Early transient dysautonomia predicts the risk of infantile epileptic spasm syndrome onset: A prospective cohort study

Abstract

Background: Infantile epileptic spasm syndrome (IESS) is an age-dependent epileptic encephalopathy with a significant risk of developmental regression. This study investigates the association between heart rate variability (HRV) in infants at risk of IESS and the clinical onset of IESS.

Methods: Sixty neonates at risk of IESS were prospectively followed from birth to 12 months with simultaneous electroencephalogram (EEG) and electrocardiogram recordings for 60 min at every 2-month interval. HRV metrics were calculated from 5 min time-epoch during sleep including frequency domain measures, Poincare analysis including cardiac vagal index (CVI) and cardiac sympathetic index (CSI), and detrended fluctuation analysis (DFA α1, DFA α2). To assess the effect of each HRV metric at the 2-month baseline on the time until the first occurrence of either hypsarrhythmia on EEG and/or clinical spasm, univariate cox-proportional hazard models were fitted for each HRV metric.

Results: Infantile epileptic spasm syndrome was diagnosed in 20/60 (33%) of the cohort in a 12-month follow-up and 3 (5%) were lost to follow-up. The median age of developing hypsarrhythmia was 25 (7-53) weeks and clinical spasms at 24 (8-40) weeks. Three (5%) patients had clinical spasms without hypsarrhythmia, and 5 (8%) patients had hypsarrhythmia before clinical spasms at the initial presentation. The infants with high CSI (hazard ratio 2.5, 95% CI 1.2-5.2, P = 0.01) and high DFA α1 (hazard ratio 16, 95% CI 1.1-240, P = 0.04) at 2 months were more likely to develop hypsarrhythmia by the first year of age. There was a trend toward decreasing CSI and DFA α1 and increasing CVI in the first 8 months of age.

Conclusion: Our data suggest that relative sympathetic predominance at an early age of 2 months may be a potential predictor for developing IESS. Hence, early HRV patterns may provide valuable prognostic information in children at risk of IESS allowing early detection and optimization of cognitive outcomes. Whether early intervention to restore sympathovagal balance per se would provide clinical benefit must be addressed by future studies.

Keywords: autonomic dysfunction; heart rate variability; hypsarrhythmia; infantile spasm (IS); perinatal brain injury.

Figure 1

Schematic demonstrating the steps in the analysis of heart rate variability. Raw ECG data were sampled at 256 Hz and the 5 min segment of ECG recordings was extracted from seizure-free epochs during NREM sleep. Pan–Tompkin's algorithm was applied to compute RRi and generate an equidistantly sampled RR time series. Standard frequency domain HRV measures included normalized power of variability in two standard frequency bands HF (0.15–0.4 Hz) and LF (0.04–0.15 Hz). The Poincaré plot is a scatter plot representing the correlation between consecutive RRi. SD1 represents the standard deviations of the points perpendicular to the line of identity, while SD2 represents the standard deviation of the points along the line of identity. DFA measures the correlations within the data for different time scales. Variability is characterized by the slope of the regression line obtained as alpha1 (range of 4–16 beats) and alpha 2 (range of 16–64 beats). The slope ranges in value from 0.5 (random) to 1.5 (correlated). ECG, electroencephalogram; NREM, Non-Rapid eye movement sleep; R-R, R-R peak interval; FTT, Fast Fourier transform; HF, High-frequency filter; LF, Low-frequency filter; CVI, cardiac vagal index; CSI, cardiac sympathetic index; SD1, Standard deviation direction perpendicular; SD2, standard deviation direction diagonal. Adapted from Goswami et al. (23).

Figure 2

Comparison of HRV metrics between infants with and without the diagnosis of IESS at each age group. There was no statistically significant difference in any HRV metric between infants with and without the diagnosis of IESS at all time points tested. IESS, Infantile epileptic spasm syndrome; HRV, heart rate variability.

Figure 3

Heart rate variability trajectory at 1 year of age in infants with IESS. There was a trend toward an increase in mean CVI values at consecutive visits until the postnatal age of 8 months after which there was a plateau. There was also a simultaneous decrease in CSI and DFA α1, which also plateaus around 8 months. IESS, Infantile epileptic spasm syndrome; HRV, heart rate variability; CVI, cardiac vagal index; CSI, cardiac sympathetic index; DFA α1, detrended fluctuation analysis alpha 1.