Brief interval heart period variability by different methods of analysis correlates highly with 24 h analyses in normals

R. P. Sloan, P. A. Shapiro, E. Bagiella, M. M. Myers, J. T. Bigger, R. C. Steinman, J. M. Gorman

Research output: Contribution to journalArticlepeer-review

18 Scopus citations

Abstract

Heart period variability (HPV) measured from 24 h ECG recordings predicts mortality following myocardial infarction and may be a measure of cardiovascular health in the general population. Since epidemiologic evaluation of healthy people will require alternatives less intensive than 24 h recording, we investigated the relationship between HPV derived from 24 h and 5 min recordings, using two approaches for obtaining RR intervals. Template-matching (TM) algorithms were applied to 24 h ECG recordings from 41 normal subjects (mean age 35.7 ± 13 years). Five min of ECG data during this 24 h period also were collected by an on-line microcomputer-based system for peak detection (PD) analysis. Intraclass correlations comparing the TM and PD approaches on the 5 min period were .80 or greater for all measures of HPV. Pearson correlation coefficients between the 5 min (TM) estimates and 24 h data and 5 min (PD) estimates and 24 h data exceeded .60 and .55, respectively, for all but one variable, with all p values < .05. Thus, in healthy adults, TM and PD approaches to HPV estimation from short segments of ECG data are highly consistent and the correlations between HPV obtained from brief intervals and 24 h measures were substantial, suggesting that assessment of HPV as a screening measure of cardiac autonomic control in healthy adults may be feasible.

Original languageEnglish
Pages (from-to)133-142
Number of pages10
JournalBiological Psychology
Volume38
Issue number2-3
DOIs
StatePublished - Oct 1994
Externally publishedYes

Fingerprint

Dive into the research topics of 'Brief interval heart period variability by different methods of analysis correlates highly with 24 h analyses in normals'. Together they form a unique fingerprint.

Cite this