🔴 Pressure_Max 14.60 cmH2O — new 26-night window RECORD (z = +3.08 vs 30d). Previous window high was 2026-10-05's 13.98; tonight exceeds it by +4.4%. The 30d mean Pressure_Max is 12.00 (sd 0.84), so 14.60 is a true outlier — the auto-titrator pushed harder on at least one segment than it has in the entire 26-night window. Pressure_Avg 7.78 (z = +0.53, flat) and Pressure_90 10.18 (z = +0.04, flat) are both unremarkable, so the profile is flat avg/90, extreme max — the machine saw a brief event demanding a sharp 14.60 cmH2O push, not a sustained elevation. This is almost certainly the mechanism behind the simultaneous Leak_90 spike (flag 3): a high-pressure segment overcame the mask seal.
🔴 Deep sleep 8.59% — rank 2/26 in the window; the 3-night soft stretch flagged yesterday now extends to a 4-night stretch with a NEW low. Only 2026-10-04 (7.88%) in the 26-night window has less deep sleep than tonight. z = -1.32 vs 30d; -36.3% vs 30d (13.48); -31.5% vs 7d (12.54). The 4-night slice 10-08 → 10-11 averages deep 10.57% vs the 26-night mean of 13.35 — ~2.8 points below baseline, worse than yesterday's 3-night 11.23. Deep minutes 34.62 is the lowest in the 26-night window after 10-04's 32.65. The one note of balance: tonight's REM at 26.65% is well above baseline (rank 22/26) — so this is a selective deep-sleep deficit, not an architecture collapse. REM compensated; N3 did not.
🔴 Leak_90 9.60 L/min — rank 23/26, z = +1.81, +160.9% vs 30d, +124.0% vs 7d. Doubles yesterday's 4.80. Ties the window-highs set on 09-26 (9.6); only 09-23 (12.0) and 10-03 (10.8) are higher in the 26-night window. Yesterday's "Leak_Max streak BROKEN" flag is partially reversed — Leak_Max 24.00 tonight is back above the 22.8 cluster-threshold (ending the one-night break at 18.0 on 10-10), but it is not aggressively above (rank 14/26, z = +0.21). The signal has moved from Leak_Max to Leak_90: brief peaks moderate, sustained-moderate leak elevated. Leak_Avg 1.89 (rank 22/26, z = +1.05, +66.8% vs 30d) corroborates. 7d Leak_90 jumps from 4.29 (yesterday) to 4.97 — not a trend yet, but reverses yesterday's cooling.
🟡 Hypopneas drive AHI despite OA rank 2/26 — a different event profile than recent nights. OA 4 (z = -1.24, rank 2/26 — second-lowest OA count in the window, tied with 2026-10-01's 4). CA 0 (rank 1/26). Yet H 6 (z = +1.07, rank 19/26, +81.8% vs 30d) — in the 26-night window only 10-05 (11), 09-20 (7), 09-22 (7) have more. AHI 1.42 stays clean (6th consecutive sub-2.0 AHI: 1.85 → 1.24 → 0.85 → 1.49 → 1.26 → 1.42), but tonight's events are near-exclusively hypopneas rather than apneas — airway partially collapsing without fully occluding. The AHI-RDI gap is only 0.14 (RDI 1.56) — same as yesterday's 0.14 — so RERAs are quiet, further isolating the signal to H events specifically.
🟢 Yesterday's FL spike reverted overnight — the "watch next 2-3 nights" verdict is BASELINE. All sustained-FL metrics are back near 30d norms: FL_P95 0.09 (z = -0.28, rank 8/26, -5.9% vs 30d), Min_Above_010 15.6 min (z = -0.30, -9.1% vs 30d), Min_Above_020 4.1 min (z = -0.08, -4.1% vs 30d), FL_Avg 0.0150 (z = -0.22). Yesterday's 26-night-peak-on-three-metrics simultaneously was a single-night event, not the start of a sustained trend. One qualifier: FL_Max 0.56 remains elevated (z = +0.76, rank 21/26, +16.5% vs 30d) — brief single-sample peaks continued even as the sustained tail dropped. The simplest story that ties it to flag 1: yesterday's sustained FL triggered the auto-titrator to push harder tonight, hitting P_Max 14.60 to clear brief spikes while leaving the sustained-FL distribution near normal.
