Personal Pain Sensitivity Prediction from Ultra-short-term Resting Heart Rate Variability

Pain is a subjective experience with interpersonal perception sensitivity differences. Pain sensitivity is of scientific and clinical interest, as it is a risk factor for several pain conditions. Resting heart rate variability (HRV) is a potential pain sensitivity measure reflecting the parasympathetic tone and baroreflex function, but it remains unclear how well the prediction can achieve. This work investigated the relationship between different ultra-short-term HRV features and various pain sensitivity representations from heat and electrical pain tests. From leave-subject-out cross-validated results, we found that HRV can better predict a composite pain sensitivity score built from different tests and measures than a single measure in terms of the agreement between predictions and observations. Heat pain sensitivity was more possibly predicted than electrical pain. SDNN, RMSSD and LF better predicted the composite pain sensitivity score than other feature combinations, consis-tent with pain's physical and emotional attributes. It should be emphasized that the validity is probably limited within HRV at the resting state rather than an arbitrary measurement. This work implies a potential pain sensitivity prediction possibility that may be worth further validation.