doi:10.1152/japplphysiol.01424.2010 112:1248-1257, 2012. First published 19 January 2012;J Appl Physiol

Yeganeh, Mikkel Fishman, Christopher G. Wilson and Kingman P. StrohlLee Friedman, Thomas E. Dick, Frank J. Jacono, Kenneth A. Loparo, AmirsubjectsCardio-ventilatory coupling in young healthy resting

You might find this additional info useful...

41 articles, 23 of which can be accessed free at:This article cites /content/112/8/1248.full.html#ref-list-1

1 other HighWire hosted articlesThis article has been cited by

  [PDF] [Full Text] [Abstract]

, March 1, 2014; 306 (5): H755-H763.Am J Physiol Heart Circ PhysiolSungwoo Ahn, Jessica Solfest and Leonid L. RubchinskyFine temporal structure of cardiorespiratory synchronization

including high resolution figures, can be found at:Updated information and services /content/112/8/1248.full.html

can be found at:Journal of Applied Physiologyabout Additional material and information http://www.the-aps.org/publications/jappl

This information is current as of December 22, 2014. 

ISSN: 0363-6143, ESSN: 1522-1563. Visit our website at http://www.the-aps.org/.Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2012 by the American Physiological Society.those papers emphasizing adaptive and integrative mechanisms. It is published 12 times a year (monthly) by the American

publishes original papers that deal with diverse areas of research in applied physiology, especiallyJournal of Applied Physiology

on Decem

ber 22, 2014D

ownloaded from

on Decem

ber 22, 2014D

ownloaded from

Cardio-ventilatory coupling in young healthy resting subjects

Lee Friedman,1,2 Thomas E. Dick,1,3 Frank J. Jacono,1,2 Kenneth A. Loparo,4 Amir Yeganeh,2

Mikkel Fishman,1 Christopher G. Wilson,3 and Kingman P. Strohl1,2

1Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, Case Western Reserve University,Cleveland; 2Louis Stokes Cleveland Veterans Affairs Medical Center, Cleveland; Departments of 3Neurosciences and4Electrical Engineering and Computer Science, Case Western Reserve University, Cleveland, Ohio

Submitted 9 December 2010; accepted in final form 18 January 2012

Friedman L, Dick TE, Jacono FJ, Loparo KA, Yeganeh A, FishmanM, Wilson CG, Strohl KP. Cardio-ventilatory coupling in young healthyresting subjects. J Appl Physiol 112: 1248–1257, 2012. First publishedJanuary 19, 2012; doi:10.1152/japplphysiol.01424.2010.—In this work,cardio-ventilatory coupling (CVC) refers to the statistical relationshipbetween the onset of either inspiration (I) or expiration (E) and thetiming of heartbeats (R-waves) before and after these respiratoryevents. CVC was assessed in healthy, young (�45 yr), resting, supinesubjects (n � 19). Four intervals were analyzed: time from I-onset toboth the prior R-wave (R-to-I) and the following R-wave (I-to-R), aswell as time from E-onset to both the prior R-wave (R-to-E) andfollowing R-wave (E-to-R). The degree of coupling was quantified interms of transformed relative Shannon entropy (tRSE), and �2 testsbased on histograms of interval times from 200 breaths. Subjects werestudied twice, from 5 to 27 days apart, and the test-retest reliability ofCVC measures was computed. Several factors pointed to the relativeimportance of the R-to-I interval compared with other intervals.Coupling was significantly stronger for the R-to-I interval, couplingreliability was largest for the R-to-I interval, and only tRSE for theR-to-I interval was correlated with height, weight, and body surfacearea. The high test-retest reliability for CVC in the R-to-I intervalprovides support for the hypothesis that CVC strength is a subjecttrait. Across subjects, a peak �138 ms prior to I-onset was charac-teristic of CVC in the R-to-I interval, although individual subjects alsohad earlier peaks (longer R-to-I intervals). CVC for the R-to-I intervalwas unrelated to two separate measures of respiratory sinus arrhyth-mia (RSA), suggesting that these two forms of coupling (CVC andRSA) are independent.

Shannon entropy; respiratory sinus arrhythmia; heart rate variability

CARDIO-VENTILATORY COUPLING (CVC) can refer to several con-cepts. Specifically, CVC can refer to an increase in heart rateduring inspiration (5, 10, 20, 21, 26, 40), as well as to atendency for heartbeats to occur at particular times relative tothe onsets of inspiration (I) or expiration (E). In this report, weuse the term CVC in this latter sense and use the termrespiratory sinus arrhythmia (RSA) to refer to CVC in theformer sense.

Coleman (6) noted that animals tend to have a stable numberof heartbeats in each breath, i.e., a whole number ratio. Galli(14) noted the same thing in patients with heart failure. Weissand Salzano (41) also reported whole number ratios betweenheartbeats and the respiratory cycle in anesthetized dogs. Morerecently, this basic relationship between heartbeats and breathshas been observed in healthy athletes (32) and anesthetizedpatients undergoing elective surgery (25).

The presence of whole number ratios between heartbeats andbreaths implies CVC in the context of heartbeats occurring atspecific times in the respiratory cycle. Indeed, as early as thelate 1960s, several German researchers identified CVC in thata heartbeat occurred preferentially before I-onset in anesthe-tized as well as awake, healthy subjects (11, 18, 19). Theseobservations were confirmed and extended in the 1990s and2000s by Galletly, Larsen, and coworkers (12, 38). Typically,these researchers defined the time of a heartbeat from theR-wave peak in the electrocardiogram (ECG) and found thatheartbeats have a tendency to occur �500 ms before I-onset(12, 38).

Prior reports have not compared CVC for all four of thefollowing intervals: R-wave occurrence to I-onset (R-to-I),R-wave to E-onset (R-to-E), I-onset to R-wave (I-to-R), andE-onset to R-wave (E-to-R). We assessed and quantified CVCwith respect to I-onset and E-onset, determining the strength ofCVC for intervals before and after respiratory phase transi-tions. These comparisons may provide further insight into thenature, interpretation, and implications of CVC.

We note that there is no universally agreed on definition ofcoupling. In the present manuscript we use the term couplingas it has been used by Galletly, Larsen, and coworkers. In thissense, evidence for coupling emerges out of an analysis ofhistograms for various intervals between heartbeats and I- andE-onsets. Evidence of a statistically significant departure fromchance in these histograms is evidence of coupling by thisdefinition. As such, this definition makes no claims of causal-ity. Causality, i.e., does the heartbeat facilitate respiratoryphase transition, would be addressed in a study with a differentdesign.

