Sources: health-report-2026-07-18-to-2026-07-19 2.json (Apple Watch sleep stages, exported 22:47 on 07-19) + OSCAR_Greg_by_Day_2026-07-18_2026-07-19.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (23:54 on 07-18).
Awake time = 36.6 min — worst in the full matched dataset. The prior 4-night average (Jul 14–17) was 8.8 min. Tonight's 36.6 min is a 4.2× spike, producing the lowest sleep efficiency recorded (90.3% vs. prior range of 96.3–98.8%). This is the most significant outlier in the dataset to date.
Leak avg rising: 0.13 → 0.27 → 0.54 L/min over Jul 16–17–18. Tonight's 0.54 is 64% above the prior 7-day average (0.33 L/min). A leak max of 25.2 L/min indicates at least one sharp mask-dislodgment or mouth-opening event overnight. High-leak nights in the historical dataset (Jun 24–27, Jul 01) coincided with elevated AHI; tonight that pattern does not repeat (AHI=2.35), but the leak spike does appear temporally linked to the awake-time spike.
RERA = 0 for second consecutive night. The Jul 12–16 streak of RERA ≥ 3 has now been broken for two nights running (Jul 17: 0, Jul 18: 0). However, the continuation of high awake time on a RERA=0 night argues that RERA was not the sole arousal driver — either the leak event(s) caused awakening, or there is a third variable (positional, thermal, or circadian) operating independently.
Deep sleep continues to decline from recent peak. Jul 17 deep was 16.4%; tonight drops to 14.7%. The staged-nights average (Jul 14–17) was 17.1%. Two-night directional decline, but not yet outside the prior range.
| Metric | Value |
|---|---|
| Bedtime (session start) | 23:54 |
| Time in bed | 6.29 h (377.3 min) |
| Total sleep duration | 5.68 h (340.7 min) |
| Sleep efficiency (TST/TIB) | 90.3% — lowest recorded |
| Awake time | 36.6 min — highest recorded |
| Core sleep | 3.53 h (211.8 min) — 62.1% |
| Deep sleep | 0.84 h (50.2 min) — 14.7% |
| REM sleep | 1.31 h (78.8 min) — 23.1% |
| CPAP usage | 6.38 h |
| AHI | 2.35 |
| RDI | 2.35 |
| OA / H / CA / UA / RERA | 10 / 4 / 1 / 0 / 0 |
| Pressure avg | 8.78 cmH₂O |
| Pressure 90th %ile | 10.98 cmH₂O |
| Leak avg | 0.54 L/min |
| Leak max | 25.2 L/min |
| Leak 90th %ile | 0 L/min |
| Date | AHI | RERA | Leak avg | Deep % | REM % | Awake (min) | Efficiency |
|---|---|---|---|---|---|---|---|
| 07-11 | 4.66 | 0 | 0.15 | 14.3% | — | 24.6 | — |
| 07-12 | 2.70 | 3 | 0.68 | 12.6% | — | 16.1 | — |
| 07-13 | 1.44 | 3 | 0.53 | 11.4% | — | 16.1 | — |
| 07-14 | 3.18 | 3 | 0.30 | 19.3% | 17.8% | 10.5 | 96.8% |
| 07-15 | 3.01 | 4 | 0.25 | 16.1% | 25.9% | 5.0 | 98.8% |
| 07-16 | 1.50 | 3 | 0.13 | 16.6% | 20.6% | 12.5 | 96.3% |
| 07-17 | 2.78 | 0 | 0.27 | 16.4% | 30.0% | 7.0 | 98.2% |
| 07-18 | 2.35 | 0 | 0.54 | 14.7% | 23.1% | 36.6 | 90.3% |
Splitting by leak threshold (≤ 0.30 vs. > 0.30 L/min):
The directional signal is consistent: higher-leak nights average 11.2 more awake minutes (+94%) and 3.1% less deep sleep. July 18 is the strongest data point in this pattern, with leak = 0.54 and awake = 36.6 min.
Important caveat on causation: the leak max of 25.2 L/min on Jul 18 suggests an acute event (e.g., mask shifting, brief mouth opening) rather than sustained chronic leak (the 90th-percentile figure is 0, meaning the vast majority of the night was leak-free). The awake time may have preceded the leak event if position shifts caused both, or the awakening may have been caused by the leak. No intra-night timing data is available to distinguish.
| Metric | Tonight (07-18) | 7-day avg (07-11–07-17) | 30-day avg (06-19–07-17) |
|---|---|---|---|
| AHI | 2.35 | 2.75 | 2.78 |
| RERA | 0 | 2.29 | 1.26 |
| Leak avg (L/min) | 0.54 | 0.33 | 1.04* |
| Pressure avg (cmH₂O) | 8.78 | 8.90 | 7.66† |
| Usage hours | 6.38 | 6.48 | 6.45 |
| Deep sleep % | 14.7% | ~15.1%‡ | ~15%‡ |
| REM sleep % | 23.1% | ~23%‡ | ~23%‡ |
| Awake time (min) | 36.6 | ~14.1‡ | ~16‡ |
| Sleep efficiency | 90.3% | ~97.5%‡ | ~97%‡ |
*30-day leak elevated by Jun 24–27 and Jul 01 high-leak nights.
†30-day pressure depressed by zero-pressure baseline nights (Jun 15–22).
‡Sleep-stage trailing averages approximate from nights with stage data only.
health-report-2026-07-18-to-2026-07-19.json (exported 09:54 Jul 18) contains a session starting at 00:04 Jul 18, likely capturing the tail of the Jul 17 sleep. The "2" file (exported 22:47 Jul 19, session starting 23:54 Jul 18, TIB=6.29h) is the better match for CPAP Jul 18 (6.38h, delta 0.09h) and is used for this report. The 00:04 session is not analyzed here; it may be a duplicate measurement artifact.