This is a very interesting question. Observe that the neural activity (sympathetic and vagal tones) alters the dynamics of the sinoatrial node, which causes the quasi-periodicity of the RR intervals (or the asymmetry observed in the Poincare plot). Without the neural control of the heart, this asymmetric behavior vanishes, as observed in cardiac transplanted subjects. Consequently, the asymmetric response of the sinoatrial node depends on the neural activity. Physiologically speaking, the pacemaker cells have different intrinsic resistances. Therefore, the sinoatrial node can be considered asymmetric due to the structural or topological organization. Does this answer your question? You may find more details about the neural processing underlying the cardiac regulation in a recent paper published in PLoS ONE entitled "Statistical Coding and Decoding of Heartbeat Intervals" (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0020227).

Hi Teodor,

I am not sure about the effects of sinus node properties, but think that "symmetric/asymmetric neural control" would be a significant source of RR asymmetry. A good starting point to look at would be respiration as it is one of the main drivers of RR variability. For example, inspiration and expiration are not "symmetric" in duration. Then you have the effect of the pulmonary stretch receptors on the vagal nerve, which sense inspiration. On longer time scales sympathetically mediated inhibition of the vagal nerve might add further asymmetry.

Hi Mathias!. Of course you are right, but on the other hand there is a whole internal dynamics of sinus node cells, in which I may easily imagine some inertia and some assymetry in time: that the onset phase of response to the vagal stimulus is different than the decay phase.

This is a very interesting question. Observe that the neural activity (sympathetic and vagal tones) alters the dynamics of the sinoatrial node, which causes the quasi-periodicity of the RR intervals (or the asymmetry observed in the Poincare plot). Without the neural control of the heart, this asymmetric behavior vanishes, as observed in cardiac transplanted subjects. Consequently, the asymmetric response of the sinoatrial node depends on the neural activity. Physiologically speaking, the pacemaker cells have different intrinsic resistances. Therefore, the sinoatrial node can be considered asymmetric due to the structural or topological organization. Does this answer your question? You may find more details about the neural processing underlying the cardiac regulation in a recent paper published in PLoS ONE entitled "Statistical Coding and Decoding of Heartbeat Intervals" (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0020227).

Do you have a hard link to "Without the neural control of the heart, this asymmetric behavior vanishes, as observed in cardiac transplanted subjects"? The transplanted hearts observe the 20% of the RSA via right atrium stretch receptors and I would not say that the pattern is necessarily symmetric.

Also I have never seen the term "intrinsic resistances", what do you actually mean?

can we go back to animal experiments such as Langendroff preperation for isolated heart in rabbit? because the possibility of changes in the sinus functions that to intrinsic will depend upon its surrounding electrolytes and membrane properties.

I have been enjoying this thread!

Let us not forget Leon Glass' pioneering work on sinus node cells self-orgnization. If we are to think of intrinsic versus extrinsic variability of heart rate, the degree of weak nonlinear coupling between individual sinus cells (or ensembles thereof) should be influenced by the milieu interior and contribute to HRV. I agree I do not think this has been studied.

Hi Teodor,

Sorry for the late replay. The hard link is http://www.ncbi.nlm.nih.gov/pubmed/2793748. I agree that changes in the heart rate have been observed in heart transplant subjects (14.3% during single deep breath). However, the cross-correlation between RR intervals and respiration shows minimal oscillations when compared to normal volunteers. Moreover, the heart rate (at most of the time for transplanted subjects) is relatively stable with small variations (3.8% and 4.9% in Vasalva and Mueller maneuvers, respectively). These variations are very small, and are likely to be caused by (quantization) errors introduced by the A/D converter or noise.

Regarding the “intrinsic resistance”, I want to talk about that electrical activity of SA node cells in terms of cell capacitance. Basically, SA node cells of different size have different cell capacitance (http://cardiovascres.oxfordjournals.org/content/47/4/658.full), meaning that they have different action potential configurations. Therefore, we can assume that their organization is asymmetric.

I agree to all of you who speak about the role of collective phenomena in sinus node function.

Fausto: I was aware of spatial inhomogeneity of the SA node but I would rather think of it in terms of shorter or longer intrinsic rates that are generated in different parts of the SA node (pacemaker wandering phenomenon). I don't know if "size matters". If you have larger cell it has larger capacitance but also larger area. Of course volume (r^3) grows faster than area (r^2) (take a spherical cow) but maybe for larger cells the ionic currents are stronger too? But the observation is interesting indeed.

This discussion does not simplify my question, but provokes to ask another: what difference does it make to the dynamics: the fact that we have rather a tissue than a single cell.

Of course credit goes to Glass, Guevara, Dokoss, Jalife, Boyett, Demir, Zhang, Noble, Slenter & Salata and innumerable others (Hi Martin!).

Jyotsana: as far as I know Oliver Monfredi from Boyett's lab does measurements on Langendorff: note e.g. this http://spo.escardio.org/eslides/view.aspx?eevtid=40&fp=P4078

Buchner, thanks for sending me such a nice link.

I agree with the approach of collective phenomena in a net of NON LINEAR contributions. I see also the limits in the possibility to actually investigate this matter since HRV analysis continues to apply (mainly from clinicians ) standard but limited methods of investigation. As example we know that HRV analysi is based in frequency domain on FFT that instead should be applied, in rigorous terms, only to strictly periodic, stationary and linear processes. R-R signals seem to violate all such conditions but the FFT dominates in HRV studies. Certainly it gives good results good results in VLF,LF and HF bands analysis but we cannot forget that it has several limits.

Please, consider also that in R-R analysis usually is not performed sampling at equally spaced time intervals as it should be required by using FFT or DFT. Often softwares resample at a given and prefixed frequency. This is a good methodlogical approavh from a certain view point but of course we cannot ignore that so doing we lose important informations. So in conclusion, R-R is one of the most important signals in Nature, also the most complex from a certain point of view but of course we are missing under the profile of its satisfactory investigatiuon. The problem is not posed from a general view point. In fact we have in this case advanced procedures under the linear and non linear methodology. However, the problem is that R-R is modulated by two essential armos that are given from the ANS. We need to quantify VLF,LF and HF for imprtant evaluations under the clinical profile and to understand the inner structure of this signal as the posed question indicates. Here , in analysis of such bands in frequency domanins , we are missing.

If RR intervals are not resampled, the spectrum has a mysterious frequency unit: cycles per beat. Although there are methods as spline-based resampling or Lomb evolutionary periodogram that either provide constant sampling or offer an alternative.

Lomb. certainly represents and aletrnative but you certainly know that FFT or DFT are mainly used.. I any case resampling , at least in principle, gives a lose of information.....but it is not the only current probelm. As example there is still the problem to identify the most appropriate window .. and still more. Finally, there is the problem of the controls( pathologic cases. I am not so sure that FFT may give correct results when examining pathological subjects.That is still another final problem. By using FFT or DFT we arrive to the spectrum in frequency domain. We have PSD or amplitude spectrum. Strictly from a physiological point of view, the power in PSD or the Ampplitudes in the Aplitude(frequency spectrum what should represent from a conceptual point of view?