Cardioventilatory coupling in heart rate variability: the value of standard analytical techniques

P. D. Larsen and D. C. Galletly

Section of Anaesthesia, Wellington School of Medicine, PO Box 7343, Wellington, New Zealand*Corresponding author

Accepted for publication: August 14, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
In a group of spontaneously breathing anaesthetized subjects, we examined the ability of simple spectral and non-linear methods to detect the presence of cardioventilatory coupling in heart rate time series. Using the proportional Shannon entropy (HRI–1) of the RI–1 interval (interval between inspiration and the preceding ECG R wave) as a measure of coupling, we found no correlation between HRI–1 and either the fractal dimension or approximate entropy of the heart rate time series. We also observed no difference in the distribution of heart rate variability (HRV) spectral power in three frequency ranges (high, 0.15–0.45 Hz; low, 0.08–0.15 Hz; very low, 0.02–0.08 Hz) between uncoupled epochs and coupling patterns I, III and IV. Because of its association with low breathing frequencies, pattern II coupling epochs showed exaggerated low-frequency power as the high-frequency ‘respiratory’ peak fell into the low-frequency range. We conclude that coupling pattern is largely independent of autonomic tone and that these standard methods of HRV analysis are limited in their ability to detect the presence of cardioventilatory coupling in heart rate time series.

Br J Anaesth 2001; 87: 819–26

Keywords: heart, heart rate; heart, cardioventilatory coupling


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
Beat-to-beat heart rate variability (HRV) is mediated primarily by variations in the level of sympathetic and parasympathetic modulation of the sinoatrial node. Certain methods of HRV analysis are therefore useful for examining cardiac autonomic tone.1 Because different reflexes within the heart rate control system are active within different frequency ranges, frequency domain analysis of heart rate time series allows the relative contribution of each reflex pathway to be distinguished. Three component periodicities make up the typical heart rate time series. Fast periodicities in the range 0.15–0.45 Hz are largely due to the influence of respiratory phase on vagal tone [respiratory sinus arrhythmia (RSA)]. Low-frequency periodicities, in the region of 0.08–0.15 Hz, are produced by baroreflex feedback loops associated with sympathetic, activity and periodicities in the frequency range 0.02–0.08 Hz have been variously ascribed to modulation by chemoreception, thermoregulation and the influence of vasomotor activity.2 These regions commonly overlap, as in the case of subjects breathing slowly at respiratory frequencies approaching 0.1 Hz.

In addition to frequency domain analysis, there has been considerable interest in the analysis of HRV using non-linear methods, such as the fractal dimension (DF) and approximate entropy (ApEn), both of which attempt to quantify the complexity of heart rate dynamics.

The fractal dimension is a quantification of the space-filling propensity of the heart rate time series. While in Euclidean geometry a line is one-dimensional and a surface is two-dimensional, an irregular curve can be thought of as a line which is attempting to fill a surface. Mathematically, we can therefore describe such a line as having a dimension between one and two. The more the line fills the plane, the higher will be its fractal dimension DF.3

Entropy is a quantification of the repetition of patterns within a given signal.4 Small values of entropy are associated with regularity of patterns within a signal, such that it may be possible to predict the recurrence of a previously identified pattern. Larger values of entropy are associated with greater apparent randomness. Steven Pincus has developed a measure of entropy, ‘approximate entropy’ (ApEn), as a modification of Kolmogorov–Sinai entropy to allow measurements of entropy from shorter data segments, with simpler mathematics and lower computational demands.5 6 ApEn has been applied to the study of HRV by a number of authors.5 7 8

Cardioventilatory coupling is a temporal coherence of cardiac rhythm and inspiratory timing.9 It is seen in resting subjects, during sleep and spontaneous-breathing general anaesthesia. Experimental and clinical observations are consistent with coupling being a triggering of inspiratory onset by a cardiovascular afferent(s) associated with a preceding heart beat.10 11 During a coupled or cardiac-initiated breath, inspiratory onset will occur a fixed interval (the coupling interval, typically 0.5 s) after an ECG R wave. However, the exact relationship between the timing of the ECG R wave and inspiratory onset is complex and will vary according to whether the breath has been initiated by the intrinsic inspiratory pacemaker or by the cardiac trigger. This variation leads to multiple patterns of coupling, which can be described in terms of variation in coupling interval and entrainment ratio (number of heart beats within each breath).911 These patterns have been classified as I, II, III, IV and uncoupled, although other patterns have been suggested both from clinical observation and from computer modelling.11