We hypothesized that CVC is a phenotypic trait in normal,healthy human subjects. The concept is that the presence andstrength of CVC will vary from individual to individual butwill be consistent within an individual. The test-retest reliabil-ity of CVC intensity will provide insight into the extent towhich CVC is a subject trait or is influenced by occasionvariance.

Finally, we were also interested in relating the intensity ofCVC to the strength of RSA. Although others have indicatedno correlation (22, 23, 32, 38), this is a fundamental issuebecause both RSA and CVC have been proposed to support acommon goal of efficient gas exchange.

Thus, in the present report, we describe analytical toolsfor quantifying CVC, determine the test-retest reliabilityand the strength of CVC for all four types of intervals(R-to-I, R-to-E, I-to-R, E-to-R), and compare the relation-ship between CVC intensity and RSA strength in young,healthy subjects at rest.

Address for reprint requests and other correspondence: L. Friedman, Divi-sion of Pulmonary, Critical Care and Sleep Medicine, Dept. of Medicine, CaseWestern Reserve Univ., 10900 Euclid Ave., Cleveland, OH 44106 (e-mail:lfriedman10@gmail.com).

J Appl Physiol 112: 1248–1257, 2012.First published January 19, 2012; doi:10.1152/japplphysiol.01424.2010.

http://www.jappl.org1248

on Decem

ber 22, 2014D

ownloaded from

METHODS

Subjects

Young, adult (�45 yr), healthy subjects were recruited fromlaboratory personnel, medical staff, and/or their relatives and associ-ates (n � 26). Of these, only 19 subjects were included because onedid not have 200 breaths in the recording period (minimum 200breaths required for our analysis); one had premature ventricularcontractions during the recording period; one had poor recordingquality; and four did not return for a second study (see below). Thefinal group of 19 subjects (8 men, 11 women) had a mean age of 28.5yr (range: 17–43 yr), no history of major medical illness, and nocurrent prescription medications. Subjects were studied on two occa-sions (from 5 to 27 days apart, median: 7 days). Each subject’s visitwas considered a study. The median height and weight of the subjectswas 170.2 cm (range: 154.9–185.4 cm) and 72.6 kg (range: 45.4–97.5kg). The median body mass index (BMI) of the subjects was 23.7(range: 18.8–35.8). Three of the 19 subjects had BMI values over 30(30.8, 32.2, and 35.8). Body surface area (BSA) was calculatedaccording to the Mosteller formula (28).

The study was approved by the Institutional Review Board of theLouis Stokes Veterans Affairs Medical Center and all subjects gavewritten informed consent.

Data Acquisition

Studies were conducted during the day in a quiet room at theCleveland Veterans Affairs Medical Center at Wade Park, Cleveland,OH. Each study lasted �20 min (median, range: 9.6 to 24.9 min). Tomeasure respiratory flow, a pneumotachograph (Validyne) attached toa conventional plastic mouthpiece plus filter was suspended on aswing arm over the hospital bed and the subject adjusted the bed andswing arm so that the fit of the breathing device was comfortable in asemi-reclining position. The output of the Validyne demodulator wasfiltered (low-pass cutoff frequency of 300 Hz) and amplified (modelUSBPGF-S1, Alligator Technologies, Costa Mesa, CA). The ECGwas recorded using a modified V5 lead placement, with band-passfiltering (0.3 to 300 Hz) and amplification (Grass model QP-511, WestWarwick, RI). The respiratory flow and ECG signals were sampled at2,000 Hz using LabView software and a National Instruments A/Dboard (Austin, TX). The signals were stored for offline analysis.

Digital Signal Preprocessing

Both the flow and ECG signals were low-pass filtered with azero-phase (forward-reverse) digital filter (using the MATLAB “filt-filt” function, MathWorks, Natick, MA) with a 3-dB cutoff at 34 Hz.The signals were then decimated to 1 kHz. Careful visual inspectionof the signals before and after digital filtering revealed that theessential components of each signal were preserved after digitalfiltering without phase shifts. The amplitude of the ECG signal at theR-wave peak was slightly reduced by the digital filter, but the time ofthe peak was preserved within 1 sample (1 ms). Furthermore, in caseswith 60-Hz noise, the filtered signal was an improvement over the rawsignal as a basis for detection of the peak of the R-wave.

Event Detection

Custom algorithms were created in MATLAB for event detection.For the detection of I- and E-onsets, the flow signal obtained from thepneumotachograph was integrated to obtain volume. The troughs andpeaks of the volume curve were then used as indicators of the onsetsof inspiration and expiration, respectively. However, each mark forthe I- and E-onsets was inspected visually. This labor-intensive stepwas especially important because the minimum volume point may ormay not be the start of I and the maximum volume point may or maynot be the start of E. Both the volume and flow signals were inspectedfor every breath to ensure that the chosen point was either the start of

inspiration or the start of expiration. The placement of each could beedited manually based on this assessment. The detection of the peakof each R-wave was based on both velocity (1st derivative) andamplitude criteria. We employed software for evaluating the accuracyof R-wave detection, with user-editable event times and examinationof short or long R-to-R intervals.

Assessment of CVC

As a result of the abovementioned event detection steps, the times,in seconds (ms resolution) for each I-onset, E-onset, and R-wave peakwere used as inputs to the coupling analysis. To achieve uniformity ofstatistical power for each subject and each visit, exactly 200 consec-utive breaths were selected for analysis. These breaths were either theset that had the lowest variability for the I-to-I time intervals or the setthat had the highest variability for the I-to-I time intervals. For eachsubject (n � 19), for each visit (n � 2), the following four intervalswere evaluated: 1) R-to-I: the time from I-onset to the previousR-wave; 2) I-to-R: the time from I-onset to the next R-wave;3) R-to-E: the time from E-onset to the previous R-wave; and4) E-to-R: the time from each E-onset to the next R-wave.

Assessment of Temporal Autocorrelation of Event Intervals

For each subject (n � 19), for each visit (n � 2), there were 200each of R-to-I, I-to-R, R-to-E, and E-to-R intervals. Interval lengthwas regressed onto interval number (1�200), and the residuals fromthis regression were tested for temporal autocorrelation using theDurbin-Watson (DW) test. A DW statistic of 2.0 indicates the absenceof temporal autocorrelation. In the present case, a DW statistic�1.758 indicated positive temporal autocorrelation, and �2.242 in-dicated negative temporal autocorrelation.

