. 2023 Jan 4;10:1016211. doi: 10.3389/fped.2022.1016211

Table 4.

HRV multivariable analysis (HRV prediction model), n = 101 infants.

	OR^a (95% CI)	p value	AUROC (95% CI)
Mean NN (ms)^b	2.73 (1.34–5.53)	0.005	0.837 (0.759–0.914)
HF power (ms²)^c	1.45 (0.44–4.81)	0.544
LF/HF ratio^c	1.48 (0.51–4.34)	0.471
TINN (ms)^c	0.61 (0.19–1.93)	0.398
MSE Complexity Index^b	0.60 (0.32–1.13)	0.115
MSE slope short^b	0.31 (0.12–0.79)	0.014

^{^a}

For the non-transformed variables the odds ratio represents the change in odds for a one-standard deviation increase in the HRV variable. For the transformed variables the odds ratio represents the change in odds for a one-standard deviation increase in the log HRV variable.

^{^b}

HRV variable was standardised prior to conducting the logistic regression.

^{^c}

HRV variable was log transformed (log base 10) and then standardised prior to conducting the logistic regression.

The multivariable model is: log (p/(1 − p)) = 0.49 + 1.00 (standardized Mean NN) + 0.37 (standardised log 10 HF power) + 0.40 (standardised log 10 LF HF ratio) −0.50 (standardised log 10 TINN) − 0.51 (standardised MSE c index) −1.16 (standardised MSE slope short) p = 0.480 for Hosmer-Lemeshow goodness of fit.

HRV, Heart Rate Variability; NN interval, normalised RR interval; HF power, high frequency power (0.15–0.4 Hz), LF/HF ratio, low frequency/high frequency ratio; TINN, triangular interpolation of the NN interval histogram; MSE, multiscale entropy.