The temporal coherence of heart beats and inspiration is a major determinant of breath-to-breath fluctuations in breathing frequency during anaesthesia.10 As breathing influences HRV through respiratory sinus arrhythmia, it is to be expected that cardioventilatory coupling will contribute significantly to the pattern of HRV.9 12 13 By temporally aligning heart beats to the breathing cycle, and given a constant breathing period, heart beats will occur at constant positions within the breathing cycle from breath to breath. Heart beats are therefore subject to repeating patterns of respiratory mediated fluctuations in vagal tone (RSA), giving rise to repeating patterns of HRV. In a previous paper we have demonstrated that, during one particular pattern of coupling (pattern I), the precise positioning of heart beats within each breath results in maximal fluctuations in heart rate due to RSA.13 We have suggested that, on the basis of these repeating patterns of HRV, it may be possible to detect cardioventilatory coupling from heart rate time series. In a preliminary description we have shown that several geometrical features of heart rate time series may occur in association with cardioventilatory coupling12, and in the present work we have examined the relationship between three standard measures of HRV and coupling.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
Ethical approval was obtained for all data collection. The 98 subjects, the technique of anaesthesia and the data recording methods have been described in detail in a previous paper on the effect of coupling on ventilatory variability.10 All subjects were breathing spontaneously during general anaesthesia for elective surgical procedures, had no evidence of cardiorespiratory disease and were taking no medication which was likely to influence autonomic function.

Anaesthesia was induced in all subjects with propofol and maintained with isoflurane, nitrous oxide and oxygen. Opioids were given according to surgical indications. Subjects breathed through a laryngeal mask airway, with FIO2 adjusted to maintain arterial oxygen saturation (SpO2) at or above 96% at all times.

We monitored SpO2, end-tidal carbon dioxide (Datex Oscar, Datex-Ohmeda, Helsinki, Finland), non-invasive blood pressure (Dinamap) and ECG (lead CM5; Neo-trak 502; Corometrics, Connecticut, USA). Ventilatory timing was measured by an electronically triggered non-return valve within the breathing system. Continuous recordings of ECG and ventilatory timing were made using a Macintosh IIcx computer with a 16-bit ADC board (MIO-16; National Instruments, Austin, TX, USA) and a sampling rate of 500 Hz.

Data analysis
Quantitative and qualitative measures of cardioventilatory coupling
From the ECG and ventilatory timing signals, we determined the time from each R wave to the onset of the following inspiration (RI interval) and plotted successive RI intervals as a time series. Horizontal banding within these RI plots indicates a constant temporal alignment between heart beats and inspiration (i.e. cardioventilatory coupling). The heart beat which immediately precedes inspiratory onset is given a negative subscript (R–1) and the interval between that heart beat and the inspiration is designated the RI–1 interval. During coupling, RI–1 is generally the most constant RI interval.

Patterns of coupling were defined as follows:

Pattern I: constant RI–1 alignment and a constant number of heart beats in each ventilatory period (i.e. constant entrainment ratio);

Pattern II: constant RI–1 alignment but with a varying entrainment ratio;

Pattern III: RI–1 intervals alternate between two values from breath to breath;

Pattern IV: RI–1 interval alignment slowly changes (increases or decreases) from breath to breath, but holds transiently at the RI alignment associated with coupling;

Uncoupled: No consistent RI–1 interval alignment observed.