Histograms

Histograms of the four interval types were constructed. The as-sumption of serial independence of observations was met becausetemporal autocorrelation was absent in these intervals. We employedtwo binning schemes (Fig. 1), one for each of the two couplingmeasures: 1) transformed relative Shannon entropy (tRSE); (2)�log10(P) from a �2 test. We computed the histograms expected bychance. For tRSE, Shannon entropy was computed for each histogramfor only the portion of the chance distribution that was flat (up to thelower 0.25% of the R-R distribution; Fig. 1A). This portion wasdivided into 20 bins (Fig. 1C). Shannon entropy was converted torelative Shannon entropy (RSE) by dividing by the maximum Shan-non entropy [�log(1/k), where k � number of bins]. Because thismeasure is censored at 1.0, has values clustered very close to 1.0, anddecreases as coupling strength increases, RSE was transformed totRSE as tRSE � (�100*RSE)�100; tRSE ranges from 0 to 100, with0 indicating no coupling and 100 indicating extreme coupling.

An RSE of 0.95 would yield a tRSE of 5 and an RSE of 0.9 wouldyield a tRSE of 10. This measure had a gamma distribution that wasamenable to parametric statistical analysis using a generalized linearmodel. This measure was based on a subset of the entire potentialchance distribution (the flat part), so only a subset of the available 200intervals was used in the analysis (n � 200 intervals). The number ofintervals varied slightly from subject to subject.

The second coupling measure was based on a �2 goodness-of-fittest, using all the data (up to 99.75% of the R-R distribution). This fulldata range was divided into 27 bins (Fig. 1D). However, to keep theexpected counts �5, bins 23–27 were combined into a single 23rd binfor �2 analysis. The negative log (base 10) of the P value from the �2

goodness-of-fit test was computed and will be referred to as�log10(P). A P value of 0.01 would yield a �log10(P) of 2, and a Pvalue of 0.0001 would yield a �log10(P) value of 4. Thus, a largevalue for �log10(P) is strong evidence for coupling. This analysisused all of the 200 intervals.

1249Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

Assessment of Respiratory Sinus Arrhythmia

Direct respiratory sinus arrhythmia. For each I-onset, we measuredthe two preceding R-to-R intervals (RR-2, RR-1), the two followingR-to-R intervals (RR�1, RR�2), and the R-to-R interval that spannedI-onset (RR_Insp). The means for each of these intervals werecalculated and used to create an index of the magnitude of respiratorysinus arrhythmia (RSA) consisting of the average of RR�1 andRR�2 divided by the average of RR-1 and RR_Insp.

Fourier-based RSA. To compute a Fourier measure of RSA fromthe R-to-R intervals, we followed the general guidelines from the TaskForce of the European Society of Cardiology and the North AmericanSociety of Pacing Electrophysiology (36) and the specific approachoutlined by Singh et al. (34). We took exactly 8 min of ECG data. TheR-to-R time series was interpolated to 4-Hz resolution. Next, wesubtracted the mean of the time series and divided the time series bythe mean. The data were then windowed using a Hann Window andthen calibrated so that total power in the spectrum would equal thetotal variance in the time series. Our measure of RSA was the total

power in the high frequency (HF-RSA) bin that ranged from 0.15 to0.40 Hz (9 to 24 breaths/min).

Statistical Analysis

The distributions of tRSE values were non-normal and were bestapproximated by a gamma distribution. Similarly, the distributions of�log10(P) values were generally skewed and were best approximatedby a negative binomial distribution. All ANOVA models involvingthese dependent variables were analyzed with generalized linearmixed model ANOVAs (Proc GLIMMIX, SAS, Cary, NC) using thebase distributions noted above. These ANOVA models produce meanestimates for each coupling measure and tests that these mean esti-mates were non-zero. These tests will be referred to as “tests fornon-zero mean estimates”. Post hoc tests were based on the “analysisof means” concept and employed a “Nelson-Hsu” adjustment formultiple comparisons (SAS). All correlation analyses involved eitherPearson coefficients or Spearman rank order correlation coefficients.Autocorrelation analysis involved detrending the time series and

Fig. 1. Illustration of the steps involved in setting up the binning schemes. The first step is to create an R-to-R interval histogram (A). Next, a cumulative densitydistribution is created from 0 to the 99.75 percentile of the R-to-R interval histogram (B). Then this distribution was inverted and normalized to a maximum of1.0 (C). This final distribution is the expected distribution of R-to-I intervals (or any interval between an R wave and the immediately following or precedingI- or E-onset) assuming no coupling (chance). For transformed Relative Shannon Entropy (tRSE), the distribution from 0 to the 0.25 percentile of the R-Rdistribution (the flat part of the chance distribution) was divided into 20 bins (C). For the �log10(P) analysis, almost the entire distribution (up to 99.75 percentileof the R-R distribution) was divided into 27 equal bins (D). Because the sum of the expected counts in bins 23–27 was often quite small, these bins were combinedinto a single 23rd bin.

1250 Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

computing the DW statistic on the residuals. Test-retest reliability wasassessed with intraclass correlation coefficients (ICC) that were esti-mated directly from an analysis of variance table (33). The ICC is anumber that ranges from 0.0 to 1.0, and Shrout and Fleiss (33) definesix types. We employed type ICC(1,1) for an assessment of test-retestreliability in the absence of visit effects and type ICC(3,1) in thepresence of visit effects. Cicchetti and Sparrow (3, 4) presentedguidelines for interpretation of ICCs as follows: poor (below 0.40),fair (0.41–0.59), good (0.60–0.74), and excellent (above 0.75). R-to-R intervals for the RSA analysis were evaluated using a mixed-model repeated-measures ANCOVA (SAS Proc Mixed, SAS) withsubjects treated as random and interval type and visit treated as fixedeffects. Median heart rate was a covariate in this analysis.

RESULTS

Comparison of the Least to Most Variable 200I-to-I Intervals

Coupling measures for the set of 200 consecutive breathswith the least variable respiratory rate were compared with themeasures with the most variable respiratory rate. For our dataset, respiratory rate variability did not affect the tRSE and�log10(P) coupling measures [tRSE: F � 0.04, df � 1,18, P �0.8360; �log10(P): F � 0.17, df � 1, 18, P � 0.68]. Henceforth,all results will be reported for the least variable 200 consecutivebreaths.

