leak_avg vs. deep_pct | −0.19 | Weak negative — high leak slightly costs deep |
| leak_avg vs. awake_min | −0.07 | Essentially zero |
| leak_90 vs. deep_pct | −0.21 | Weak negative — parallels leak_avg |
| leak_90 vs. awake_min | −0.05 | Zero |
| ahi vs. deep_pct | +0.16 | Weak positive — decoupling holds |
| ahi vs. awake_min | +0.06 | Zero |
| ahi vs. rem_pct | 0.00 | Zero |
| p_avg vs. deep_pct | +0.10 | Weak positive |
| p_avg vs. awake_min | −0.13 | Weak negative |
| p_avg vs. rem_pct | −0.10 | Weak negative |
| tib_s vs. deep_pct | −0.18 | Weak negative — longer nights have slightly less deep by percentage |
Reading. Adding tonight (leak_avg = 0.72, deep = 17.98%, REM = 31.27%, awake = 4.52 min) to the mid-leak bucket pulls the bucket means slightly toward the good end: deep 15.16% → 15.28%, awake 18.54 → 17.96 min, REM 23.00% → 23.34%. Every aggregate correlation moved by ≤ 0.02, consistent with tonight being a strong outlier that the correlation structure absorbs without shifting the underlying decoupling thesis. The decoupling thesis holds firmly: no correlation exceeds |0.22|, and tonight's near-record architecture came on a mid-band CPAP night with no exceptional inputs on any channel. This is the mirror image of the 08-25 outlier — where 08-25 was near-record-bad architecture with median-band CPAP, tonight is near-record-good architecture with median-band CPAP. Two consecutive outliers of opposite sign, neither predicted by any CPAP channel — this is what decoupling looks like at its most demonstrated.
CA count distribution across 52 matched nights: CA=0 on 20 (38%), CA=1 on 19 (37%), CA=2 on 10 (19%), CA=3 on 2 (4%), CA=4 on 1 (2%). Tonight's CA=1 is at the 55th percentile — completely typical. The 08-25 CA=2 has been re-classified as a single-night blip. Historical low-pressure CA sub-pattern (5 of 7 CA≥2 nights at P_avg ≤ 7.6) remains valid as a background statistical observation but is not producing sustained CA elevation at the current 7.0–7.1 operating point. Off active watch for CA.
| Metric | Tonight (08-26) | 7-day avg (through 08-25) | 30-day avg (through 08-25) |
|---|---|---|---|
| AHI | 1.33 | 1.85 | 1.78 |
| OA count | 7 | 9.33 | 8.04 |
| OA per hour | 1.03 /h | 1.37 /h | 1.19 /h |
| H count | 1 | 2.83 | 3.04 |
| CA count | 1 | 0.50 | 0.93 |
| RERA count | 0 | 1.00 | 0.93 |
| Leak avg (L/min) | 0.72 | 0.61 | 0.82 |
| Leak 90th (L/min) | 0.0 | 1.20 | 2.53 |
| Pressure avg (cmH₂O) | 7.09 | 7.45 | 7.41 |
| Pressure 90th (cmH₂O) | 9.42 | 10.08 | 9.94 |
| Usage hours | 6.77 | 6.75 | 6.77 |
| Deep sleep % | 17.98% | 11.51% | 13.11% |
| REM sleep % | 31.27% | 23.87% | 23.46% |
| Awake time (min) | 4.52 | 10.29 | 16.82 |
| Efficiency | 98.87% | 97.34% | 95.95% |
| TIB (h) | 6.68 | 6.64 | 6.70 |
Verdict: tonight beats every trailing benchmark on the metrics that matter. AHI is 28% below 30-day baseline. OA is 13% below. H is 67% below. RERA is at floor. Leak_90 is at floor. Pressure is 4–5% below the 30-day baseline (as expected — new operating point). Deep is +37% vs. 30-day baseline; REM is +33% vs. baseline; awake is −73% vs. baseline; efficiency is 2.92 pts above baseline. No metric is meaningfully worse than baseline. This is the strongest single-night composite in the trailing 30-day window and probably in the full matched history.