The dispersion of RI–1 interval values correlates inversely with coupling. During coupling patterns I and II the RI–1 interval will be virtually identical from breath to breath, whereas the RI–1 interval varies to a greater degree during coupling patterns III and IV, and in uncoupled time series will give the appearance of randomness. Statistical measures of RI–1 dispersion therefore correlate with the degree of heart beat to inspiratory alignment. On the basis of computer modelling of the coupling process, the RI–1 dispersion is a measure of the proportion of the breaths which have been triggered by a cardiac signal.11 In the present study the RI–1 interval dispersion was measured using proportional Shannon entropy of the RI–1 interval (HRI–1), the details of which have been published previously14 and are given in the Appendix. Values of HRI–1 below 0.7 are typically associated with the appearance of cardioventilatory coupling in RI–1 time series plots.

Spectral analysis
To determine the effect of cardioventilatory coupling on spectral measures of HRV, we extracted epochs of data, 256 s in length, displaying as near as possible a constant single pattern of coupling throughout.

Spectral analysis was performed using a method based on that of Akselrod and colleagues,1 15 which is given in the Appendix. The spectral power (integrated area under the power spectrum curve) was calculated in each subject for total power and in three frequency bands: high (0.15–0.45 Hz), low (0.08–0.15 Hz) and very low (0.02–0.08 Hz). The proportion of the total power in each frequency band was also calculated.

Data epochs for ApEn and DF
The ApEn and DF of RR interval time series are generally computed from data series of at least 500–1000 heart beats.6 8 16 In this study, data epochs of 500 heart beats were used. However, because the pattern and degree of coupling changes over time it was not possible to obtain a sufficient number of 500-heart beat epochs demonstrating a single pattern of coupling throughout for statistical analysis. We therefore extracted a single 500-beat epoch of heart rate time series from as many subjects as possible, without reference to the RI plot. A maximum of one epoch was extracted from each subject and, where possible, the heart rate time series was free of non-stationary trends.

For each epoch of heart rate data, we determined ApEn and DF and calculated HRI–1 from the associated RI interval plot. ApEn was calculated using the algorithm of Pincus5 and DF was calculated using the method of Katz;3 both of these methods are given in the Appendix.

Analysis
Data were acquired and variables were calculated using custom-written software in LabVIEW 5 (National Instruments, Austin, TX, USA). Statistical analysis was performed using Statview 5.

We used the Kruskal–Wallis test to examine the relationship between the total spectral power and the proportions of power in the high-, low- and very low-frequency bands, and the different patterns of coupling.

We examined the relationship between HRI–1 and total spectral power, proportional band power, DF and each of the ApEn scores (for m=2 as well as m=3, at rEn=0.025, 0.05, 0.075, 0.1, 0.15, 0.2, 0.3, 0.4, 0.5 and 0.6; m is the number of heart periods in a sequence and rEn is the tolerance range within which sequences of m RR intervals are considered to be similar) using the Spearman rank correlation.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
Cardioventilatory coupling was observed in all subjects, although, as we have described previously,10 the duration and degree of coupling varied considerably over time and between individuals. We extracted 70 epochs of 256 s for spectral analysis where a single coupling pattern dominated throughout the epoch. Where other patterns were interspersed briefly within these epochs, we ensured that these periods occurred towards the beginning or end of the selected epoch in order to minimize the effect on the Fourier analysis. These spectral epochs comprised 12 pattern I, 21 pattern II, 6 pattern III, 9 pattern IV and 23 uncoupled. For ApEn and DF analysis, we obtained 52 epochs of 500 heart beats meeting the criteria for ectopy and stationarity.

Spectral analysis
No significant difference in total spectral power was observed between the different patterns of coupling. However, there was a significant difference in the distribution of power, with a significantly greater proportion of power in the low-frequency band and a corresponding decrease in the proportion of high-frequency power observed in pattern II coupling epochs. This observation was related to the expected10 low breathing frequency in epochs with pattern II coupling (Table 1). As the breathing frequency decreased, movement of the high-frequency respiratory-related spectral peak into the range of the low-frequency band (defined as 0.08–0.15 Hz) occurred, resulting in an apparent increase in low-frequency power in this group. If we exclude pattern II epochs, there were no significant differences in distribution of power between epochs of different coupling patterns.