Temporal Autocorrelation of Intervals

The proportions of significant values for positive and nega-tive autocorrelation (DW statistics) were calculated and com-pared with the proportion expected by chance. This was donefor each interval type separately as well as all intervals takentogether. In no case did the proportion of significant DWstatistics differ significantly from that which would be ex-pected by chance (0.05). Therefore, temporal autocorrelationwas not present in these intervals.

Comparison of the Two Coupling Measures

tRSE and �log10(P) were significantly correlated to a mod-est degree (Pearson r � 0.66, n � 152, P � 0.001; Spearmanr � 0.60, n � 152, P � 0.001). Although these two couplingmeasures were correlated, one measure could not replace theother because at best one accounted for only �45% of theother’s variance.

Sample Interval Histograms

To illustrate our analysis, histograms for two subjects (#12and #20) with strong CVC are presented in Fig. 2. (Resultsfrom the analysis of the full dataset are given below.) Table 1contains the tRSE values for the data in Fig. 2. Table 2 containsthe �log10(P) measures, and Table 3, the median heart rate andrespiratory rate for the data presented in Fig. 2. The heart andrespiratory rates were within normal limits. However, respira-tory rates appeared to have increased slightly on visit 2. Thiswas found in the complete data set as well and is discussedbelow.

For subject #12, visit 1, the R-to-I histogram had a maximalvalue at the bin centered at 116 ms (Fig. 2, arrow a) and asecondary peak at 347 ms (Fig. 2, arrow b). The tRSE and the�log10(P) for this histogram were robust (Tables 1 and 2,respectively). The tRSE was greatest for the R-to-I interval

compared with the other intervals, although the �log10(P)indicated significant coupling at several other intervals.

On visit 2, coupling for the R-to-I interval was again strongfor both measures [tRSE and �log10(P)]. Furthermore, theprimary peak in the R-to-I histogram also occurred in the bincentered at 116 ms (Fig. 2, arrow c). There was a second peakat 429 ms (not marked) and another at 611 ms (Fig. 2, arrowd). Thus, for subject 12, the R-to-I interval showed robustcoupling on both test days.

For subject 20, for the R-to-I interval, the coupling measureswere modest on visit 1 but were very strong on visit 2.Nonetheless, the R-to-I interval histograms had similar peakson visits 1 and 2. On visit 1, peaks occurred at 118 ms (Fig. 2,arrow e) and �588 ms (Fig. 2, arrow f) before inspiration. Onvisit 2, the primary peak occurred at 118 ms again (Fig. 2,arrow g), but the later peak occurred around 878 ms (Fig. 2,arrow h).

Composite 27-Bin Histograms

Composite (or cumulative) 27-bin histograms for the fourintervals were based on a sum across all subjects and visits(Fig. 3). All histograms, regardless of statistical significance,contributed to this composite. The most striking features werea peak at 138.5 ms and a trough at 217.7 ms before I-onset(Fig. 3, top). For the I-to-R histogram, a peak appeared at�811 ms. The time from the main peak in the R-to-I histogramto the late peak in the I-to-R histogram (811�138.5 � 949 ms)was almost exactly the same as the mean R-R interval spanninginspiration (945 ms, see below). Both of the expiration-relatedhistograms (R-to-E and E-to-R) began to roll off toward zeromore quickly than the inspiration-related intervals. This indi-cates that long R-to-E and E-to-R intervals occurred lessfrequently than for R-to-I or I-to-R intervals. There does seemto be a small increased likelihood of a heartbeat immediatelyfollowing the phase transition to expiration as well.

Fig. 2. The 27-bin interval histograms for 2 subjects (#12 and #20) showingstrong CVC (1 subject per row). The first 2 columns present the data from thefirst visit and the second 2 columns present the data from the second visit. Thehistograms are centered on either onset of inspiration (I-onset; columns 1 and3) or expiration (E-onset; columns 2 and 4). I-onset and E-onset occur at time0, and 27 bins are shown before and after time 0. The number below eachhistogram is the bin width in ms. The alphabet marks and associated arrows(e.g., a, f) are discussed in Sample Interval Histograms.

1251Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

Visit Effects

There were no statistically significant differences betweenvisits for heart rate (F � 1.48, df � 1,18, P � 0.24, 2-tailed).However, respiratory rate increased slightly from visit 1 to 2(mean: from 15.9 to 17.3 breaths/min) (F � 6.6, df� 1,18,P � 0.02). For the coupling measures [tRSE and �log10(P)],visit effects were tested as part of larger models includinginterval type and the interaction between visit and intervaltype. There were no statistically significant visit effects orvisit-by-interval type interactions for the two coupling mea-sures (smallest P value was 0.15).

tRSE by Interval Type

Histograms of tRSE values and gamma distribution fits aredisplayed by interval type in Fig. 4. The highest mean tRSEwas for the R-to-I interval followed by the I-to-R interval. Themean tRSE for the E-onset-related intervals was lower thanthe mean tRSE for the I-onset-related intervals. The F-test forthe interval effect from the generalized linear mixed modelANOVA was statistically significant (F � 7.84, df�3, 54, P �0.0002). Table 4 provides the tests for non-zero mean esti-mates, as well as effect size estimates for each test. Only themean estimate for the R-to-I interval was significantly differentfrom zero, and this effect size was large. Post hoc analysis ofmeans indicated that the R-to-I interval was statistically larger(more coupled) than the overall mean across intervals (P value,adjusted for multiple comparisons � 0.0019) and that the R-to-E interval was significantly smaller (less coupled, P value,adjusted for multiple comparisons � 0.0094) than the overallmean.

�log10(P) by Interval Type

Histograms of �log10(P) values and negative binomial fitsare displayed by interval type in Fig. 5. The highest mean�log10(P) was for the R-to-I interval followed by the I-to-Rinterval. The F-test for the interval effect from the generalizedlinear mixed model ANOVA was statistically significant (F �3.01, df � 3,54, P � 0.04). Table 4 provides the tests fornon-zero mean estimates, as well as effect size estimates foreach test. All of the mean estimates were significantly largerthan 0, indicating significant coupling for all intervals. The

largest mean, with the largest effect size, was for the R-to-Iinterval. Post hoc analysis of means indicated that the R-to-Iinterval was statistically larger (more coupled) than the overallmean across intervals (P value, adjusted for multiple compar-isons �0.03). At the P � 0.05 level [�log10(0.05)�1.301], forthe R-to-I, I-to-R, R-to-E, and E-to-R intervals, respectively,the number of statistically significant �2 were 18, 20, 17, and13 (of a total of 38). At the P � 0.001 level, the numbers ofsignificant coupling studies was 9, 5, 3, and 4, respectively.

