Sleep 7.2 Cardiopulmonary Coupling: Sleep Staging Classifications

Learn how CPC compares to traditional sleep staging, its advantages and limitations, and how it connects to sleep stability and autonomic physiology.

Video Breakdown:

00:00 Introduction to Cardiopulmonary Coupling

00:52 History and Development of CPC

01:54 Technical Aspects of CPC Analysis

04:32 Clinical Applications and Case Studies

09:03 CPC in Different Populations

09:48 Conclusion and Future Directions

This is where I wanna go into cardiopulmonary coupling and how cardiopulmonary coupling has been used to sleep staging.
So, cardiopulmonary coupling has been espoused by some of its various influencers. This is Andy Galpin who's gone on a number of like Andy Huberman and, The Diary of a CEO and this is just a segment of something he talked about. "But PSG would generally be the gold standard. Now, another way to approach sleep at the highest standard is to use a method called cardiopulmonary coupling. Now, this has actually been around for a long time. It's got dozens and dozens of studies behind it, though most people have never heard of it.

It's a little bit different than PSG. It approaches different things. It has a different target. It's looking more at your physiology, your autonomic nervous system, your respiratory rate. And things like that. Um, it was developed by Dr. Robert Thomas, from Harvard. And again, it's been studied in a lot of areas.
It's FDA approved as a way to diagnose sleep disorders, and so there are some pros and and cons to each of these approaches. I would recommend if you're interested in that, to talk to your physician or medical team to decide which approach PSG or CPC is better for you based on the questions you have and trying to get answered." Anyway, that's a, a segmentas an example.
So the landmark paper that, you know, brought CPC to the forefront was the one that was done by Robert Thomas in 2005, at Beth Israel Deaconess Medical Center. I really never asked Robert why he intuitively thought to assess the coupling between heart rate and respirations as a way to assess quality of sleep because, I know Robert, I was also at Beth Israel Deaconess Medical Center while he was there, but I would guess
that it was a way to capture the autonomic nervous system changes that are seen with unstable or CAP sleep, and which is not routinely captured on the normal EEG. What he did along, with Dr. CK Peng and Ary Goldberger, you know, basically, Labfront is a spinoff from the lab that came out of for this group is, there was, the use of just a single EKG.
And this EKG was beat-labeled, and then it was divided into two components. One, it was analyzed for the RR or the N-N interval to establish the heart rate variability. And then they used, EKG derived respiration to obtain the respiration waveforms or the time series.
And then they, do a cross spectral and coherence calculation, which is the cardiopulmonary coupling method. Then to calculate the CPC. And just a little information about how they got the respiration from the EKG. So this is an example of a raw EKG you could see here. But if you take the magnitude of a QRS waveform, you could see that there's a fluctuation there.
And this fluctuation happens because as you breathe in, your lung expands. The air has a higher electrical impedance, and so that diminishes the signal here. It also causes the heart to shift down so it changes the axis. So in other words, the QRS changes correspond to respiration, and you can see that in gray, which corresponds nicely with a piezoelectric belt assessment here as well.
And basically, this team, Robert Thomas's team, trained a number of data sets using a dataset, so there's a training set, and determined the various markers that were needed in order to categorize into non-CAP, which is the stable, CAP was the unstable, wake, REM and unknown.
And then you have a number of cutoff variables. So you could see that the high frequency, low frequency, and very low frequencies are based on these ranges from 0.1 to one hertz for high frequency. 0.01 to 0.1 for low frequency and zero to 0.01 hertz for very low frequency.
And you can see sort of the cutoffs in order to determine whether there's a non-CAP and CAP. And then they test it on the testingdataset. And you know, I would say the cardiopulmonary coupling did decently well with sensitivity about 81% for non-CAP, specificity about 84% for CAP. But recognizing that even CAP and non-CAP is a visual scoring method to begin with, so that in itself is imperfect.
So the gold standard is imperfect. But it became interesting to see that actually, when you take a look at the data itself, the analytical methodologies, you can see that it is quite helpful. So this is an example of a 22-year-old healthy woman. You can see the traditional hypnogram. There's four stages, because this was study was done back in 2005.
You see the delta power here from the EEG to suggest the slow wave activity, which indicates restful sleep. And then you have the CAP stages, so wake, REM, and CAP. So CAP is the unstable, non-CAP is the stable one. And then, the low to high ratio, and the low or high ratio for the CPC analysis.
So the CPC actually is demonstrated here. So this is the high frequency coupling, which shows again, good, stable non-REM sleep correlating with non-CAP. And then the low frequency, which indicates unstable sleep. And you can see that with the high frequency, it's correlating with deeper stages of sleep.
More importantly with higher delta power, indicating that this is really good restful restorative sleep. And again, you can see that when you have more low frequency component as indicated here, that correlates with the CAP stages. So that was an indication there as well. And this is another example of that as well, for a 56-year-old healthy woman. Again, in this case, you're seeing high frequency coupling, not as prominent in this younger woman.
More low frequency coupling, which indicates more unstable sleep. But interestingly here is where you see sort of,evidence of high frequency coupling, but it's not categorized as deep sleep here. It's because this elderly woman probably didn't fit that 75 microvolt criteria for delta sleep.
But, you could still see that the delta power is there and in some ways the cardiopulmonary coupling may do a, better job, in terms of differentiating good stable sleep compared to the sleep staging that's conventionally used. And then it became useful from a clinical standpoint. So this is a 56-year-old man with sleep apnea.
You could see before CPAP, you see much more low frequency coupling, again, indicating unstable sleep. But the moment you added CPAP here, you could see that the high frequency couplings increased. And then you, interestingly, you don't see that much of a difference 'cause even with pre-CPAP, you're not seeing deep stages of sleep, and you're not really seeing that much of a difference. So the CPC is differentiating the appropriate response to CPAP. And then this individual with complex sleep apnea, who has a combination of obstructive and complex sleep apnea is showing, basically the poor sleep quality during the positive airway pressure, despite getting the CPAP, but with the addition of dead space, which addresses the essential sleep apnea, you could see that it's doing its job. So again, it looks like, this cardiopulmonary coupling is doing a good job. The other thing is that, as mentioned before, is that the cardiopulmonary coupling particularly a high frequency component, does collaborate or corroborate with the delta power of the EEG. So you can see that here and it matches very nicely for the two. And, interestingly, there's a lag between the two, between the, the cardiopulmonary coupling and the delta that you're getting from the EEG and if anything,the cardiopulmonary coupling, the high frequency component precedes the delta power. So it precedes about 4.3 minutes. And to indicate that the autonomic tone, like the, appropriate parasympathetic setting is in some ways necessary for the brain to get into that sort of deep sleep. So the body itself is pretty important for that, or at least that's the suggestion, from that lag there.