View this table:
[in this window]
[in a new window]
 
Table 1 Comparison of mean RR interval (mean RR), mean consecutive difference RR interval (mean CD RR), mean respiratory frequency (mean f), total spectral power 0.02–0.45 Hz (total power) the percentage power in the high, low and very low frequency bands (% high, % low, % very low) according to coupling pattern. Values given are mean (SD). Groups were compared with ANOVA (mean RR, mean CD RR, f, % power) or the Kruskil–Wallis test (total power). Where differences were identified, the unpaired t-test was used to compare differences between groups. * denotes significantly different compared to pattern I, {dagger} denotes significantly different compared to pattern II, {ddagger} denotes significantly different compared to pattern IV
 
Figure 1 shows an example of heart rate, respiratory frequency and RI interval time series, as well as power spectra of HRV for an epoch of pattern I coupling and an uncoupled epoch. Pattern I coupling is seen in the RI interval plot in (Fig. 1A), where a constant temporal relationship between heart beats and inspiration (occurring at time=0) was observed, with a constant three heart beats per breath. In contrast, there is no apparent structure in the RI interval plot in (B). The two power spectra demonstrate that, for both of these epochs, the vast majority of power was in the high-frequency band (at the respiratory frequency).



View larger version (28K):
[in this window]
[in a new window]
 
Fig 1 Representative plots of RR interval, respiratory frequency, RI interval and power spectra for (A) a pattern I coupling epoch and (B) an uncoupled epoch.

 
No significant correlation was observed between HRI–1 and the total HRV power, or proportion of power in the high-, low- and very low-frequency ranges in the group of epochs comprising patterns I, III, IV and uncoupled.

Approximate entropy
We observed no correlation between HRI–1 and ApEn for any combination of m or r values selected in this study (Table 2). ApEn correlated negatively with RR interval at r=0.025 for both m=2 and m=3 and positively with RR interval at 0.075, 0.1, 0.15 and 0.2 for both m=2 and m=3. We also observed a negative correlation between respiratory frequency and ApEn at all values of m and rEn with the exception of r= 0.025 for both m=2 and m=3, where no significant relationship was observed. The high-frequency spectral power correlated with all ApEn values except for that at rEn=0.05 with m=2 or 3; for low rEn values the correlation was negative and for values above 0.05 it was positive.


View this table:
[in this window]
[in a new window]
 
Table 2 Correlation between ApEn and RR interval, respiratory frequency, HRI–1, high-frequency (HF) power and DF from the 500 heart beat epochs. Spearman rank correlation coefficient ({rho}) is given for ApEn vs mean RR interval, mean respiratory frequency, proportional Shannon entropy of the RI–1 interval (HRI–1), high-frequency power and DF. *P=0.05; **P=0.01; ***P=0.005; ****P<0.001; n.s. = not significant
 
Fractal dimension
We observed no correlation between DF and HRI–1. DF correlated (1) positively with mean RR interval (P<0.05); (2) with all ApEn values except for that at r= 0.075 with m=2 or r=0.05 and 0.075 for m=3; this correlation was negative for r=0.025 and positive for the higher values; (3) positively with high-frequency spectral power ({rho}=0.93, P<0.0001) and the proportion of power in the high-frequency band ({rho}=0.57, P<0.0001); and (4) inversely with age ({rho}=–0.36, P=0.02). DF did not correlate with respiratory frequency.


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
In this study we explored the relationship between cardioventilatory coupling and standard spectral and non-linear measures of HRV. We examined heart rate time series from clinical subjects, identified the specific coupling pattern and measured coupling using entropy of the RI–1 interval. We observed no correlation between cardioventilatory coupling and any of the frequency and non-linear domain measures of HRV used.

RSA is determined by ventilatory phase-related vagal modulation. The degree of modulation is influenced by a variety of factors, including age, position, underlying vagal tone and respiratory frequency.17 Vagal tone is reduced soon after inspiratory onset, causing a brief acceleration in heart rate. Vagal tone then returns and heart rate decreases before the onset of the next inspiration. The vagal modulation, and hence heart rate variation, is therefore cyclic, with a maximum and minimum heart rate at specific phases of the breathing cycle. This cyclic RSA response can be demonstrated in the form of an RSA curve, where the duration of the RR interval between two consecutive R waves (Rn and Rn+1) is plotted against the time that Rn+1 occurs after inspiratory onset (or against the phase of the ventilatory cycle). Coupling relates to this curve in at least two ways.

