Sources: health-report-2026-07-25-to-2026-07-26.json (Apple Watch sleep stages, exported 15:13 on 07-26 — this export had not yet synced when this report was first generated on 07-26 morning; regenerated same day once it landed) + OSCAR_Greg_by_Day_2026-07-25.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (23:11 on 07-25).
The predicted leak-arousal link from this morning's report is confirmed: tonight's leak spike coincided with the 4th-lowest deep sleep % and 3rd-highest awake time of all 34 matched nights in the dataset. Deep sleep fell to 9.9% (only 07-20 at 8.3%, 06-26 at 8.7%, and 06-24 at 9.7% were lower) and awake time rose to 30.6 min (only 07-18 at 36.6 and 06-25 at 31.6 were higher). This lines up with — and strengthens — the dataset's existing directional pattern: high-leak nights (>0.30 L/min) now average 15.2% deep / 19.3 min awake vs. 17.4% deep / 13.9 min awake on low-leak nights (n=34, leak avg 1.69 tonight vs. bucket means).
Leak avg 1.69 L/min was the highest single-night reading since 07-01 (6.22) — 25 nights ago — and the 6th-highest of all 40 tracked CPAP nights. Leak 90th %ile hit 7.2 L/min vs. a 0.0–2.4 range across the entire prior three weeks. Leak max was only 18.0 L/min — lower than 5 of the prior 8 nights — meaning this reads as sustained elevated leak rather than one brief dislodgment, a different signature than recent nights' pattern of low leak average with occasional high leak max.
AHI and OA partially reverted from 07-24's spike but did not fully normalize: AHI 1.72 (down from 2.34) and OA 10 (down from 15), both still above the 07-20–07-23 baseline (AHI 0.88–1.59, OA 2–7). Worth watching whether 07-26 completes the reversion.
Sleep efficiency (92.5%) and awake time (30.6 min) are both notably worse than the trailing 7-day averages (efficiency ~95%, awake 20.1 min excluding tonight) — consistent with a rough night overall, not just a leak-only anomaly. REM was elevated (32.9%, highest of the 8-night window) which sometimes co-occurs with fragmented sleep in this dataset, though the correlation for REM specifically hasn't been tested.
| Metric | Value |
|---|---|
| Bedtime (session start) | 23:11 |
| Time in bed | 6.84 h (410.4 min) |
| Total sleep duration | 6.33 h (379.8 min) |
| Sleep efficiency (TST/TIB) | 92.5% (below 7-day avg) |
| Awake time | 30.6 min ↑ (3rd-highest in dataset) |
| Core sleep | 3.62 h (234.0 min) — 57.2% |
| Deep sleep | 0.63 h (37.6 min) — 9.9% ↓ (4th-lowest in dataset) |
| REM sleep | 2.08 h (124.9 min) — 32.9% |
| CPAP usage | 6.98 h |
| AHI | 1.72 (↓ from 2.34, but ↑ vs. 07-20–23 range) |
| RDI | 1.72 |
| OA / H / CA / UA / RERA | 10 / 1 / 1 / 0 / 0 |
| Pressure avg | 7.92 cmH₂O |
| Pressure 90th %ile | 10.36 cmH₂O |
| Leak avg | 1.69 L/min ↑↑ (highest since 07-01) |
| Leak max | 18.0 L/min |
| Leak 90th %ile | 7.2 L/min ↑↑ (prior 3-week range: 0.0–2.4) |
| Date | AHI | OA | RERA | Leak avg | Leak 90th | Deep % | Awake (min) | Efficiency |
|---|---|---|---|---|---|---|---|---|
| 07-18 | 2.35 | 10 | 0 | 0.54 | 0.0 | 14.7% | 36.6 | 90.3% |
| 07-19 | 1.29 | 6 | 1 | 0.92 | 0.0 | 10.1% | 14.0 | 96.2% |
| 07-20 | 0.94 | 2 | 0 | 0.03 | 0.0 | 8.3% | 21.1 | 94.4% |
| 07-21 | 0.88 | 4 | 1 | 0.85 | 1.2 | 28.4% | 27.1 | 93.3% |
| 07-22 | 1.59 | 7 | 1 | 1.08 | 2.4 | 12.0% | 10.0 | 97.5% |
| 07-23 | 1.29 | 7 | 0 | 0.48 | 2.4 | 25.5% | 13.0 | 96.8% |
| 07-24 | 2.34 | 15 | 0 | 0.48 | 0.0 | 16.7% | 24.6 | 93.9% |
| 07-25 | 1.72 | 10 | 0 | 1.69 | 7.2 | 9.9% | 30.6 | 92.5% |
Split at leak ≤ 0.30 vs. > 0.30 L/min:
the user r (leak avg vs. deep %): −0.30 (n=34, was −0.29). the user r (leak avg vs. awake min): +0.12 (was +0.09). the user r (AHI vs. awake min): +0.18 (was +0.22). the user r (AHI vs. deep %): +0.05 (was +0.03).
