Prognostic relevance of the interaction between short-term, metronome-paced heart rate variability, and inflammation: results from the population-based CARLA cohort study.

ABSTRACT

AIMS: To determine the interaction between HRV and inflammation and their association with cardiovascular/all-cause mortality in the general population. METHODS AND RESULTS: Subjects of the CARLA study (n = 1671; 778 women, 893 men, 45-83 years of age) were observed for an average follow-up period of 8.8 years (226 deaths, 70 cardiovascular deaths). Heart rate variability parameters were calculated from 5-min segments of 20-min resting electrocardiograms. High-sensitivity C-reactive protein (hsCRP), interleukin-6 (IL-6), and soluble tumour necrosis factor-alpha receptor type 1 (sTNF-R1) were measured as inflammation parameters. The HRV parameters determined included the standard deviation of normal-to-normal intervals (SDNN), the root-mean-square of successive normal-interval differences (RMSSD), the low- and high-frequency (HF) power, the ratio of both, and non-linear parameters [Poincaré plot (SD1, SD2, SD1/SD2), short-term detrended fluctuation analysis]. We estimated hazard ratios by using covariate-adjusted Cox regression for cardiovascular and all-cause mortality incorporating an interaction term of HRV/inflammation parameters. Relative excess risk due to interactions (RERIs) were computed. We found an interaction effect of sTNF-R1 with SDNN (RERI 0.5; 99% confidence interval (CI) 0.1-1.0), and a weaker effect with RMSSD (RERI 0.4; 99% CI 0.0-0.9) and HF (RERI 0.4; 99% CI 0.0-0.9) with respect to cardiovascular mortality on an additive scale after covariate adjustment. Neither IL-6 nor hsCRP showed a significant interaction with the HRV parameters. CONCLUSION: A change in TNF-α levels or the autonomic nervous system influences the mortality risk through both entities simultaneously. Thus, TNF-α and HRV need to be considered when predicating mortality.