First, in the majority of individuals with pattern I coupling, the two consecutive R waves which immediately follow inspiratory onset (R+1 and R+2) tend to fall at the extremes of the RSA curve.13 R+1 is therefore preceded by the longest RR interval and R+2 is preceded by the shortest RR interval that can occur for the particular RSA curve. This leads to an apparent maximization of RSA; that is, for most breaths the difference between the maximum and minimum RR intervals within the breath is as great as is possible for the given RSA curve. It would be expected that this would lead to an increase in power in the high (respiratory) frequency band. That we did not observe this in our pattern I data epochs may relate to at least two factors: (1) we know that the effect does not occur in all subjects;13 and (2) high-frequency power is already high and there is a wide inter-individual difference in spectral power; because the numbers of epochs with pattern I of sufficient length were small, it is therefore difficult statistically to detect significant variation in high-frequency power.

The absence of a relationship between coupling and distribution of spectral power suggests that the coupling pattern is independent of cardiac autonomic tone. This observation is consistent with our demonstration that the coupling pattern is determined by the ratio of heart rate to intrinsic breathing frequency as well as the magnitude of the cardiac-related afferent trigger.11

The HRV power spectrum will be influenced by factors other than autonomic tone. For example, the position and character of the high-frequency peak will be influenced by ventilatory frequency fluctuations. As coupling, specifically the coupling pattern, determines much of the ventilatory variability of spontaneously breathing anaesthetized subjects, the coupling pattern should affect the characteristics of the respiratory spectral peak. These minor effects are, however, beyond the scope of this paper.

The second relationship between the RSA curve and coupling arises because coupling is a triggering of inspiratory onset by cardiac activity and therefore inspiration is temporally aligned to the preceding, triggering heart beat (usually RI–1). It follows that each of the subsequent heart beats (R+1, R+2, etc.) will also align themselves to the preceding inspiratory onset. Thus each heart beat will become aligned to constant phases of the ventilatory cycle. Each heart beat within a ventilatory cycle will therefore be modulated differently according to its position on the RSA curve. In pattern I coupling, where the entrainment ratio remains constant (i.e. there is a constant number of heart beats within each ventilatory period), identically repeating sequences of HRV will occur with each breath. One observes this repetition as discrete banding of the RR interval time series.12 Coupling should therefore be associated with regularity of the heart rate time series. To a degree, this regularity may be disrupted if the entrainment ratio is varying from breath to breath, as it does in coupling patterns II, III and IV. Thus, although regularity should be a feature of HRV during pattern I coupling, this may be less likely for patterns II, III and IV and uncoupled epochs.

It has been suggested that ApEn has value in distinguishing between ‘regularity’ in HRV under pathophysiological conditions and greater ‘irregularity’ in HRV in ‘normal’ individuals.18 19 ApEn is, however, a family of statistics, and its correlation to any physiological process will depend upon the selection of values for N (number of RR intervals), m and r.6 In the present study the correlation between ApEn and DF or high-frequency spectral power (i.e. RSA) was critically dependent upon the selection of rEn in the range 0.025–0.075. It is possible that this dependence on the control variables may lead to entirely different conclusions. Thus, Palazzolo and colleagues16 observed a good correlation between ApEn (N=2048, m=2, r=0.07) and the high-frequency spectral component of HRV in resting dogs and found that sympathetic blockade did not alter ApEn, whereas parasympathetic blockade produced significant decreases in ApEn. It was therefore concluded that ApEn was positively related to RSA. In contrast, Mansier and colleagues4 computed ApEn (N=1000, m=2, r=0.05) in the mouse and found that the administration of atropine was associated with a significant increase in ApEn. The authors noted that there was a decrease in the total variability associated with atropine treatment, as measured by spectral analysis, but that the ‘complexity’ of the series increased according to ApEn and other non-linear measures. These authors therefore concluded that ApEn was inversely related to RSA. It seems probable that the divergent opinions of Mansier and Palazzolo can be explained by their choice of control variables.