| Week (Wed–Tue) | AHI avg | Leak avg | P_avg | P_90 avg | OA avg | Hours | n |
|---|---|---|---|---|---|---|---|
| 07-15 to 07-21 | 1.89 | 0.48 | 8.39 | 11.05 | 8.2 | 6.43 | 5 |
| 07-22 to 07-28 | 1.68 | 0.87 | 7.41 | 10.22 | 8.6 | 6.76 | 7 |
| 07-29 to 08-04 | 1.74 | 0.70 | 7.03 | 9.27 | 7.3 | 7.12 | 7 |
| 08-05 to 08-11 | 1.83 | 1.16 | 7.51 | 10.13 | 7.6 | 6.37 | 7 |
| 08-12 to 08-18 | 1.57 | 0.61 | 7.76 | 10.34 | 7.9 | 6.62 | 7 |
| 08-19 to 08-25 | 1.90 | 0.80 | 7.55 | 10.11 | 9.4 | 6.76 | 7 |
| 08-20 to 08-26 (trailing 7 through tonight) | 1.78 | 0.63 | 7.40 | 9.98 | 9.0 | 6.75 | 7 |
Weekly reading. The 08-19 → 08-25 close-out at OA 9.43/night was the highest full-week OA count on record, driven by the mid-week cluster peak. The trailing-7 through tonight (08-20 → 08-26) has already begun the correction: OA 9.0 vs. 9.43 (−4.6%), AHI 1.78 vs. 1.90 (−6.3%), Leak_avg 0.63 vs. 0.80 (−21%), P_avg 7.40 vs. 7.55 (−0.15). The next full Wed–Tue week (08-26 → 09-01) is on track to close well below the 9-OA/night threshold if the current pattern holds. Weekly-level cluster residue is dissipating on schedule and no new escalation trigger is present.
PRIMARY FINDING: STRONGEST COMPOSITE NIGHT IN THE MATCHED HISTORY. Tonight is at or near the top decile on every architecture axis (REM #3/52, awake #4/52, efficiency #2/52, deep #9/52) and at or near the bottom quintile on CPAP event burden (AHI #12/52, H #5/52, RERA #1/52, Leak_90 #1/52), and the new pressure operating point is confirmed at night 2 (P_avg 7.09, rank 12/52). This is an unprecedented conjunction — the summer's decoupling analysis has consistently shown architecture varying independently of CPAP metrics, but tonight both channels peaked simultaneously with no obvious causal link. If the user noticed anything special about last night — meal, mood, exercise, environment — it would be worth flagging as an experimental condition worth reproducing.
DEEP-SLEEP QUESTION: DECISIVELY RESOLVED. The 08-25 collapse to 6.63% was a one-night Watch stage-detection artifact, not a physiological signal. Tonight's 17.98% (+11.35 pts) is a fully symmetric bounce that cannot be explained by a rapid clearing of an unmeasured stressor (a stressor sufficient to suppress deep by 5.31 pts would take multiple nights to fully clear, not one). The 51-nights-of-decoupling thesis absorbs the outlier without any parameter shift, and reading (b) from the 08-25 report is now validated. Close the deep-sleep question unless deep drops below 10% again inside the next 5 nights.
PRESSURE CHANNEL: NEW OPERATING POINT CONFIRMED. P_avg 7.05 → 7.09 (+0.04) is a stationary state after the 08-24 → 08-25 partial reinflation. The 7.0–7.2 P_avg / 9.1–9.5 P_90 band is now the operating baseline, ~0.35 cmH₂O below the pre-cluster 30-day baseline. Event burden at this lower pressure is better, not worse, than at the higher baseline — AHI trended down over the same window (1.81 → 1.61 → 1.33) and H count is now near floor (rank 5/52). This is a strictly Pareto-improved operating point. Expected to hold barring a fresh mask/leak or airway event.
EVENT CHANNELS: CLUSTER FULLY RETIRED. AHI 1.33 is 25% below 30-day baseline. OA 7 is at the low edge of the normal 7–11 band. H 1 is near floor. RERA 0 is at floor. CA 1 is at median. All four obstructive/event channels are quiet. No active watch remains from the 08-18 → 08-25 cluster.
LEAK CHANNEL: AT FLOOR. Leak_avg 0.72 (near median), Leak_90 = 0.0 (at floor). Leak_max 16.8 is elevated in the single-event tail but not sustained (compare Leak_90 = 0 — the elevated max is a transient spike, not a leak pattern). Off active watch.
BEDTIME CHANNEL: FLOOR IDENTIFIED AT 22:45–22:55. The 4-night improvement arc peaked at 22:46 on 08-25 and reversed to 22:53 tonight. Current operating band is 22:45–22:55, 15–25 min past the 22:15–22:30 target. The "trajectory has capacity to hit 22:30" hypothesis is now weaker but not falsified — a single-night reversal within a 22:45–22:55 band is consistent with normal variance, not a directional signal that improvement is exhausted. If the user wants 22:30, a 15-min behavioral intervention (wind-down cue, screen shutoff, book at 22:15) remains the plausible closer. Not health-critical at the current 6.7-h TIB and 98.87% efficiency.
ARCHITECTURE CHANNEL: NEAR-RECORD HIGH. REM 31.27% (rank 3/52), deep 17.98% (rank 44/52), awake 4.52 min (rank 4/52), efficiency 98.87% (rank 2/52). The 11-night efficiency streak (≥ 95%) is now the longest on record. Whatever variance drove tonight up was not captured by any CPAP channel — a repeat of the decoupling thesis at the opposite pole from 08-25. The next 3–5 nights will show whether tonight is a genuine local peak or the start of a sustained high-quality streak.
What to watch on 08-27.