Sources: health-report-2026-07-10-to-2026-07-11.json (Apple Watch sleep stages, exported 08:58 on 07-11) + OSCAR_Greg_by_Day_2026-06-15_2026-07-10.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (22:44 on 07-10).
Leak resolved. After the sharp spike on 07-09 (Leak_Avg 1.57 L/min, 95th %ile 6.0), last night's leak dropped to 0.18 L/min (95th %ile 0.0) — back in line with the 07-02 to 07-08 baseline (avg 0.26 L/min). The 07-09 event appears to have been a one-night mask-seal/positioning issue.
Deep sleep suppressed despite low leak — event burden the likely driver. Deep sleep was 15.2% last night, well below the 7-day average of 21.1% and below the low-leak-night average of 19.9% (n=14 matched nights). With leak ruled out, the most plausible driver is the elevated event burden: AHI 2.98 (above 7-day avg 2.37), H=6 hypopneas (highest single-night count in the dataset), and RERA=2. Pressure reached a 95th %ile of 11.42 cmH₂O — the highest in the dataset (prior max 11.10 on 07-07), suggesting the device was working hard to resolve flow limitation. The updated correlation across all 20 matched nights: AHI vs. awake-minutes r ≈ +0.40; awake time last night was 24.1 min vs. a 7-day average of 15.1 min.
AHI plateau now 3 weeks old. Week-over-week AHI: 06-15–06-22 → 0.87, 06-23–06-29 → 3.41, 06-30–07-06 → 2.80, 07-07–07-10 → 2.81. AHI has not returned to the early-window baseline in any week since the late-June jump — now running 3–4x above baseline for three consecutive weeks. Worth discussing with your care provider if it continues.
Longest usage session in the dataset. CPAP usage: 7.38 hours (prior max: 7.05 on 07-08).
| Metric | Value |
|---|---|
| Bedtime (session start) | 22:44 |
| Time in bed | 7.24 h (7h 15m) |
| Total sleep duration | 6.84 h (6h 50m) |
| Sleep efficiency (TST/TIB) | 94.5% |
| Awake time | 24.1 min |
| Core sleep | 4.03 h — 58.9% |
| Deep sleep | 1.04 h — 15.2% |
| REM sleep | 1.77 h — 25.9% |
| CPAP usage hours | 7.38 h (dataset high) |
| AHI | 2.98 events/hr |
| RDI | 3.25 events/hr |
| Obstructive apneas (OA) | 16 |
| Hypopneas (H) | 6 (dataset high) |
| Central apneas (CA) | 0 |
| Unclassified/other (UA) | 0 |
| RERA | 2 |
| Pressure avg / 95th | 8.73 / 11.42 cmH₂O (95th: dataset high) |
| Leak avg / 95th | 0.18 / 0.00 L/min |
Across all 20 matched nights (all nights with both CPAP record and full Watch sleep-stage data):
Note: adding tonight's data slightly updated the low-leak deep-sleep average downward (was 20.25% over 13 nights, now 19.89% over 14), reflecting that last night is a mild outlier in the low-leak group.
Last night vs. trailing 7-day averages (07-03 to 07-09):
| Metric | Last night (07-10) | 7-day avg | Δ |
|---|---|---|---|
| AHI | 2.98 | 2.37 | +0.61 |
| Leak avg | 0.18 | 0.54 | −0.36 (normalized) |
| Pressure avg | 8.73 | 8.27 | +0.46 |
| Pressure 95th | 11.42 | 10.67 | +0.75 (new dataset high) |
| Usage hours | 7.38 | 6.45 | +0.93 |
| Deep sleep % | 15.2% | 21.1% | −5.9 pts |
| REM sleep % | 25.9% | 24.5% | +1.4 pts |
| Awake min | 24.1 | 15.1 | +9.0 min |
| TIB | 7.24 h | 6.44 h | +0.80 h |
| TST | 6.84 h | 6.19 h | +0.65 h |
Last night vs. trailing 30-day averages (06-15 to 07-09):
| Metric | Last night | 30-day avg | Δ |
|---|---|---|---|
| AHI | 2.98 | 2.41 | +0.57 |
| Leak avg | 0.18 | 1.07 | −0.89 (well below avg) |
| Usage hours | 7.38 | 6.28 | +1.10 |
| Deep sleep % | 15.2% | 17.5% | −2.3 pts |
| REM sleep % | 25.9% | 22.7% | +3.2 pts |
| Awake min | 24.1 | 16.4 | +7.7 min |
| TST | 6.84 h | 6.04 h | +0.80 h |
Net read: despite the longest CPAP session on record and a clean leak number, deep sleep came in below average and awake time was elevated — driven by elevated hypopneas (H=6) and the highest 95th-percentile pressure in the dataset. The 3-week AHI plateau and the pressure escalation are the most meaningful signals to monitor going forward.
health-report-2026-07-10-to-2026-07-11.json (single-night export, 22:44 on 07-10 to 08:58 on 07-11).