Test-Retest Reliability of tRSE

The intra-class correlation coefficients (ICCs) were deter-mined for each of the four intervals for both coupling mea-sures. The ICC for tRSE for the R-to-I interval was in the“good” range (0.69) and was significantly greater than zero(95% confidence limits do not include 0). The ICC for tRSE forthe I-to-R interval was in the “fair” range (0.49) and was alsosignificantly greater than 0. The ICCs for the expiration-relatedintervals were not significantly different from 0.

For �log10(P) measures, the ICC for the R-to-I interval wasin the “excellent” (0.83) range and significantly greater thanzero and significantly larger than the other interval types. TheICCs for the remaining intervals were not statistically signifi-cantly different from 0.

Test-Retest Reliability of Heart Rate and Respiration Rate

The ICC for median heart rate was 0.65 (good). For therespiration rate, we used ICC(3,1) rather than ICC(1,1) (seeabove) due to the presence of a visit effect for this measure.This ICC was 0.73 (top of the good range).

Correlations Between Coupling Measures and Heart Rate orRespiratory Rate

Spearman correlation coefficients were calculated, relatingthe mean tRSE and mean �log10(P) (across visits) for intervaltypes and either mean heart rate or mean respiratory rate(2*4*2 � 16 correlation coefficients). Two correlations werestatistically significant before but not after correction for mul-tiple comparisons. These are mean tRSE for the E-to-R intervalwas positively correlated with the median respiratory rate(0.46, P � 0.04) and mean �log10(P) measure for the E-to-R

Table 1. tRSE and subject ranks across interval types for the histograms displayed in Fig. 2

Subject #

Visit 1 Visit 2

R-to-I I-to-R R-to-E E-to-R R-to-I I-to-R R-to-E E-to-R

12 9.24 (19) 3.83 (14) 2.55 (15) 1.92 (7) 8.43 (19) 4.58 (18) 2.32 (12) 2.55 (14)20 4.40 (15) 4.09 (15) 3.03 (17) 2.08 (9) 6.62 (18) 3.48 (14) 2.71 (14) 2.55 (7)

Rank within the set of 19 subjects per visit shown in parentheses. A high rank means a high transformed relative Shannon entropy (tRSE).

Table 2. �log10(P) and subject ranks across interval types for the histograms displayed in Fig. 2

Subject #

Visit 1 Visit 2

R-to-I I-to-R R-to-E E-to-R R-to-I I-to-R R-to-E E-to-R

12 13.45 (18)‡ 3.42 (17)‡ 2.16 (12)† 1.30 (12)* 10.89 (18)‡ 2.77 (18)‡ 0.40 (5) 0.35 (6)20 1.64 (13)* 1.99 (9)* 1.69 (11)† 0.56 (7) 6.60 (17)‡ 2.96 (19)‡ 0.43 (6) 1.23 (13)†

Rank within the set of 19 subjects per visit shown in parentheses. A high rank means a high �log10(P), which is an index of strong cardio-ventilatory coupling(CVC). *P � 0.05; †P � 0.010; ‡P � 0.001.

1252 Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

interval was positively correlated with the median heart rate(0.49, P � 0.04). Thus, the coupling of R-to-I intervals was notdependent on heart or respiratory rates.

Relationship Between Coupling and Demographic Measures

Mean tRSE for the R-to-I interval was highly correlated withheight, weight, and especially BSA (Table 5; Fig. 6). Couplingwas stronger for larger subjects (taller, heavier, and largerBSA). Coupling strength was not correlated with BMI. Bodysize was confounded with sex: of 9 subjects in the small group(Fig. 6, bottom left), 8 were women. Given differences inheight and weight, sex differences would have to be comparedafter controlling for height or weight, but this analysis was notconducted given the modest sample size (and low statisticalpower) to detect an effect. Other correlations with the otherintervals were not statistically significant (Table 5). There wasno correlation between age and either coupling measure for thisage-restricted (�45 yr) sample.

Direct RSA Analysis

The means and standard errors for the five R-to-R intervaltypes (RR-2, RR-1, RR_Insp, RR�1, and RR�2) were calcu-lated across subjects and visits (Fig. 7). These R-to-R intervalswere significantly different (F � 49.8, df � 4, 144, P �0.0001). RSA consisted of shorter RR intervals in inspiration(RR�1 and RR�2) relative to the RR intervals beginningduring expiration (RR-1 and RR_Insp). The difference be-tween the means of these two sets of intervals was 149.5 msand was statistically significant (t � 12.9, df � 144, P �0.0001). (RR_Insp was 170 ms and RR-1 was 130 ms longerthan the mean of RR�1 and RR�2.) The mean of our directmeasure of RSA intensity across subjects and visits was 0.92(� 0.05 SD) and was significantly lower than 1.0 (t � �7.26,df � 18, P � 0.0001). Our measure was also very reliable (ICC �0.90) across visits.

Relationship Between Direct RSA and HF Power

To validate our direct measure of RSA, we compared therank order of direct RSA with the rank order of high-frequencypower from a traditional fast Fourier transform analysis ofheart rate variability. As expected, these RSA measures werevery highly and positively correlated (Spearman r � 0.84, P �0.0001).

Relationship Between RSA and Coupling

The relationship between both measures of RSA (direct RSAand HF power) and both measures of coupling [tRSE and�log10(P)] was assessed for the R-to-I interval. The rank ofeach measure of coupling was treated as a dependent variable

and regressed onto the rank of each measure of RSA, with visiteffects also in the model. The F-values for the regression werevery low (0.0, 0.01, 0.15, 0.36) and do not support the hypoth-esis that RSA and CVC [tRSE and �log10(P)] for the R-to-Iinterval are correlated.

Relationship Between RSA and Demographic Measures

Mean direct RSA and mean HF power were not correlatedwith height, weight, or BMI (largest magnitude Spearman r �0.12, 0.23, respectively).