Interpretation: Adding tonight strengthens the leak→deep-sleep relationship slightly (−0.29 → −0.30) and the leak→awake relationship modestly (+0.09 → +0.12), since 07-25 is a high-leak, low-deep, high-awake data point right in line with the existing direction. The AHI-based correlations weakened slightly, since tonight's AHI (1.72) was only moderate despite the rough sleep architecture — reinforcing that leak, not AHI, is the better predictor of tonight's disruption. Still, |r| values remain in the 0.05–0.30 range — real but weak relationships, not strong predictors on any single night.
| Metric | Tonight (07-25) | 7-day avg (07-19–07-25) | Prior 7-day avg (07-12–07-18) | 30-day avg (06-26–07-25) |
|---|---|---|---|---|
| AHI | 1.72 | 1.44 | 2.42 | 2.54 |
| Leak avg (L/min) | 1.69 | 0.79 | 0.39 | 0.89 |
| Pressure avg (cmH₂O) | 7.92 | 7.65 | 8.79 | 8.44 |
| Usage hours | 6.98 | 6.73 | 6.44 | 6.59 |
| Deep sleep % | 9.9% | 15.9% | — | 16.4% |
| REM sleep % | 32.9% | 25.3% | — | 24.1% |
| Awake time (min) | 30.6 | 20.1 | — | 16.7% |
Tonight's deep sleep (9.9%) is nearly 6 points below its own trailing 7-day average (15.9%) and 6.5 points below the 30-day average (16.4%) — a meaningful single-night dip. Awake time (30.6 min) is roughly 1.5x the 7-day average and nearly double the 30-day average. AHI and pressure remain unremarkable relative to trailing figures; the disruption tonight is concentrated in leak, deep sleep, and awake time specifically.
| Week | AHI avg | Leak avg | n |
|---|---|---|---|
| 06-22 to 06-28 | 2.84 | 1.98 | 7 |
| 06-29 to 07-05 | 3.24 | 1.41 | 7 |
| 07-06 to 07-12 | 3.00 | 0.43 | 7 |
| 07-13 to 07-19 | 2.22 | 0.42 | 7 |
| 07-18 to 07-24 (prior trailing 7d) | 1.53 | 0.63 | 7 |
| 07-19 to 07-25 (current trailing 7d) | 1.44 | 0.79 | 7 |
The multi-week AHI downtrend continues (new trailing-7-day low of 1.44). Trailing leak ticked up (0.63 → 0.79) almost entirely because of tonight — the other six nights in the window ranged 0.03–1.08.
Primary watch item: does the leak spike repeat on 07-26? Given tonight's clear architecture disruption (lowest deep sleep in over three weeks, 3rd-highest awake time on record), a repeat would be worth a mask cushion/fit check rather than waiting for a 3-night streak — the single-night severity here is already dataset-notable.
AHI/OA: Still not fully back to the 07-20–23 baseline. Watch 07-26 for full reversion vs. a sustained mild elevation.
Sleep architecture: Tonight is now the clearest single data point yet supporting the leak→arousal/deep-sleep relationship in this dataset. Continue tracking to see if the correlation strengthens further with more high-leak nights.
health-report-2026-07-25-to-2026-07-26.json (exported 15:13 on 07-26) has since landed and is incorporated here.