Meta-Analysis of Heart Rate Variability as a Psychophysiological Indicator of Posttraumatic Stress Disorder

Abstract

Background: Posttraumatic stress disorder (PTSD) syndrome is accompanied by the changes in autonomic nervous system, and heart rate variability (HRV) parameters assess the balance of sympathetic and parasympathetic influences on heart rate. HRV is a promising psychophysiological indicator of PTSD. The aim of this meta-analysis is to provide a quantitative account of the literature findings on PTSD and co-occurring HRV parameters. We first examined the effect size of PTSD on HRV in available published studies, and we then examined the effects of PTSD treatments on various HRV parameters. Methods: A literature search was done using three electronic databases; Pubmed, PsycINFO and Google Scholar. From each of the selected studies, we extracted study characteristics and outcomes measured. The metaanalysis was performed according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and Cochrane Handbook guidelines, using Comprehensive Meta-analysis Software, ver. 2.0. We calculated the Hedges’g effect size with 95% confidence interval (CI), statistical significance (p), and heterogeneity for each effect size estimate. We considered several variables, and for each an individual meta-analysis was performed. Results: Meta-analysis reveals the potential utility of HRV index variables as indicators of PTSD. For high frequency (HF) power, the aggregate mean effect size of 11 studies with 379 subjects was large and negative, showing reduced HF for PTSD patients relative to controls and indicating lower vagal activity in these patients. There were also significantly decreased RMSSD values, indicating lower vagal tone in subjects with PTSD. The aggregate mean effect size of low frequency (LF) power was small, negative, and significant, indicating slightly reduced LF in PTSD patients. The LF/HF ratio was statistically significantly higher in PTSD patients. The positive LF/HF effect size indicates increase in sympatho-vagal balance under the PTSD conditions as compared with the normal conditions, and further reflects non-linearity in co-occurring shifts in LF and HF power with proportionately greater reduction in HF than LF. Heart rate (HR) was significantly elevated in PTSD patients. Lastly, there was a decrease in SDNN in the patients, indicating decrease in the variability in RR intervals and a increase in sympathetic tone. These analyses showed no significant between-study heterogeneity for any of these parameters. Very few studies have examined changes in HRV variables pre- and post-treatment; using conservative random effect modeling, a significant increase in RMSSD could be discerned, and decrease in HR was nearly significant (p 1-tailed=0.08). Conclusion: Currently available scientific literature clearly shows that reductions HF, RMSSD and LF/HF have utility as indicators of autonomic effects of PTSD, which can be associated with impaired vagal activity. Increased LF/HF index in patients suggests non-linear decreases in both LF and HF yielding excess of sympathetic relative to vagal imbalance in PTSD patients. Data are scarce for treatment effects on HRV, but RMSSD appears to be increased.