DISCUSSION

CVC in awake, resting, supine, healthy subjects is strongestand most consistent for the interval between I-onset and theprevious R-wave (R-to-I interval). For both coupling measures[tRSE, �log10(P)], the R-to-I interval had the largest mean andwas the only interval that was significantly larger than themean across intervals. For tRSE, the mean estimate was sig-nificantly larger than 0 only for the R-to-I interval. Althoughthe pattern of CVC could vary across subjects, the R-waveoccurred preferentially at 138 ms before I-onset. The test-retestreliability of both CVC measures was significantly greater thanzero for the R-to-I interval, whereas the coupling measures forthe expiratory-related intervals were unreliable. These datasupport the hypothesis that CVC strength for the R-to-I intervalis a subject trait under these conditions. Furthermore, CVC forthe R-to-I interval only (as indexed by tRSE) was stronglyrelated to the size (height, weight, and BSA) of the subjects.Finally, CVC was unrelated to the two measures of RSA(direct RSA and HF power).

CVC has been observed in intact awake (6) and anesthe-tized animals (24), in clinical populations (14, 22), and inhealthy awake and anesthetized subjects (11, 12, 18, 19, 38).These studies identified CVC for the R-to-I interval,whereas our study compared each interval (R-to-I, I-to-R,

Fig. 3. Composite 27-bin histograms for all four intervals based on a sumacross subjects and visits. All histograms, regardless of statistical significance,are included in these composites. The numbers below each bar represent theaverage bin center, in ms, across subjects and visits. The average bin width was39.58 ms.

Table 3. Median heart rate and respiratory rate for the datapresented in Fig. 2

Subject #

Visit 1 Visit 2

Median heartrate

Median respirationrate

Median heartrate

Median respirationrate

12 78.7 13.2 75.1 15.420 55.7 15.3 60.8 15.5

Values are in cycles/min.

1253Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

R-to-E, and E-to-R) systematically and statistically. Weverified that the R-to-I interval is the key interval. As notedabove, the R-to-I interval was the only interval significantlylarger than the mean interval for both CVC measures. Ourfindings are consistent with Hildebrandt (17) who reportedmuch stronger CVC in R-to-I than R-to-E intervals and withLarsen et al. (25) who identified stronger coupling in theR-to-I than I-to-R intervals.

The time between the R-wave and I-onset varies from thatreported recently but is consistent with that reported in the1960s. We found an increased probability of R-waves oc-curring �138 ms prior to I-onset. Larsen and coworkers (12,38) reported this interval as 510 ms in anesthetized subjects(12) and 670 ms in healthy awake supine subjects (38).Although individual subjects had peaks in this time range(e.g., Fig. 2), the peaks were not sufficiently synchronous at

these earlier time points to be apparent in the group average(Fig. 3). In contrast, the peak at �150 ms is consistent withreports by German researchers in the 1960s (11, 17, 18, 19,31).

Differences in the method of assessing respiration and de-termining I-onset may account for these differences in R-to-Iinterval duration. Previous methodologies include the openingand closing of a Ruben nonreturn (one-way) valve (12), nasalpressure (38), a “thorakometer” (an instantaneous measure ofchest circumference; 19), a pneumatic inflatable cuff aroundthe chest (18), and a thermistor (11). All found CVC of R-Iintervals in most subjects. Whether one or another provides thebest or most consistent result remains to be determined andcomparison of methods might be useful.

This is the first report that has evaluated the test-retest reliabilityCVC [tRSE: 0.69, �log10(P): 0.83]. The ICCs, in the good and

Fig. 4. Histograms of tRSE values by interval type along with gamma function fits.

Table 4. Tests for non-zero mean estimates

Measure Interval Mean Estimate SE DF T Value P Value Cohen’s d Effect Size Desc.

tRSE R-to-I 1.99 0.21 54 3.34 0.0015 0.91 largetRSE I-to-R 1.41 0.22 54 1.57 0.1223 0.43 moderatetRSE R-to-E 1.19 0.22 54 0.79 0.4352 0.22 smalltRSE E-to-R 1.04 0.23 54 0.18 0.8568 0.05 —�log10(P) R-to-I 3.09 0.08 54 13.87 �.0001 3.77 very large�log10 (P) I-to-R 2.75 0.08 54 12.43 �.0001 3.38 very large�log10(P) R-to-E 2.10 0.08 54 9.13 �.0001 2.48 very large�log10(P) E-to-R 2.22 0.08 54 9.81 �.0001 2.67 very large

These mean estimates do not match the mean estimates in Figs. 4 and 5. ANOVA models use the logarithm of the raw value, and control for visit and visit*interval type interactions. The reported estimated means are in log units and have been converted back to the natural scale as e(estimated mean).

1254 Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

excellent range, respectively, suggest that these measures arereasonably similar when assessed in the same subject 5–27 daysapart. This is particularly noteworthy in light of the evidence forwaxing and waning of CVC during a single recording session(38). Our findings suggest that a single estimate of CVC based on

200 breaths is a reliable characteristic of an individual, whentested under comparable conditions.

In our resting, awake, supine subjects, RSA consisted of arelative decrease in R-to-R intervals following I-onset (theRR�1 and RR�2). The average of these two intervals was�150 ms shorter than the two R-to-R intervals at the end ofexpiration. Our measure of RSA (average of RR�1 and RR�2divided by the average of RR-1 and RR_Insp) was significantlyless than 1.0 (0.92 � 0.05 SD) and was the most reliable(test-retest ICC � 0.90) subject characteristic in this study.Furthermore, our measure of RSA was not related to eitherCVC measure for the R-to-I interval. This is consistent withseveral previous reports (13, 23, 24, 32).

The strong relationship noted between the strength of CVC(tRSE) and height, weight, and BSA has not been reportedpreviously. However, cardiovascular measures have been cor-related to height, weight, and BSA. Chirinos et al. (2) measured11 aspects of arterial load and ventricular-arterial coupling andreported that all these measures were positively correlated withheight, weight, and body size. The largest correlations werebetween height or weight or BSA, and stroke volume, totalarterial compliance, arterial load, and end-systolic left ventric-ular elastance. In other studies, height was positively correlatedwith skin microvascular resistance in adult women (15) andwith blood pressure in children, including 17 year olds (29).Two papers have reported that small vessel elasticity waspositively correlated with height and/or weight (9, 43). Perhapsthese correlations, including our correlation between CVC and

Fig. 5. Histograms of �log10(P) values by interval type along with negative binomial fits.

