Week 5 is the fifth active Sleepgenic Weekly Report, and like Week 4 before it, it answers a question the previous report asked. The Week 4 report ended with an explicit test: REM had declined for three consecutive weeks, from 1.10 to 0.67 to 0.17 hours, and the question was whether the pressure coming off would let REM rebuild. If REM returned toward the 1.0-hour range, Week 4 would read as the trough of a recoverable strain cycle. If REM stayed suppressed, the interpretation would shift from strain to something more persistent.
Week 5 gave a partial answer. REM rebuilt — but not all the way.
Average REM sleep in Week 5 was 0.62 hours, up from Week 4's 0.17. That is a substantial reversal. After three weeks of decline and a near-total collapse in Week 4, REM moving back to 0.62 hours is the first upward move in the trajectory since Week 2. The freefall stopped, and it reversed. That is the encouraging half of the week.
The qualifying half is that 0.62 hours is still below the Week 0 baseline median of 0.81 hours, and below the 1.0-hour range that Week 4 named as the threshold for full recovery. REM moved decisively in the right direction without reaching the level that would let the prior month read as a cleanly closed strain cycle. The within-week pattern reflects this: REM was effectively absent on several nights — 0.08 hours on May 20, no reading on May 17 — but closed the week strong, with 0.82 hours on May 21 and 1.25 hours on May 22, the single best REM night since Week 2.
At the score layer, Week 5 recovered in step with REM. The average Overall Score was 64.3, up from Week 4's 52.1 — a 12-point rebound that recovers most of the Week 4 loss, though it remains just under the 67.5 baseline. Recovery Score rose to 68.7 from 63.4. Duration Score reached 75.0, the highest of the entire run, and total sleep averaged 6.84 hours, also the highest of the run. The body was spending ample time asleep in Week 5. The limiting factor was not duration; it was the architecture inside that duration, where REM remained below baseline.
The physiological layer was stable rather than dramatic. HRV averaged 37.7 ms, essentially flat against Week 4's 37.6 — a body neither under acute strain nor in strong recovery, holding steady. Resting heart rate averaged 61.6 bpm, improved from Week 4's 62.9. Deep sleep held at 1.41 hours, again above baseline, continuing the pattern seen across the entire run: deep sleep is consistently defended while REM is the variable stage. Sleep Stress averaged 20.4, down from Week 4's 23.6 but still elevated against Weeks 2 and 3 — the stress signal eased but did not fully settle.
The most important signal in Week 5 only becomes visible because the training calendar has not changed across the entire run. Every week has followed the same stimulus pattern — Rest, Summit, two Recovery days, Rucking, Run, Walk — by design, for exactly this kind of comparison. When the load on each day is held constant, the sleep response to each specific day can be tracked across weeks, and one day stands out.
The Summit night — the highest-load day of the week — is where the strain is concentrating. Across the run, the night following the Summit scored 78 in Week 2 and 73 in Week 3. In Week 4 it fell to 43. In Week 5 it fell further, to 38 — the single worst night of Week 5, with a Recovery Score of 34, Sleep Stress of 35.8, and no recorded REM despite 8.37 hours in bed. A long night that produced almost no restoration. The same summit effort that the body absorbed cleanly in Weeks 2 and 3 is now producing its worst sleep of the week.
This matters because the rest of Week 5 recovered well. The Run night scored 73 and was tagged "Positive Recovering." The Walk night scored 81, the week's high, tagged "Positive Long And Deep." The lower-load days are restoring normally. It is specifically the highest-load day that the body is no longer absorbing. That is a more precise finding than a flat weekly average can show: Week 5's recovery is real but uneven, and the unevenness is organized around load intensity. The harder the day, the worse the night that follows — and the gap between the easy days and the hard day is widening.
The interpretation is that Week 5 is a partial reversal with the strain concentrating at high load. At the whole-week level, the trajectory is repairing: REM turned upward, scores recovered most of their loss, resting heart rate improved, Sleep Stress eased. But underneath the recovering average, the response has become load-dependent in a way it was not earlier in the run. In Weeks 2 and 3, the Summit night was among the better nights of the week. By Week 5, it is the worst. The body is still handling the moderate and light days — and is now failing to convert the single hardest day into restorative sleep.
Sleepgenic classifies Week 5 as partial reversal with concentrating high-load strain. The Week 4 trajectory question is answered in the encouraging direction at the aggregate level: REM is not persistently blocked, because it rebuilt. But the same-day comparison surfaces a second, more specific signal — the strain has not gone away so much as relocated, from a broad multi-week REM decline into a sharp deficit on the highest-load night. Across five weeks, the longitudinal picture is now clearly oscillatory rather than linear — Overall Score has moved 60.2, 70.7, 68.1, 52.1, 64.3 — and within that oscillation, the foundation holds (deep sleep, total sleep, HRV, resting heart rate all at or near baseline) while the cost is increasingly paid on the Summit night specifically.
The watch item for Week 6 is the Summit night. The aggregate question — does REM complete its return toward 0.81 hours — still stands. But the sharper question is whether the Summit-night deficit continues to deepen. If the night after the highest-load day keeps falling while the easier days recover, the signal is that the current Summit load has crossed the threshold of what the body can presently absorb, and the methodology's next job is to read that threshold directly. If the Summit night recovers alongside the rest of the week, the Week 4–5 stretch will read as a full strain-and-recovery cycle that briefly concentrated at high load before resolving.