The other thing is that high frequency coupling is associated with heart rate dipping. You can see the high frequency component here, and you can see that the heart rate is reduced at the moments that the high frequency coupling is elevated. And this is also true from blood pressure dipping as well.
So this high frequency coupling component is a great sign. It means that you're getting restorative sleep. This is the time when you're getting more delta power in your Delta sleep. Your heart rate is down, your blood pressure is down, and so that's the good type of sleep. And it's, shown to that it's relevant, at least in various clinical populations.
This is a study done by Yang in 2011, who is, also part of our group at Beth Israel Deaconess. And you can see a depressed patient without medications. And you can see that there's a reduction in the high frequency, more of this low frequency coupling and then this healthy subject, on the other hand having, a more prominent high frequency coupling.
And then even with medications, it's sort of tracked with medications. So hypnotics, are, are able to improve the sleep for these individuals with depression. And that is corroborated by the CPC measures. So there's more increase in high frequency coupling when you get the medications equivalent to the amount of high frequency coupling that are getting in healthy individuals.
And there is a correlation also with age. So we know that in young children, as you get into the adolescent years, your adolescent years are characterized by poor sleep, you know, just erratic, irregular sleep, and you could see that by the total sleep time as you increase in age from, three years of age to 22.
And you could see that each of these dots is an individual and this black marker here, the fill dots are the average for each of these ages, and you can see that the total sleep time decreases as you get older into your adolescent years. And then the high frequency coupling also decreases, and then the low frequency coupling also increases.
So indicating that you have an less good sleep, more bad sleep. So that is showing how CPC is used for staging.

Course episodes

Sleep 7.1 Understanding Cardiopulmonary Coupling in Sleep Studies

Sleep 7.2 Cardiopulmonary Coupling: Sleep Staging Classifications

Sleep 7.3 High vs. Low Frequency Coupling in Cardiopulmonary Physiology

Sleep 7.4 Cardiopulmonary Coupling and Sleep Apnea Detection

Sleep 7.5 Traditional vs. Modern CPC Methodology

Sleep 7.6 Cardiopulmonary Coupling in Sleep: Strengths, Limitations & Future