Table 5. Spearman correlations between coupling measuresand height, weight, BSA, and BMI vs. tRSE

Interval Type: R-to-I I-to-R R-to-E E-to-R

Height 0.66 0.24 0.29 0.24P � 0.0022 ns ns ns

Weight 0.67 0.18 0.45 0.36P � 0.0017 ns ns ns

BSA 0.70 0.25 0.45 0.33P � 0.0008 ns ns ns

BMI 0.40 �0.01 0.32 0.28ns ns ns ns

�log10(P)

Interval type R-to-I I-to-R R-to-E E-to-RHeight 0.25 0.21 0.02 �0.02

ns ns ns nsWeight 0.45 0.23 0.02 0.05

ns ns ns nsBSA 0.44 0.25 0.04 0.01

ns ns ns nsBMI 0.37 0.09 �0.12 �0.01

ns ns ns ns

BSA, body surface area; BMI, body mass index; ns, not statisticallysignificant.

1255Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

height, weight, and BSA, represent adaptations to the extraload demands of large people. In this light, our correlationsmight argue for a role of CVC in enhancing the efficiency ofoxygen delivery. It is interesting to note, in this context, thatneither measure of RSA was correlated with body size mea-sures.

Significance

We used the R-wave peak of the ECG signal because this isthe largest and most salient component and the R-wave oc-curred �138 ms before inspiratory flow. In future studies thatexamine cause and effect, determining the cardio-ventilatoryvariables will impact the latency measure. The R-wave occursjust prior to ventricular contraction. However, other cardiac-related events occur earlier; the P-wave, which indicates atrialdepolarization, occurs between 120 and 200 ms before theR-wave, and sino-atrial node depolarization occurs slightlyearlier. Regarding the I-onset, diaphragmatic activity beginsbefore air volume changes at the mouth (37), and upper airwaymuscles contract well before diaphragmatic activity (1, 30, 35,42). Thus, determining CVC based on onset of EMG activity inupper-airway muscles, the diaphragm and air flow at the mouthand nose may provide substantial insight into the mechanismsand effects of CVC.

As noted above, our definition of coupling does not includethe concept of causality. With the methodology we presentedwe cannot determine if the presence of a heartbeat just beforeonset of I facilitates I-onset. We can say that our data areconsistent with this interpretation but we can go no further.Proof of causality requires an experiment in which one of thevariables (R-wave timing or arterial pulse pressure) is manip-ulated and the effect on I-onset observed. This is hard toimagine doing in human subjects but could be attempted in

animal studies. However, several analytical methods examinecausality (for review, see 16). Perhaps the application of suchmethods in the future can provide additional clues to the causalrelationships.

Although neural control of cardio-ventilatory functions ap-pears to be integrated at the brain stem, integration at a systemslevel (typically focused on the efficiency of gas exchange)needs to be defined. At a brain stem level, although expiratoryneural activity can be primarily pulse modulated (7, 8), baro-receptor input only weakly resets the respiratory rhythm (27).Both the magnitude and the direction of resetting vary, mod-ulated by the afferent radiation of inputs from the lungs andchest wall and interdependence of medullary and pontinenetworks. However, in humans, chest wall activation patternscontrolling tidal volume are primarily inspiratory with passiveexpiration during quiet wakefulness. Furthermore, the activa-tion of upper airway, or flow modulating muscles, may occurbefore the diaphragm. Hence, CVC as we describe it fromcardiac activation and inspiratory flow is an example wheresystems are modulated at a level beyond that explained cur-rently by conceptual models based on reduced preparations.Speculation as to whether CVC is a primary or secondarycontrolled variable is beyond the scope of this paper yet thisbegs further investigation. Where and how this develops maybe an important element to determine, in addition to its phys-iological importance.

GRANTS

This research was supported by grants from the National Institutes of Health(HL-087377 and HL-087340) and by Award Number I01BX000873 from theBiomedical Laboratory Research and Development Service of the VA Officeof Research and Development.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: L.F., T.E.D., K.A.L., and K.P.S. conception anddesign of research; L.F. performed experiments; L.F. analyzed data; L.F.,T.E.D., F.J.J., K.A.L., A.Y., M.F., C.G.W., and K.P.S. interpreted results ofexperiments; L.F. prepared figures; L.F. drafted manuscript; L.F., T.E.D.,F.J.J., K.A.L., A.Y., M.F., C.G.W., and K.P.S. edited and revised manuscript;

Fig. 6. Scatter plot relating rank body surface area (BSA) to rank tRSE for theR-to-I interval. �Female subjects.

Fig. 7. R-to-R intervals before, after and spanning I-onset. Values are means(�SE) for all subjects and visits, after controlling for differences in medianheart rate.

1256 Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from

L.F., T.E.D., F.J.J., K.A.L., A.Y., M.F., C.G.W., and K.P.S. approved finalversion of manuscript.

REFERENCES

1. Brancatisano TP, Dodd DS, Engel LA. Respiratory activity of posteriorcricoarytenoid muscle and vocal cords in humans. J Appl Physiol 57:1143–1149, 1984.

2. Chirinos JA, Rietzschel ER, De Buyzere ML, De Bacquer D, GillebertTC, Gupta AK, Segers P. Arterial load and ventricular-arterial coupling:physiologic relations with body size and effect of obesity. Hypertension54: 558–566, 2009.

3. Cicchetti DV. The precision of reliability and validity estimates re-visited:distinguishing between clinical and statistical significance of sample sizerequirements. J Clin Exp Neuropsychol 23: 695–700, 2001.

4. Cicchetti DV, Sparrow SA. Developing criteria for establishing interraterreliability of specific items: applications to assessment of adaptive behav-ior. Am J Ment Defic 86: 127–137, 1981.

5. Cohen MA, Taylor JA. Short-term cardiovascular oscillations in man:measuring and modelling the physiologies. J Physiol 542: 669–683, 2002.

6. Coleman WM. The psychological significance of bodily rhythms. J CompPsychol 1: 213–220, 1921.

7. Dick TE, Morris KF. Quantitative analysis of cardiovascular modulationin respiratory neural activity. J Physiol 556: 959–970, 2004.

8. Dick TE, Shannon R, Lindsey BG, Nuding SC, Segers LS, BaekeyDM, Morris KF. Arterial pulse modulated activity is expressed inrespiratory neural output. J Appl Physiol 99: 691–698, 2005.

9. Duprez DA, Kaiser DR, Whitwam W, Finkelstein S, Belalcazar A,Patterson R, Glasser S, Cohn JN. Determinants of radial artery pulsewave analysis in asymptomatic individuals. Am J Hypertens 17: 647–653,2004.

