Original Research: Pupillary Constriction Velocity and Latency to Predict Excessive Daytime Sleepiness

Background: Excessive daytime sleepiness (EDS) is common in adults. A need exists for an easier and faster objective clinical measures of EDS. The autonomic nervous system controls pupil size and prior pupillometry studies have demonstrated associations with sleepiness. We used a novel portable device to assess pupillometry and prospectively evaluated a sleep clinic cohort. Methods: Following IRB approval Pupillometry (The ForSiteTM, NeurOptics, Irvine, CA), was performed on 113 sleep clinic patients. Constriction and dilation velocity and latencies, minimum and maximum aperture were obtained along with Epworth Sleepiness Score (ESS), 10 point Visual Analog Scale (VAS), BMI, gender, age and AHI. Three sets of measures were obtained and analyzed with ANOVA, t-test, Linear Regression and Pearson correlation coefficients (SAS, Cary, NC). Results: Both constriction velocity and latency correlated with VAS (n = 88, r = 0.28, p = 0.007 and r = 0.31, p = 0.004). Only constriction velocity correlated with AHI (n = 78, r = -0.27, p = 0.016). Multivariate linear regression which includes VAS and age predicted constriction velocity (r = 0.36, p = 0.002) and latency (r = 0.38, p = 0.001). Using Pearson correlation, AHI and VAS combined were associated with constriction velocity (-0.273 (0.016), and 0.284 (0.007), respectively). Using a maximum constriction velocity threshold value (age adjusted) of 2.8, VAS ≥ 6 was predicted with a sensitivity of 83% and specificity of 84%. Conclusions: Pupillary constriction velocity and latency predict self-reported VAS state of sleepiness. While both are affected by age, only constriction velocity is affected by apnea severity. These data suggest that a portable pupillometer may provide a method to identify individuals with abnormal sleepiness.