Breathing Rate Analysis & BOS Screening
Breathing Rate Trends
Distribution Comparison
Mann-Whitney U: p<0.001 · Cohen's d: -2.83 (large) · 95% CI for mean difference: (-1.52, -1.18) brpm
Ruxolitinib Effect on Breathing Rate
Respiratory-Autonomic Coupling
Respiratory sinus arrhythmia (RSA): In healthy individuals, higher HRV correlates with lower breathing rate through vagal modulation. Disrupted coupling may indicate autonomic dysfunction.
Patient 1: r=0.10, p=0.342, n=88 -- Coupling disrupted/weak
Patient 2: r=0.01, p=0.900, n=518 -- Coupling disrupted/weak
Patient 3: r=-0.44, p=0.043, n=22 -- Coupling intact
BOS Screening Panel
Clinical Interpretation
BOS Screening: P1's breathing rate trend is classified as Low Risk. The linear trend slope is +0.0052 brpm/day (0.0% of nights above the 18.0 brpm elevated threshold). Post-HSCT patients require ongoing monitoring for BOS, especially with chronic GVHD.
Ruxolitinib Effect: Statistically significant difference in breathing rate pre vs post Ruxolitinib (delta: +0.51 brpm, effect: medium). Ruxolitinib, a JAK inhibitor, reduces inflammatory cytokines and may modulate respiratory drive indirectly through GVHD suppression.
Autonomic Coupling: Respiratory-autonomic coupling is disrupted or weak in Patient 1 (Spearman r=0.10). Disrupted breath-HRV coupling in HSCT patients may reflect autonomic neuropathy or chronic inflammatory burden affecting vagal tone.
Anomalous Nights (Patient 1)
Nights with breathing rate >2 SD above personal mean (3 detected):
| Date | Breath (brpm) | Z-Score | HRV (ms) | HR (bpm) |
|---|---|---|---|---|
| 2026-01-20 | 15.9 | 2.3 | 5.0 | 96.5 |
| 2026-01-22 | 15.8 | 2.1 | 8.0 | 91.625 |
| 2026-04-11 | 16.0 | 2.5 | 12.0 | 74.625 |
This analysis uses nighttime breathing rate from Oura Ring sleep periods (long sleep only). Clinical BOS diagnosis requires pulmonary function tests (FEV1, DLCO). Wearable breathing rate is a screening adjunct, not diagnostic.
Methodology
Data Source: Oura Ring sleep periods (type=long_sleep), average_breath column. Both patients' data loaded from separate SQLite databases.
BOS Screening: Linear regression on daily breathing rate. Thresholds: elevated >18.0 brpm, concerning trend >0.02 brpm/day. Normal sleep breathing range: 12.0-20.0 brpm.
Ruxolitinib Comparison: Mann-Whitney U test (non-parametric, appropriate for non-normal distributions). Cohen's d for effect size, bootstrap 95% CI for mean difference (10,000 iterations).
Cross-Patient Comparison: Mann-Whitney U with Cohen's d. Z-score normalization relative to each patient's own mean/SD for fair comparison.
Coupling Analysis: Spearman rank correlation between breathing rate and HRV (RMSSD) on overlapping dates. Respiratory sinus arrhythmia predicts negative correlation (higher HRV, lower breath rate) when vagal modulation is intact.
Anomaly Detection: Nights with breathing rate >2.0 SD above personal mean are flagged. Co-occurring HRV and HR values provide clinical context.