Across 16,965 wearable-device users contributing over 5.2 million person-days of data, nightly marijuana use was associated with measurably altered cardiovascular and sleep physiology. Within individuals, use nights showed lower resting heart rate (−0.12 bpm) and reduced heart rate variability (−0.47 ms) — a marker of parasympathetic suppression — alongside longer sleep (+7.6 minutes) that was simultaneously more fragmented, with more disturbances (+0.47 events), greater wake time (+1.79 minutes), and altered sleep architecture. A notable secondary finding: self-reported marijuana use probability declined by 11.1% over 72 weeks of tracking, suggesting that wearable self-monitoring itself may reduce consumption.
This preprint, not yet peer-reviewed, carries real weight simply through its scale — it is among the largest objective real-world analyses of marijuana's physiological footprint to date, vastly exceeding typical laboratory studies of tens of participants. The HRV reduction is clinically meaningful context: chronically lower HRV associates with elevated cardiovascular risk and poorer resilience. The sleep finding is nuanced — longer duration sounds beneficial, but fragmentation undermines restorative sleep quality, particularly slow-wave and REM stages. Limitations are important: data are observational and self-reported, wearable sleep staging has known inaccuracies, and causality cannot be assumed. Reverse causation (poor sleepers self-medicating) is plausible. Still, the consistency of within-person effects strengthens inference. For adults using marijuana as a sleep aid, this large dataset offers a cautionary signal: sedation may lengthen sleep while degrading its architecture.