Sources: health-report-2026-07-29-to-2026-07-30.json (Apple Watch sleep stages, exported 06:18 on 07-30) + OSCAR_Greg_by_Day_2026-07-29.csv (ResMed AirSense/AirFlow 11 via OSCAR). Matched by calendar date of sleep-session start (23:59 on 07-29).
Leak spiked to 1.52 L/min — the second-highest reading in the last 8+ nights — yet produced the best sleep-quality numbers of the window. Leak avg is nearly double both the 7-day average (0.93) and the 30-day average (0.80), and the leak 90th-percentile (6.0 L/min) is the highest since 07-25's 7.2. Despite that, awake time fell to 4.0 minutes (98.9% efficiency) — the lowest awake-time reading and highest efficiency in the entire recent trend table. This is a strong data point against leak driving sleep disruption on a given night, consistent with the weak leak↔awake correlation (r = +0.07) established below.
Pressure has now declined for three straight nights (6.87 → 6.77 → 6.63 cmH₂O, 07-27 through 07-29), each meaningfully below both the 7-day average (7.27) and 30-day average (8.11). This extends what the 07-28 report flagged as a 2-night dip into a sustained 3-night trend. No adverse effect on AHI or sleep architecture has appeared yet, but this is now a real trend worth continued tracking rather than a blip.
Third consecutive strong night for awake time/efficiency: 07-27 (5.5 min / 98.5%), 07-28 (11.5 min / 97.2%), 07-29 (4.0 min / 98.9%). This reinforces that the 07-25/07-26 disruption (30.6 min and 26.6 min awake) was a short-lived event, and tonight sets a new low for the window even with elevated leak in play.
| Metric | Value |
|---|---|
| Bedtime (session start) | 23:59 |
| Time in bed | 6.02 h (361.3 min) |
| Total sleep duration | 5.95 h (357.3 min) |
| Sleep efficiency (TST/TIB) | 98.9% — new high for the trend window |
| Awake time | 4.0 min — lowest in the trend window |
| Core sleep | 3.32 h (199.2 min) — 55.8% |
| Deep sleep | 1.06 h (63.7 min) — 17.8% |
| REM sleep | 1.57 h (94.3 min) — 26.4% |
| CPAP usage | 6.18 h |
| AHI | 1.62 |
| RDI | 1.94 |
| OA / H / CA / UA / RERA | 7 / 2 / 1 / 0 / 2 — tied for highest RERA of last 8 nights |
| Pressure avg | 6.63 cmH₂O — 3rd straight low reading |
| Pressure 90th %ile | 8.68 cmH₂O |
| Leak avg | 1.52 L/min — 2nd-highest of last 8 nights |
| Leak max | 25.2 L/min |
| Leak 90th %ile | 6.0 L/min — highest since 07-25 |
| Date | AHI | OA | RERA | Leak avg | Leak 90th | Deep % | Awake (min) | Efficiency |
|---|---|---|---|---|---|---|---|---|
| 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% |
| 07-26 | 1.24 | 5 | 0 | 0.39 | 0.0 | 14.3% | 26.6 | 92.9% |
| 07-27 | 1.6 | 8 | 0 | 1.19 | 2.4 | 18.0% | 5.5 | 98.5% |
| 07-28 | 2.01 | 8 | 2 | 0.77 | 2.4 | 17.8% | 11.5 | 97.2% |
| 07-29 | 1.62 | 7 | 2 | 1.52 | 6.0 | 17.8% | 4.0 | 98.9% |
Split at leak ≤ 0.30 vs > 0.30 L/min (07-29's leak of 1.52 falls into the higher-leak bucket):
the user r (leak avg vs. deep %): −0.29 (n=38). the user r (leak avg vs. awake min): +0.07. the user r (AHI vs. awake min): +0.18. the user r (AHI vs. deep %): +0.04.
