Sources: health-report-2026-07-14-to-2026-07-15.json (Apple Watch sleep stages, exported 05:53 on 07-15) + OSCAR_Greg_by_Day_2026-07-14.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (22:49 on 07-14).
Deep sleep recovery: 19.3% — the 5-night suppression streak is broken. July 9–13 averaged 12.8% deep sleep (5 consecutive nights below 16%, the longest streak in the dataset). Tonight bounced back to 19.3%, near the 30-day average of ~17.3%. Awake time also dropped sharply to 10.5 min (best in recent weeks; 7-day avg was 17.8 min). This is a meaningful reversal, though a single night is not yet a confirmed trend shift.
RERA=3 for the 3rd consecutive night — a new record in this dataset. The July 13 report flagged back-to-back RERA=3 (July 12 and 13) as unprecedented. July 14 extends that to three nights in a row. Despite deep sleep recovering tonight, RERA remains persistently elevated. The three-night RERA=3 streak (July 12–14) now stands as the clearest sustained signal in this dataset of upper airway resistance / flow limitation independent of scored AHI events.
REM compressed to 17.8% — sharp drop vs. recent baseline. The 7-day REM average through July 14 is 24.5%; tonight is 7 pp below that. While deep sleep recovered, REM gave substantial ground. This architectural shift — deep recovering at REM's expense — is unusual and worth watching over the next few nights. If REM stays compressed while deep normalizes, it may reflect a persistent arousal dynamic that alternates which restorative stage it suppresses.
OA count = 16, AHI = 3.18 — obstructive burden elevated. OA=16 equals the July 7 count (the highest since July 11's OA=21). With AHI at 3.18 (above the 7-day avg of 2.78 and 30-day avg of 2.51), obstructive events are driving tonight's event load. RERA=3 on top of this is a combined picture of both frank obstruction and residual flow limitation in the same night.
| Metric | Value |
|---|---|
| Bedtime (session start) | 22:49 |
| Time in bed | 5.99 h (5h 59m) |
| Total sleep duration | 5.81 h (5h 49m) |
| Sleep efficiency (TST/TIB) | 97.1% |
| Awake time | 10.5 min (best in recent weeks) |
| Core sleep | 3.65 h — 62.9% |
| Deep sleep | 1.12 h — 19.3% (streak broken) |
| REM sleep | 1.04 h — 17.8% (below recent baseline) |
| CPAP usage | 6.28 h |
| AHI | 3.18 |
| RDI | 3.66 |
| OA / H / CA / UA / RERA | 16 / 4 / 0 / 0 / 3 |
| Pressure avg | 9.32 cmH₂O |
| Pressure 95th %ile | 11.36 cmH₂O |
| Leak avg | 0.30 L/min |
| Leak 95th %ile | 0* (see caveats) |
Correlations from prior reports remain directionally valid; adding tonight (now 24 matched nights) does not meaningfully shift them.
| Relationship | Direction | Notes |
|---|---|---|
| AHI vs. awake time | Moderate positive (~r=0.44) | Tonight: AHI 3.18, awake 10.5 min — below trend (deep/REM absorbed instead?) |
| Leak vs. deep sleep % | Moderate negative (~r=−0.44) | Tonight: leak low (0.30), deep recovered — consistent |
| RERA vs. deep sleep % | Inverse (qualitative) | RERA=3 yet deep=19.3% tonight — prior nights showed suppression under RERA=3 |
Tonight's AHI vs. awake time: Despite AHI of 3.18 (above recent average), awake time was only 10.5 min — the lowest in the dataset's recent window. This is unusual. One interpretation: the obstructive events tonight were clustered in non-arousal windows, while the RERA events (which previously correlated with sleep-stage disruption) may have occurred without causing scored awakenings even as they shaped stage architecture (suppressing REM rather than generating awakenings).
OA=16 context across recent high-OA nights:
| Date | OA | AHI | Deep % | Awake (min) | RERA |
|---|---|---|---|---|---|
| 07-07 | 16 | 3.71 | 26.3% | 10.5 | 1 |
| 07-09 | 13 | 2.12 | 10.6% | 25.6 | 0 |
| 07-11 | 21 | 4.66 | 14.3% | 24.6 | 0 |
| 07-14 | 16 | 3.18 | 19.3% | 10.5 | 3 |
July 7 is the closest analog to tonight: OA=16, awake=10.5 min, deep=26.3%. July 7 had minimal RERA (1) but excellent deep sleep. Tonight adds RERA=3 to a similar OA load, yet deep sleep also improved. The deep-sleep recovery tonight likely reflects normal night-to-night variation rebounding off the multi-night suppression trough, rather than RERA having resolved.
Three-night RERA=3 streak (July 12–14):
| Date | AHI | Deep % | RERA | Awake (min) |
|---|---|---|---|---|
| 07-12 | 2.70 | 12.6% | 3 | 16.1 |
| 07-13 | 1.44 | 11.4% | 3 | 16.1 |
| 07-14 | 3.18 | 19.3% | 3 | 10.5 |
RERA=3 has now persisted through a low-AHI night (1.44) and a higher-AHI night (3.18), with deep sleep both suppressed and normal. The persistence of RERA=3 is the consistent thread regardless of AHI or deep-sleep outcome. This sustained flow-limitation signal, across 3 consecutive nights, points to a structural or positional contributor (e.g., jaw position, nasal resistance, pillow height) that isn't resolving night-to-night.
| Metric | Tonight (07-14) | 7-day avg (07-08–07-14) | 30-day avg (06-15–07-14) |
|---|---|---|---|
| AHI | 3.18 | 2.78 | 2.51 |
| RDI | 3.66 | — | — |
| Leak avg (L/min) | 0.30 | 0.52 | ~1.00 |
| Pressure avg (cmH₂O) | 9.32 | 8.61 | ~8.55* |
| Pressure 95th (cmH₂O) | 11.36 | — | — |
| Usage hours | 6.28 | 6.76 | 6.49 |
| Deep sleep % | 19.3% | 15.6% | ~17.3% |
| REM sleep % | 17.8% | 24.5% | ~23% |
| Awake time (min) | 10.5 | 17.8 | ~17 |
*Excludes early zero-pressure nights (06-15 to 06-22).
Deep sleep: prior 5-night suppression streak (July 9–13) vs. recovery:
| Date | Deep % | AHI | Leak | RERA |
|---|---|---|---|---|
| 07-05 to 07-08 (prior high-deep period avg) | ~25.4% | 2.24 | 0.13 | 0.5 |
| 07-09 (streak begins) | 10.6% | 2.12 | 1.57 | 0 |
| 07-10 | 15.2% | 2.98 | 0.18 | 2 |
| 07-11 | 14.3% | 4.66 | 0.15 | 0 |
| 07-12 | 12.6% | 2.70 | 0.68 | 3 |
| 07-13 | 11.4% | 1.44 | 0.53 | 3 |
| 07-14 (tonight) | 19.3% | 3.18 | 0.30 | 3 |
Deep sleep recovered to near the 30-day average tonight. However, REM (17.8%) is now 5+ pp below its 30-day average (~23%), suggesting the recovery may be redistributing stage architecture rather than fully resolving the underlying disruption.
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 |
| 07-13–07-14 (partial) | 2.31 | 2 |
The current partial week (July 13–14) averages 2.31/hr, below the prior three weeks. Whether this is a genuine trend improvement or driven by the low-AHI night on July 13 (1.44) remains to be seen as the week fills in.