ApEn did not correlate with the dispersion of the RI–1 interval (i.e. cardioventilatory coupling) at either m=2 or m=3 for any of the rEn values used in the present study. As coupling is associated with repeating geometrical patterns of HRV12 and ApEn is said to be a measure of regularity, this observation was surprising. Several explanations will now be considered.

(1)Coupling during anaesthesia is characterized by rapid transitions between coupling patterns, which are in turn determined by the heart rate/breathing frequency ratio.10 11 Each coupling pattern is associated with a different entrainment ratio variation from breath to breath. As the entrainment ratio variation will alter the position of heart beats on the RSA curve, transitions in coupling pattern will therefore disrupt the regularity of HRV and could thereby alter ApEn.

(2)In clinical time series, non-stationarity (an overall trend) in the heart rate time series is difficult to avoid. Thus, although repetitive patterns may occur throughout the time series, in the presence of trend they will fall outside the tolerance rEn. In order to reduce this effect, we also examined the correlation between ApEn and HRI–1 where ApEn was measured (for all 20 combinations of m and rEn) from the RR interval consecutive time series [i.e. (rr2–rr1), (rr3–rr2), (rr4–rr3)... (rrn–rrn-1)]. As with the ApEn derived from the raw RR time series, however, no significant correlation was found.

(3)In some subjects the magnitude of RSA is small and, as in paragraph (2) above, any small variation in HRV due to noise will mask the pattern repetition and cause repeating patterns to fall outside the tolerance rEn.

(4)It is possible that dispersion of the RI–1 interval, as measured by its Shannon entropy, is itself limited as a measure of coupling. In a previous paper we have demonstrated that HRI–1 correlates well with the appearance of coupling in the RI interval plot, in particular during patterns I and II. During coupling patterns III and IV, however, where RI–1 dispersion is greater, the difference in HRI–1 with uncoupled epochs is measurable but small. However, it should noted that ApEn and DF did not differ between pattern I and uncoupled epochs. In addition, having observed the lack of correlation between ApEn and HRI–1, we also examined the correlation between the set of 20 ApEn values and other measures of RI–1 dispersion, although none correlated significantly with any value from the ApEn set (these measures included the proportional Shannon entropy derived from all RI–1 intervals in each epoch rather than the median of the 10-breath moving window; RI–1 standard deviation; mean RI–1 consecutive difference; and standard deviation RI–1 consecutive difference).

(5) ApEn measures the log likelihood that short epochs of data with a similar pattern will remain similar as the epoch length is increased. As coupling is associated with repetition of a specific pattern, a statistic which examines the likelihood that a pattern is similar might be a better correlate with coupling. ApEn may therefore be measuring a quality of RR variability different from that resulting from coupling.

In conscious subjects, Yeragani and colleagues20 observed that DF was positively correlated to both ApEn and high-frequency spectral power, suggesting that both measures reflected parasympathetic modulation of heart rate, and on this basis it was suggested that the two measures (ApEn and DF) could be used interchangeably. Similar results were reported by Jartti and colleagues.8 In the present study we found high correlations between DF and ApEn, as suggested by Yeragani and colleagues, at values of rEn>=0.05, although, as noted above, at lower values of r an inverse relationship is seen. Both DF and ApEn strongly correlated with high-frequency power and therefore it is probable that the dominant influence on each of these measures is RSA. As with ApEn, we observed no relationship between DF and cardioventilatory coupling.

