Vitals Knowledge Vault

Sleep architecture

2 min read

Sleep Architecture

TL;DR

Sleep architecture refers to the structure of sleep stages across a night — N1, N2, N3 (slow-wave), and REM. Substances affect this differently; wearable detection relies on stage-distortion patterns rather than any single metric.

Why it matters for Vitals

Sleep stage architecture is one of the most directly wearable-accessible substance signatures. Apple Watch sleep staging (Series 6+) provides enough resolution to detect meaningful disruptions. Architecture changes are often the clearest next-day recovery signal.

Key Patterns by Substance

Cannabis

  • REM ↓↓ (strongest signal, ~34 min same-night reduction)
  • Sleep onset may feel easier (CB1 sedation) but objective quality is degraded
  • Chronic use → partial tolerance; withdrawal → REM rebound
  • See REM suppression for the primary cannabis sleep signature

Alcohol

  • N3 (SWS) ↑ early in night, then ↓ in second half
  • REM ↓ throughout; fragmentation ↑
  • WASO increases; recovery severely impaired despite subjective sleep quality
  • See Sleep architecture (alcohol)

Cocaine

  • Sleep onset severely delayed
  • N3 suppressed; REM delayed and reduced
  • Significant next-day sleep debt accumulation
  • See Cocaine sleep architecture

Wearable Detection

Primary Watch-accessible signals:

  • REM time and percentage (via HKSleepAnalysis stage data)
  • N3 time (SWS proxy — Apple Watch estimates, not EEG)
  • Sleep efficiency = (time asleep) / (time in bed)
  • WASO (wake after sleep onset)

Signal reliability:

  • REM detection: moderate (Apple Watch estimates REM from HR + motion, not EEG)
  • N3 detection: weak (Watch doesn’t measure delta waves; SWS proxy is unreliable)
  • Sleep efficiency: moderate-strong (time-based measure is fairly accurate)
  • WASO: moderate (individual wake events are detected but short arousals are noisy)

Substance-Specific Notes

General Caveats

  • Apple Watch sleep staging is an estimation, not clinical polysomnography
  • Individual baseline variability is high; within-subject change is more reliable than cross-subject comparison
  • Sleep stage effects compound with repeated use; acute vs chronic patterns differ substantially
  • Sleep Architecture Enhancement — device validation evidence, recovery scoring algorithm, nocturnal HRV profiling, sleep deprivation meta-analysis (deeper dive, batch 8)

Graph View

Backlinks