10. Eckberg DL. Point:Counterpoint: respiratory sinus arrhythmia is due to acentral mechanism vs. respiratory sinus arrhythmia is due to the baroreflexmechanism. J Appl Physiol 106: 1740–1742; discussion 1744, 2009.

11. Engel P, Hildebrandt G. Uber den einfluss der vigilanz auf die phasen-koppelung zwischen herzschlag und atmung. Psychologische Beitrage 15:77–86, 1973.

12. Galletly DC, Larsen PD. Cardioventilatory coupling during anaesthesia.Br J Anaesth 79: 35–40, 1997.

13. Galletly DC, Larsen PD. Relationship between cardioventilatory cou-pling and respiratory sinus arrhythmia. Br J Anaesth 80: 164–168, 1998.

14. Galli G. Deuxieme contribution a l’étude des synchronisms cardio-respiratories. Arch Mal Coeur Vaiss 17: 208–221, 1924.

15. Gerhardt U, Hillebrand U, Mehrens T, Hohage H. Impact of estradiolblood concentrations on skin capillary laser Doppler flow in premeno-pausal women. Int J Cardiol 75: 59–64, 2000.

16. Hlavackova-Schindler K, Palus M, Vejmelka M, Bhattacharya J.Causality detection based on information-theoretic approaches in timeseries analysis. Physics Reports 441: 1–46, 2007.

17. Hildebrandt G. [The coordination of rhythmic functions in man]. VerhDtsch Ges Inn Med 73: 921–941, 1967.

18. Hildebrandt G, Daumann FJ. [Coordination of pulse and respirationrhythm during work]. Int Z Angew Physiol 21: 27–48, 1965.

19. Hinderling P, Bucher K. [Synchronism between circulation and respira-tion in humans]. Helv Physiol Pharmacol Acta 23: 374–381, 1965.

20. Karemaker JM. Counterpoint: respiratory sinus arrhythmia is due to thebaroreflex mechanism. J Appl Physiol 106: 1742–1743; discussion 1744,2009.

21. Karemaker JM. Last Word on Point:Counterpoint: respiratory sinusarrhythmia is due to a central mechanism vs. respiratory sinus arrhythmiais due to the baroreflex mechanism. J Appl Physiol 106: 1750, 2009.

22. Larsen PD, Booth P, Galletly DC. Cardioventilatory coupling in atrialfibrillation. Br J Anaesth 82: 685–690, 1999.

23. Larsen PD, Galletly DC. Cardioventilatory coupling in heart rate vari-ability: the value of standard analytical techniques. Br J Anaesth 87:819–826, 2001.

24. Larsen PD, Galletly DC. Cardioventilatory coupling in the anaesthetisedrabbit, rat and guinea-pig. Pflügers Arch 437: 910–916, 1999.

25. Larsen PD, Trent EL, Galletly DC. Cardioventilatory coupling: effectsof IPPV. Br J Anaesth 82: 546–550, 1999.

26. Larsen PD, Tzeng YC, Sin PY, Galletly DC. Respiratory sinus arrhyth-mia in conscious humans during spontaneous respiration. Respir PhysiolNeurobiol 174: 111–118, 2010.

27. McMullan S, Dick TE, Farnham MM, Pilowsky PM. Effects of baro-receptor activation on respiratory variability in rat. Respir Physiol Neuro-biol 166: 80–86, 2009.

28. Mosteller RD. Simplified calculation of body-surface area. N Engl J Med317: 1098, 1987.

29. National Heart, Lung, and Blood Institute. Blood pressure tables forchildren and adolescents, 2004. http://www.nhlbi.nih.gov/guidelines/hypertension/child_tbl.htm.

30. Onal E, Lopata M, O’Connor TD. Diaphragmatic and genioglossalelectromyogram responses to CO2 rebreathing in humans. J Appl Physiol50: 1052–1055, 1981.

31. Rascke F. The hierarchical order of cardiovascular- respiratory coupling.In: Cardiorespiratory and Cardiosomatic Psychophysiology, edited byGrossman P, Janssen KHL, and Vailt D. London: Plenum, 1986, p.207–218.

32. Schäfer C, Rosenblum MG, Kurths J, Abel HH. Heartbeat synchro-nized with ventilation. Nature 392: 239–240, 1998.

33. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing raterreliability. Psychol Bull 86: 420–428, 1979.

34. Singh D, Vinod K, Saxena SC. Sampling frequency of the RR intervaltime series for spectral analysis of heart rate variability. J Med EngTechnol 28: 263–272, 2004.

35. Strohl KP, Hensley MJ, Hallett M, Saunders NA, Ingram RH Jr.Activation of upper airway muscles before onset of inspiration in normalhumans. J Appl Physiol 49: 638–642, 1980.

36. Task Force of the European Society of Cardiology and the NorthAmerican Society of Pacing Electrophysiology. Heart rate variability:standards of measurement, physiological interpretation, and clinical use.Circulation 93: 1043–1065, 1996.

37. Taylor A. The contribution of the intercostal muscles to the effort ofrespiration in man. J Physiol 151: 390–402, 1960.

38. Tzeng YC, Larsen PD, Galletly DC. Cardioventilatory coupling inresting human subjects. Exp Physiol 88: 775–782, 2003.

39. Tzeng YC, Larsen PD, Galletly DC. Mechanism of cardioventilatorycoupling: insights from cardiac pacing, vagotomy, and sinoaortic dener-vation in the anesthetized rat. Am J Physiol Heart Circ Physiol 292:H1967–H1977, 2007.

40. Tzeng YC, Sin PY, Galletly DC. Human sinus arrhythmia: inconsisten-cies of a teleological hypothesis. Am J Physiol Heart Circ Physiol 296:H65–H70, 2009.

41. Weiss HR, Salzano J. Formation of whole number ratios of heart rate andbreathing frequency. J Appl Physiol 29: 350–354, 1970.

42. Wheatley JR, Brancatisano A, Engel LA. Respiratory-related activity ofcricothyroid muscle in awake normal humans. J Appl Physiol 70: 2226–2232, 1991.

43. Winer N, Sowers JR, Weber MA. Gender differences in vascularcompliance in young, healthy subjects assessed by pulse contour analysis.J Clin Hypertens (Greenwich) 3: 145–152, 2001.

1257Cardio-ventilatory Coupling • Friedman L et al.

J Appl Physiol • doi:10.1152/japplphysiol.01424.2010 • www.jappl.org

on Decem

ber 22, 2014D

ownloaded from