In conclusion, using standard spectral and non-linear measures of HRV we were unable to demonstrate an effect of cardioventilatory coupling on heart rate time series. However, graphical displays of clinical time series plainly show that, during periods of coupling, the heart rate time series may take on a qualitative appearance which is different from that observed in uncoupled time series. Alternative analytical techniques are therefore desirable for the detection of coupling in heart rate time series, and these are explored in the accompanying paper.21


    Appendix
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
Measures of cardioventilatory coupling and HRV
Proportional RI–1 Shannon entropy (HRI–1)
The RI–1 interval dispersion was measured using proportional Shannon entropy of the RI–1 interval (HRI–1). Starting from the beginning of each RI–1 interval time series, we used a 10-breath moving window and placed the corresponding RI–1 intervals within each window into a 10-bin histogram with outer limits set at 0 and the mean of the RR intervals which had been encompassed by the 10 inspiratory onsets. From the bin occupancy of this histogram, we calculated HRI–1 as follows:

N

Shannon entropy=H=–{Sigma} Pb*log(Pb)

b=1

Maximum Shannon entropy=Hmax=–log(1/N)

HRI–1=H/Hmax

where P is the actual histogram bin probability, b is the bin number and N is the number of histogram bins.

The median value from successive moving windows over the entire RI–1 time series was used as a measure of RI–1 dispersion. Values below 0.7 are typically associated with the appearance of cardioventilatory coupling in RI–1 time series plots.14

Spectral analysis
To determine the effect of cardioventilatory coupling on spectral measures of HRV, we extracted epochs of data, 256 s in length, displaying as nearly as possible a constant single pattern of coupling throughout.

Spectral analysis was performed using a method based on that of Akselrod and colleagues.1 15 The 256-s RR interval time series, free of ectopy, was sampled at the rate of 4 Hz to produce a series of 1024 discrete RR interval values. This series was then high-pass filtered to remove fluctuations less than 0.015 Hz and low-pass filtered at 2 Hz to remove components at greater than Nyquist frequency. A fast Fourier analysis was performed using a Hanning window. The spectral power (integrated area under the power spectrum curve) was calculated in each subject for total power and in three frequency bands: high (0.15–0.45 Hz), low (0.08–0.15 Hz) and very low (0.02–0.08 Hz). The proportion of the total power in each frequency band was also calculated.

Approximate entropy (ApEn)
ApEn was calculated using the algorithm of Pincus.5 Starting from the first RR interval, a segment or window of m heart periods is taken, these forming a short sequence of RR interval variation. The entire RR time series is then examined for sequences with similar values (within a tolerance of rEn) and the number of similar sequences is counted. The window is then moved to the next consecutive RR interval and the number of similar sequences to this is then computed. This procedure is repeated as the window moves across the time series. The resulting count is then compared with a similar count for the sequence value (m+1). In this manner, ApEn measures the log likelihood that runs of data with a particular pattern will retain the same pattern for the next incremental observation. To compute ApEn, three input values must be decided according to the variables: (1) the number of data points (N) to be analysed, in this study N=500 heart beats, as discussed above; (2) the ‘embedding dimension’ (sequence length m) for comparison (for heart rate time series, m=2 and m=3 are the most commonly used embedding dimensions)6 7 16 and in the present study ApEn values were computed for both m=2 and m=3; (3) the tolerance range (rEn) within which sequences of m RR intervals are considered to be similar. Values of rEn are typically expressed as a fraction of the standard deviation of the RR interval time series and values in the range 0.05–0.25 are generally used.57 16 However, to prevent missing a correlation with coupling resulting from selecting inappropriate rEn values, we chose to use a range of r values wider than is normally used, selecting rEn=0.025, 0.05, 0.075, 0.1, 0.15, 0.2, 0.3, 0.4, 0.5 and 0.6 of the standard deviation of the RR interval time series. Thus, for our 500-beat time series, ApEn was calculated for 20 combinations of m and rEn.

Fractal dimension (DF)
DF was calculated using the method of Katz:3

DF=log(n)/[log(n) + log(d/L)]

where n is the number of data points, d is the planar extent of the curve (the ‘distance’ in time from the starting point to the nth point) and L is the total length of the curve. This algorithm generates a value for DF between 1 (representing a signal with low space-filling capacity) and 2 (representing a signal with high space-filling capacity).