🟡 REM 26.65% — rank 22/26 (z = +0.73), full rebound from yesterday's 18.96%. +11.1% vs 30d; +11.7% vs 7d. REM minutes 107.38 vs yesterday's 78.78 (+36.3% night-over-night). Tonight is REM-rich / N3-poor — the inverse of yesterday (which was also somewhat REM-poor). The 4-night deep-soft stretch has NOT been accompanied by a 4-night REM-soft stretch: 10-08 REM 24.30%, 10-09 REM 20.09%, 10-10 REM 18.96%, 10-11 REM 26.65%. Deep is the persistent concern; REM is choppy but trending back to baseline.
🟢 Usage 7.03h — rank 22/26 (z = +1.36); 3 of the last 4 nights ≥ 7.03h usage. Pattern (10-08 → 10-11): 7.08 → 6.03 → 7.17 → 7.03 — the dip on 10-09 looks like a one-off. 7d usage 6.63h vs 30d 6.43h (+3.0%); the extended-usage pattern since mid-September now looks durable. TIB 412.46 min matches usage closely; sleep efficiency 97.69 (z = +0.15).
🟡 Sleep start 23:24:08 — 46 minutes LATER than yesterday's 22:38:55, back to the typical late-10/early-11 PM pattern. This did not compress TIB (412.46 min / 6h 52.5m, rank 22/26, z = +1.17) — the user stayed in bed as long. The lateness alone does not explain the deep-sleep drop; TIB is still rank 22/26, awake 9.53 min (z = -0.03) is at baseline, and efficiency 97.69% is normal. Yesterday I noted "last night's short/late = REM drop explanation" — tonight's deep drop is similarly unexplained by sleep-window mechanics.
| Metric | Value |
|---|---|
| Sleep start | 2026-10-11T23:24:08-05:00 |
| Time in bed | 412.46 min (6h 52.5m) |
| Total sleep | 402.93 min (6h 42.9m) |
| Sleep efficiency | 97.69% |
| Awake time | 9.53 min |
| Core (N1+N2) sleep | 260.93 min / 64.76% |
| Deep (N3) sleep | 34.62 min / 8.59% |
| REM | 107.38 min / 26.65% |
| CPAP usage hours | 7.03 |
| Session count | 1 |
| AHI | 1.42 |
| RDI | 1.56 |
| Obstructive apneas (OA) | 4 |
| Hypopneas (H) | 6 |
| Central apneas (CA) | 0 |
| Unclassified apneas (UA) | 0 |
| RERA | 1 |
| Pressure_Avg | 7.78 |
| Pressure_Min | 4.00 |
| Pressure_Max | 14.60 |
| Pressure_90th | 10.18 |
| Leak_Avg | 1.89 |
| Leak_Max | 24.00 |
| Leak_90th | 9.60 |
| FL_Avg | 0.0150 |
| FL_P95 | 0.09 |
| FL_Max | 0.56 |
| Min_Above_005 (min) | 39.4 |
| Min_Above_010 (min) | 15.6 |
| Min_Above_020 (min) | 4.1 |
| Frac_NonZero (FL samples) | 26.34% |
| Metric | Tonight | 7d avg | 30d avg | Δ vs 7d | Δ vs 30d |
|---|---|---|---|---|---|
| AHI | 1.42 | 1.87 | 2.02 | -23.9% | -29.8% |
| RDI | 1.56 | 2.22 | 2.19 | -29.9% | -28.8% |
| OA count | 4.00 | 8.43 | 9.03 | -52.5% | -55.7% |
| H count | 6.00 | 3.43 | 3.30 | +75.0% | +81.8% |
| CA count | 0.00 | 0.14 | 0.53 | -100.0% | -100.0% |
