Non-linearity of the association between heart rate variability and severity of illness

I read with great interest on the study by Arbo JE and colleagues [1], which investigated the severity of illness and heart rate variability (HRV). HRV is a biomarker that low and high values will indicate unhealthy status, and there is a physiological range indicating the heart is performing well (healthy irregularity). In this situation, the correlation between HRV and severity of illness cannot be fully described by a linear function. Although the exact functional form between HRV and severity of illness is not known, it is a typical practice to check the linearity assumption in building linear regression model.

Source: The American Journal of Emergency Medicine - February 17, 2020 Category: Emergency Medicine Authors: Yishan Hu Tags: Correspondence Source Type: research