Interpretation: The moderate negative leak↔deep-sleep association persists (higher-leak nights still average ~2 points lower deep %), but the leak↔awake and AHI↔awake correlations remain weak (|r| ≤ 0.18) — and tonight is a clean illustration of that weakness: the highest leak avg in over a week paired with the lowest awake-time reading in the window. Deep % held essentially flat at 17.8% (same as 07-28) despite the leak jump, so tonight doesn't move the leak↔deep relationship much either.
| Metric | Tonight (07-29) | 7-day avg (07-23–07-29) | Prior 7-day avg (07-22–07-28) | 30-day avg (06-30–07-29) |
|---|---|---|---|---|
| AHI | 1.62 | 1.69 | 1.68 | 2.32 |
| Leak avg (L/min) | 1.52 | 0.93 | 0.87 | 0.80 |
| Pressure avg (cmH₂O) | 6.63 | 7.27 | 7.41 | 8.11 |
| Usage hours | 6.18 | 6.66 | 6.76 | 6.57 |
| Deep sleep % | 17.8% | 17.14% | 16.30% | 16.69% |
| REM sleep % | 26.4% | 25.95% | 26.69% | 24.36% |
| Awake time (min) | 4.0 | 16.56 | 17.42 | 16.49 |
Leak is running well above both trailing windows and the 30-day baseline (roughly 1.6–1.9x). Pressure continues its slide below every comparison window. AHI sits right in line with the 7-day trailing average and comfortably below the 30-day average. Awake time is the standout: less than a quarter of any trailing average, extending the post-07-25/26 recovery to a third night.
| Week | AHI avg | Leak avg | n |
|---|---|---|---|
| 06-22 to 06-28 | 2.84 | 1.98 | 7 |
| 06-29 to 07-05 | 3.60 | 1.49 | 6* |
| 07-06 to 07-12 | 3.00 | 0.43 | 7 |
| 07-13 to 07-19 | 2.13 | 0.45 | 6* |
| 07-20 to 07-26 | 1.43 | 0.71 | 7 |
| 07-22 to 07-28 (prior trailing 7d) | 1.68 | 0.87 | 7 |
| 07-23 to 07-29 (current trailing 7d) | 1.69 | 0.93 | 7 |
* n=6 because one night in that week (07-03 and 07-17 respectively) has no matched Apple Watch sleep session — see caveats.
Trailing 7-day AHI is essentially flat (1.68 → 1.69) as 07-22's 1.59 aged out and tonight's 1.62 aged in. Trailing 7-day leak ticked up (0.87 → 0.93) as tonight's elevated 1.52 replaced 07-22's 1.08 — both above 07-22's contribution, so the week's leak average is drifting upward. Still well below the 06-22–07-05 range (1.49–1.98).
Awake time/efficiency is now on a 3-night streak of strong readings (5.5 min/98.5%, 11.5 min/97.2%, 4.0 min/98.9%), and tonight set a new low for the entire trend window. This happened despite the highest leak reading in over a week — the clearest evidence yet in this dataset that leak isn't the primary driver of awake time on a night-to-night basis.
Pressure trend now needs a real look, not just tracking: three consecutive nights below 6.9 cmH₂O (6.87 → 6.77 → 6.63), all well under the 7-day (7.27) and 30-day (8.11) averages. No downside has shown up in AHI or sleep quality yet, but a declining pressure trend over 3+ nights is exactly the kind of pattern worth flagging — if it continues, worth checking whether auto-titration is settling at a genuinely lower therapeutic pressure or whether something (mask fit, positional change) is suppressing the algorithm's response.
Leak spike (1.52 L/min, 90th %ile 6.0) is the other data point to watch, even though it didn't hurt sleep quality tonight. One elevated night isn't a trend, but it's the second-highest leak reading in the 8-night window (after 07-25's 1.69) — worth checking mask seal if it recurs tomorrow.
oscar_data.csv) — included in the leak ≤0.30 bucket in Section 2 as literal low readings, though a 0.0 that early may reflect incomplete data capture rather than a true zero-leak night.