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Appendix
 References
 
1 Akselrod S, Gordon D, Ubel F, Shannon D, Barger A, Cohen R. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat to beat cardiovascular control. Science 1981; 213: 220–2[ISI][Medline]

2 van Ravenswaaij-Arts C, Kollee L, Hopman J, Stoelinga G, van Geijn H. Heart rate variability. Ann Intern Med 1993; 118: 436–47[Abstract/Free Full Text]

3 Katz MJ. Fractals and the analysis of waveforms. Comp Biol Med 1988; 18: 145–56[ISI][Medline]

4 Mansier P, Clairambault J, Charlotte N, et al. Linear and non-linear analyses of heart rate variability: a minireview. Circ Res 1996; 31: 371–9

5 Pincus SM. Approximate entropy as a measure of system complexity. Proc Natl Acad Sci USA 1991; 88: 2297–301[Abstract]

6 Pincus SM, Goldberger AL. Physiological time-series analysis: what does regularity quantify? Am J Physiol 1994; 266: H1643–56[Abstract/Free Full Text]

7 Fleisher LA, Frank SM, Sessler DI, Cheng C, Matsukawa T, Vannier CA. Thermoregulation and heart rate variability. Clin Sci 1996; 90: 97–103[ISI][Medline]

8 Jartti TT, Kuusela TA, Kaila TJ, Tahvanainen KUO, Valimaki IAT. The dose–response effects of terbutaline on the variability, approximate entropy and fractal dimension of heart rate and blood pressure. Br J Clin Pharmacol 1998; 45: 277–85[ISI][Medline]

9 Galletly DC, Larsen PD. Cardioventilatory coupling during anaesthesia. Br J Anaesth 1997; 79: 35–40[Abstract/Free Full Text]

10 Galletly DC, Larsen PD. Ventilatory frequency variability in spontaneously breathing anaesthetized subjects. Br J Anaesth 1999; 83: 552–63[Abstract/Free Full Text]

11 Galletly DC, Larsen PD. Inspiratory timing during anaesthesia; a model of cardioventilatory coupling. Br J Anaesth 2001; 86: 777–88[Abstract/Free Full Text]

12 Galletly DC, Larsen PD. The determination of cardioventilatory coupling from heart rate and ventilatory time series. Res Exp Med 1999; 199: 95–100[ISI][Medline]

13 Galletly DC, Larsen PD. Relationship between cardioventilatory coupling and respiratory sinus arrhythmia. Br J Anaesth 1998; 80: 164–8[ISI][Medline]

14 Larsen PD, Trent EL, Galletly DC. Cardioventilatory coupling: the effect of IPPV. Br J Anaesth 1999; 82: 546–50[Abstract/Free Full Text]

15 Galletly D, Corfiatis T, Westenberg A, Robinson B. Heart rate periodicities during induction of propofol–nitrous oxide– isoflurane anaesthesia. Br J Anaesth 1992; 68: 360–4[Abstract]

16 Palazzolo JA, Estafanous FG, Murray PA. Entropy measures of heart rate variation in conscious dogs. Am J Physiol 1998; 274: H1099–105[Abstract/Free Full Text]

17 Hirsch JA, Bishop B. Respiratory sinus arrhythmia in humans: how breathing pattern modulates heart rate. Am J Physiol 1981; 241: H620–9[Abstract/Free Full Text]

18 Aarimaa T, Oja R, Antila K, Valimaki I. Interaction of heart rate and respiration in new born babies. Pediatr Res 1988; 24: 745–50[Abstract]

19 Kluge KA, Harper RM, Schechtman VM, Wilson AJ, Hoffman HJ, Southall DP. Spectral analysis assessment of respiratory sinus arrhythmia in normal infants and infants who subsequently died of sudden infant death syndrome. Pediatr Res 1988; 24: 677–82[Abstract]

20 Yeragani VK, Sobolewski E, Jampala VC, Kay J, Yeragani S, Igel G. Fractal dimension and approximate entropy of heart period and heart rate: awake versus sleep differences and methodological issues. Clin Sci 1998; 95: 295–301[ISI][Medline]

21 Galletly DC, Larsen PD. Cardioventilatory coupling in heart rate variability: methods for qualitative and quantitative deter mination. Br J Anaesth 2001; 87: 827–33[Abstract/Free Full Text]