| RERA count | 1.00 | 2.29 | 1.07 | -56.2% | -6.2% |
| Pressure_Avg | 7.78 | 7.48 | 7.55 | +4.0% | +3.1% |
| Pressure_90 | 10.18 | 10.45 | 10.15 | -2.6% | +0.3% |
| Leak_Avg | 1.89 | 1.46 | 1.13 | +29.3% | +66.8% |
| Leak_Max | 24.00 | 26.91 | 22.24 | -10.8% | +7.9% |
| Leak_90th | 9.60 | 4.29 | 3.68 | +124.0% | +160.9% |
| Usage hours | 7.03 | 6.59 | 6.43 | +6.6% | +9.4% |
| FL_P95 | 0.090 | 0.104 | 0.096 | -13.7% | -5.9% |
| FL_Avg | 0.0150 | 0.0176 | 0.0158 | -14.7% | -5.0% |
| FL_Max | 0.560 | 0.524 | 0.481 | +6.8% | +16.5% |
| Min_Above_005 | 39.4 | 43.0 | 38.7 | -8.3% | +1.8% |
| Min_Above_010 | 15.6 | 20.5 | 17.2 | -23.7% | -9.1% |
| Min_Above_020 | 4.1 | 6.4 | 4.3 | -36.1% | -4.1% |
| Frac_NonZero | 0.263 | 0.254 | 0.249 | +3.8% | +5.7% |
| Deep % | 8.59 | 12.54 | 13.48 | -31.5% | -36.3% |
| REM % | 26.65 | 23.86 | 23.98 | +11.7% | +11.1% |
| Core % | 64.76 | — | — | — | — |
| Awake min | 9.53 | 6.16 | 9.68 | +54.7% | -1.6% |
| TIB min | 412.46 | 390.03 | 380.54 | +5.8% | +8.4% |
| Total sleep min | 402.93 | 383.86 | 370.85 | +5.0% | +8.7% |
| Sleep efficiency | 97.69 | 98.39 | 97.45 | -0.7% | +0.2% |
| Metric | Recent 7d | Prior 7d | Δ | Δ % |
|---|---|---|---|---|
| AHI | 1.901 | 1.461 | +0.440 | +30.1% |
| RDI | 2.194 | 1.661 | +0.533 | +32.1% |
| OA | 8.143 | 6.857 | +1.286 | +18.8% |
| H | 4.000 | 2.143 | +1.857 | +86.7% |
| CA | 0.143 | 0.286 | -0.143 | -50.0% |
| RERA | 1.857 | 1.286 | +0.571 | +44.4% |
| Pressure_Avg | 7.463 | 7.726 | -0.263 | -3.4% |
| Pressure_90 | 10.454 | 10.129 | +0.326 | +3.2% |
| Leak_Avg | 1.459 | 1.364 | +0.094 | +6.9% |
| Leak_90 | 4.971 | 4.800 | +0.171 | +3.6% |
| Leak_Max | 26.229 | 24.343 | +1.886 | +7.7% |
| Usage hours | 6.630 | 6.461 | +0.169 | +2.6% |
| Deep % (Apple) | 12.646 | 13.789 | -1.143 | -8.3% |
| REM % (Apple) | 23.682 | 24.366 | -0.684 | -2.8% |
| Awake min (Apple) | 7.312 | 12.903 | -5.591 | -43.3% |
| Sleep min (Apple) | 384.078 | 369.742 | +14.336 | +3.9% |
| TIB min (Apple) | 391.390 | 382.645 | +8.745 | +2.3% |
| Sleep efficiency | 98.12 | 96.57 | +1.55 | +1.6% |
| FL_P95 | 0.099 | 0.093 | +0.006 | +6.2% |
| FL_Avg | 0.0163 | 0.0161 | +0.0002 | +1.2% |
| Min_Above_010 | 19.014 | 16.214 | +2.800 | +17.3% |
| Min_Above_020 | 5.986 | 3.229 | +2.757 | +85.4% |
| Date | AHI | OA | CA | RERA | PAvg | P90 | LAvg | LMax | L90 | FL_P95 | MinAbv010 | MinAbv020 | Deep% | REM% | Awake |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2026-10-02 | 2.04 | 9 | 1 | 2 | 8.50 | 10.32 | 1.68 | 25.2 | 8.4 | 0.09 | 15.9 | 2.8 | 18.86 | 34.36 | 8.53 |
| 2026-10-03 | 2.33 | 7 | 1 | 2 | 7.21 | 9.30 | 2.77 | 22.8 | 10.8 | 0.08 | 11.3 | 2.5 | 9.63 | 25.46 | 10.54 |
