Association between skin temperature variability and sleep apnea severity: findings from a pilot study

Obstructive sleep apnea (OSA) may disrupt autonomic and vascular control, affecting skin temperature variability (STV). We aimed to assess whether STV (measured at distal and proximal sites) is associated with OSA severity (apnea-hypopnea index [AHI] and percent of sleep time with oxygen saturation below 90% [T90]) and if STV can aid in non-invasive OSA risk screening.

Eighteen participants underwent two nights of polysomnography with simultaneous distal and proximal skin temperature recording using iButtons. STV was calculated by standard deviation, median absolute deviation, and Poincaré plot analyses for sleep onset-to-offset, non-rapid eye movement (NREM), rapid eye movement (REM), and wake intervals. Linear mixed-effect models assessed relationships between STV and AHI/T90. Mediation analysis evaluated the role of body mass index (BMI), and receiving-operator-characteristic-curve (ROC)-based analysis assessed the screening potential of STV for moderate-to-severe OSA. All p-values were adjusted by the Benjamini-Hochberg False Discovery Rate (FDR).

AHI showed a strong, negative association with distal STV during NREM sleep, especially median absolute deviation (fixed effect coefficient = - 8.16, FDR-adjusted p < 0.001). T90 was similarly associated (- 1.90, FDR-adjusted p < 0.01). No significant relationships were found during REM or wake. BMI did not mediate these associations. ROC analysis revealed high discriminative ability for moderate-to-severe OSA (AUC: 0.93-0.97).

Reduced distal STV during NREM sleep independently signals OSA severity and may serve as a promising, non-invasive biomarker for screening OSA risk.