Sources: health-report-2026-07-15-to-2026-07-16.json (Apple Watch sleep stages, exported 11:09 on 07-16) + OSCAR_Greg_by_Day_2026-07-15.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (22:25 on 07-15).
RERA=4 with clean sleep: the streak escalates to 4 nights (July 12–15) but sleep architecture is the best in weeks. RERA has now hit ≥3 for four consecutive nights and tonight reached 4 — the dataset high and tied only by the isolated July 3 night. Yet awake time tonight is 5.0 min (lowest in this entire dataset), deep sleep is 16.1% (near the 30-day baseline of ~17.3%), and REM is 25.9% (above the ~23% baseline). This is a meaningful decoupling: the upper-airway flow-limitation signal is escalating while the sleep disruption signal has largely resolved. The RERA events appear not to be generating arousals sufficient to fragment consolidated sleep.
REM rebound to 25.9% — the strongest REM in the recent dataset. The July 14 report flagged REM compression to 17.8% as notable. Tonight REM surged to 25.9%, 3 pp above the 30-day baseline. Combined with deep sleep returning to 16.1%, tonight represents a near-full architectural recovery after the July 9–13 deep-sleep suppression trough.
Awake time 5.0 min — the lowest in the dataset. The 30-day average awake time is ~17 min; the prior best was 10.5 min on July 14. Tonight's 5 minutes is a sharp positive outlier that occurred on the same night as the dataset-high RERA count, further confirming the RERA-disruption decoupling observed from July 14 onward.
Sustained RERA streak remains unexplained and warrants monitoring. Even with good sleep quality tonight, 4 consecutive nights of RERA ≥3 is structurally significant. The device is repeatedly classifying upper airway flow limitation during sleep. Whether or not these translate to disruption on any given night, the persistence points to an anatomical or positional factor — nasal passage, jaw/tongue position, pillow height — that has been consistently present for the past 4 nights.
| Metric | Value |
|---|---|
| Bedtime (session start) | 22:25 |
| Time in bed | 6.51 h (6h 30m) |
| Total sleep duration | 6.42 h (6h 25m) |
| Sleep efficiency (TST/TIB) | 98.7% |
| Awake time | 5.0 min (dataset low) |
| Core sleep | 3.72 h — 57.9% |
| Deep sleep | 1.04 h — 16.1% |
| REM sleep | 1.66 h — 25.9% (above baseline) |
| CPAP usage | 6.65 h |
| AHI | 3.01 |
| RDI | 3.61 |
| OA / H / CA / UA / RERA | 13 / 5 / 2 / 0 / 4 |
| Pressure avg | 8.77 cmH₂O |
| Pressure 95th %ile | 11.46 cmH₂O |
| Leak avg | 0.25 L/min |
| Leak 95th %ile | 0* (see caveats) |
The addition of tonight's data sharpens the RERA-disruption picture considerably:
| Date | AHI | RERA | Leak | Deep % | REM % | Awake (min) |
|---|---|---|---|---|---|---|
| 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 |
| 07-15 | 3.01 | 4 | 0.25 | 16.1% | 25.9% | 5.0 |
Emerging pattern — leak as RERA-to-disruption mediator: The July 12–13 nights (RERA=3, high disruption) had leak of 0.53–0.68 L/min. July 14–15 (RERA=3–4, low disruption) had leak of 0.25–0.30 L/min. The hypothesis is directionally consistent across these 4 nights: when RERA events co-occur with even moderate leak, the combined stimulus may be enough to produce arousal; with a clean seal, RERA events may resolve subclinically without fragmenting sleep. Sample size is small (4 nights), so this is observational — but it's a testable pattern going forward. Nights with RERA≥3 and leak<0.35 L/min: deep 16.1–19.3%, awake 5.0–10.5 min. Nights with RERA≥3 and leak≥0.50 L/min: deep 11.4–12.6%, awake 16.1 min.
AHI vs. awake time (full dataset, qualitative): Tonight continues the pattern where AHI alone is a weak predictor of awake time — AHI=3.01 with only 5.0 min awake, vs. July 9 (AHI=2.12, awake=25.6 min). Other factors (leak, RERA co-occurrence, sleep inertia) appear more important than raw AHI for disruption.
| Metric | Tonight (07-15) | 7-day avg (07-09–07-15) | 30-day avg (06-15–07-15) |
|---|---|---|---|
| AHI | 3.01 | 2.87 | 2.53 |
| RDI | 3.61 | — | — |
| RERA | 4 | 2.14 | 1.10 |
| Leak avg (L/min) | 0.25 | 0.52 | 0.93 |
| Pressure avg (cmH₂O) | 8.77 | 8.69 | 8.61* |
| Usage hours | 6.65 | 6.70 | 6.37 |
| Deep sleep % | 16.1% | ~14.8%† | ~17.3%† |
| REM sleep % | 25.9% | ~22.5%† | ~23%† |
| Awake time (min) | 5.0 | ~14.7%† | ~17† |
*Pressure 30-day avg excludes 7 zero-pressure nights (June 15–22). †Sleep-stage 7-day/30-day figures are approximate from matched nights only; some nights in the 7-day window lack Apple Watch stage data.
Weekly AHI trend:
| Week | AHI avg | n |
|---|---|---|
| 06-15–06-21 (baseline) | 0.76 | 6 |
| 06-22–06-28 | 2.84 | 7 |
| 06-29–07-05 | 3.24 | 7 |
| 07-06–07-12 | 3.00 | 7 |
| 07-13–07-15 (partial) | 2.54 | 3 |
AHI remains elevated vs. baseline (2.54 this partial week vs. 0.76 baseline) but has not worsened week-over-week since late June.