Correlation of physiological stress as measured by heart rate and blood pressure variability and self-reported anxiety

Abstract Background Despite lack of data, there is an assumption that anxiety symptoms correspond with a physiological stress response with decreased parasympathetic and increased sympathetic nervous system. Objective To examine whether self-reported measures of anxiety correlate with physiological stress. Methods We performed a secondary analysis of data collected from a 12-week clinical trial on slow breathing that examined the changes in psychological and physiological stress (n = 88). Psychological stress was assessed with PROMIS Anxiety CAT. Physiological stress was assessed by spectral analysis of heart rate including RMSSD, LFRRI, HFRRI, LFRRI/HFRRI ratio and blood pressure variability (LFsys). Analyses consisted of Spearman correlations calculated at specific time points (baseline, 6, and 12 weeks) and changes over time (6 weeks vs. baseline, 12 weeks vs baseline) between PROMIS anxiety and physiological stress measures. Also, linear mixed-effects models were fitted on PROMIS Anxiety for physiological measures. Results Despite a decrease in mean PROMIS Anxiety from baseline to 12 weeks (-4.85 S.D. ± 5.53) from slow breathing, we found a low correlation between PROMIS Anxiety and physiological measures of stress for specific time points or changes over time. Measures of autonomic tone, including LFRRI, HFRRI, LFRRI/HFRRI ratio and LFSYS showed poor correlation to PROMIS anxiety measures at baseline, 6, and 12 weeks. This suggests a disconnect between self-reported anxiety and indexes of autonomic tone. Conclusion We report a lack of correlation between self-reported anxiety and physiological measures of stress despite reductions in self-reported stress after a behavioral intervention for relaxation.