Linear analysis is surprisingly useful for understanding hypertension in the Guyton model

Abstract

Given populations of normotensive and hypertensive “virtual individuals” (over 200,000 simulations with randomized parameter values) based on the Guyton model (Guyton et al. 1972; van Vliet & Montani 2005), we used linear analysis techniques to identify the key factors that determine which simulations result in hypertension. A Generalized Linear Model (GLM) was fitted to each population to predict the long‐term arterial pressure from the model parameters. The GLMs were then reduced to a minimal set of the model parameters, and we compared the effect of each parameter on the two GLMs, demonstrating a difference in parameter sensitivity. We also used this approach to examine the parameter sens..