Bedside diagnosis of cardiac autonomic damage by computerized analysis of heart rate-respiration relationship.

In this study we propose a method for the analysis of the relationship between heart rate changes and respiration as a possible diagnostic tool for cardiac autonomic damage. The method consists in recording R-R intervals and respiratory amplitude by a suitably equipped personal computer, and by evaluating the cross-correlation peak between the two signals. This mathematical function appeared to be more sensitive to the degree of concordance between the two signals, rather than their absolute amplitude. The cross-correlation appeared to be lower in diabetics with autonomic dysfunction, markedly decreased after injection of atropine (only in normals), slightly increased after propranolol. Hyperpnea increased the cross-correlation peak between 3-18 breaths/min in normals, but only at lower frequencies, if at all, in diabetic subjects with various degrees of autonomic dysfunction. The cross-correlation showed the best reproducibility among R-R change tests. These preliminary results suggest that this method may provide new information on autonomic integrity and a substantial advantage in terms of reproducibility.