| 2026-10-04 | 1.18 | 6 | 0 | 4 | 7.90 | 10.18 | 1.91 | 28.8 | 4.8 | 0.13 | 25.7 | 7.1 | 7.88 | 27.87 | 1.51 |
| 2026-10-05 | 5.20 | 21 | 0 | 4 | 7.93 | 12.36 | 1.76 | 22.8 | 7.2 | 0.09 | 15.7 | 8.2 | 15.92 | 25.21 | 8.03 |
| 2026-10-06 | 1.85 | 8 | 0 | 1 | 7.19 | 10.24 | 1.04 | 30.0 | 1.2 | 0.08 | 15.1 | 4.6 | 12.12 | 27.14 | 4.01 |
| 2026-10-07 | 1.24 | 7 | 0 | 0 | 7.69 | 10.72 | 1.50 | 34.8 | 3.6 | 0.12 | 23.3 | 4.6 | 18.21 | 23.43 | 6.02 |
| 2026-10-08 | 0.85 | 6 | 0 | 1 | 6.89 | 9.20 | 1.58 | 30.0 | 3.6 | 0.09 | 17.4 | 3.9 | 9.40 | 24.30 | 9.03 |
| 2026-10-09 | 1.49 | 6 | 1 | 5 | 7.38 | 9.88 | 1.34 | 24.0 | 4.8 | 0.08 | 14.0 | 6.7 | 14.02 | 20.09 | 10.54 |
| 2026-10-10 | 1.26 | 5 | 0 | 1 | 7.38 | 10.60 | 1.10 | 18.0 | 4.8 | 0.14 | 32.0 | 9.8 | 10.27 | 18.96 | 4.01 |
| 2026-10-11 | 1.42 | 4 | 0 | 1 | 7.78 | 10.18 | 1.89 | 24.0 | 9.60 | 0.09 | 15.6 | 4.1 | 8.59 | 26.65 | 9.53 |
Correlations computed across the matched set after excluding pre-2026-06-22 low-confidence CPAP nights.
| Pair | the user r | Interpretation |
|---|---|---|
| AHI vs deep_pct | -0.041 | negligible |
| AHI vs awake_min | -0.144 | weak negative |
| OA vs awake_min | -0.036 | negligible |
| CA vs deep_pct | -0.021 | negligible |
| Leak_Avg vs deep_pct | -0.080 | negligible |
| Leak_90 vs deep_pct | -0.063 | negligible |
| Leak_Max vs awake_min | -0.224 | weak negative |
| FL_P95 vs awake_min | -0.092 | negligible |
| Min_Above_010 vs deep_pct | -0.176 | weak negative |
Bucket comparison (Leak_90 median split, median = 2.40): - Low-leak nights (n=52): mean deep% 14.34, awake 14.25 min - High-leak nights (n=35): mean deep% 13.32, awake 10.16 min - Δ (high – low): deep% -1.02 pp, awake -4.09 min
Tonight's Leak_90 of 9.60 lands deep in the high-leak bucket (4× the median split, 2nd-highest high-leak-bucket Leak_90 after 2026-09-23's 12.0 and 10-03's 10.8); tonight's deep% of 8.59 is 4.73 points below the high-leak bucket mean (13.32) — a worse single-point mismatch than yesterday. The matched-night population-level correlations continue to show that leak alone explains only a small fraction of deep-sleep variance (|r| ≤ 0.08 for Leak_Avg and Leak_90 vs deep_pct). As yesterday noted, Min_Above_010 vs deep_pct r = -0.176 is the strongest signal in the matrix — but tonight Min_Above_010 is near-baseline (15.6, z = -0.30) while deep% hit a window-near-floor, so that correlation does not explain tonight either. The matched-night confident count is now 87 (two more than yesterday — one new confident night was added by including 10-10 and 10-11; the 10-10 report quoted 85).