2.3 HRV in Cardiology: Clinical Evidence & HRV in Diabetic Autonomic Neuropathy
Dr. Ahn provides his interpretation of early HRV trials, the effect of early revascularization, and diabetic autonomic neuropathy.
- 0:00 - 8:29 Early HRV Trials
- 8:30 - 14:18 Effect of Early Revascularization
- 14:19- 16:21 Diabetic Autonomic Neuropathy
- 16:22- 19:37 Personal Interpretation
- So now let's go back to the clinical indications of heart rate variability. So even as early as the 1970s, heart rate variability was recognized as a promising tool for assessing autonomic nervous system functioning. But was there any preliminary evidence to support the notion that heart rate variability can be translated to the clinical setting? And yes, there was no the first evidence was in it was done in Australia in the 1978. Evaluate to monitor 30 patients admitted to the CCU for acute myocardial infarction. It took a 62nd EKG on the day of next and calculated the our interval variance, which is essentially the standard deviation of 30 consecutive beats, and assessed for hospital mortality. And the he divided the group into two parts, those with sinus arrhythmia, and this is actually more specifically respiratory sinus arrhythmia. So it's, it's actually a good arrhythmia, not the deadly arrhythmias. And the ones with sinus arrhythmias had lower mortality about 4.1% versus 15.5%. And so you can see here that those patients with no sinus arrhythmia, so essentially low heart rate variability, compared to those with sinus with me as those with low variability had a good number of deaths, and regardless of a heart rate. So yes, the heart rate didn't seem to necessarily differentiate those who would have been at risk of dying of in the hospital. Where is the those with heart rate variability, higher heart rates are built to have lower rates of death.
- This is the sort of landmark study that was published by the multicenter, postinfarction research group in 1987. And this is the first time to really indicate why Heart Rate Variability got a lot of attention. This was a multicenter study, this was done. In fact, I think, in multiple countries, 808 patients were recruited who survived acute myocardial infarction, the measured 24 hour continuous EKG 11 to three days after plus or minus three days after the myocardial infarction, then measured 24 hour standard deviation of an interval, this was their marker of heart rate variability, and they assessed all cause mortality. And you could see the graph here. This is the Kaplan Meier curve of survival with respect to time after myocardial infarction. And for those people with standard deviation s dnn above 100. In other words, high variability have a greater chance of staying alive over the four years, those with 50 to 100 have a lower chance of survival, and those below 50 clearly did poorly. So this received a lot of attention at that time.
- And to get to a little bit of physiology. So what physiological process does SDN most correlate with? So the question is when you're looking at this, at least me as a clinician is, what is this indicating is happening here. And remember, on my This is the slide that I introduced in my first talk, it the standard deviation really depends on the duration of the data. So if you're just measuring two minutes apart rates, then the standard deviation relates to the high frequency domain, physiological process that's respiratory sinus arrhythmia, when you're five minutes, largely, it's the low frequency or bare reflex properties. But at 24 hours, you're talking about circadian hormonal rhythms. And this prior study was a 24 hour standard deviation and an interval. So the circadian hormone factors are clearly playing a factor here in distinguishing those people who's five and those who don't. And this st group basically did a follow up study, because they wanted to understand what accounted for this low variability like what was sort of the differences, what accounted for the differences in variability that they saw in these patients. And they took 10 patients with high heart rate variability, and those with low heart rate variability, they match them according to age, gender, or risk factors, etc. And then they compare their their 24 hour continuous EKG two weeks after the myocardial infarction. And they found that the low heart rate variability generally had a faster heart rate both day and at night.
- There was less of a day and versus night heart rate for heart rate difference, which again gets the this notion that there's a circadian component here that is probably important. But the other thing is that there was a lower proportion of success and difference in intervals are this the metrical PNN 50, which again is a marker of the high frequency for the parasympathetic nervous activity in the body. And this is an example of sort of the PNN 50, over time, from midnight through the rest of the day. And you can see that the vagal nerve is highly active during that night, and during the day, it goes down and high in the high heart rate, frequency variability group, but basically no change in the low HRV group. So this was for time domain measurements, because standard deviation PN and 50 are time domain measures like does frequency domain measures similarly, have information about prognosis after an acute MI, and that was confirmed by this study done in 1992.
- Again, by the same group, they took 715 patients with acute MI, measured continuously kg two weeks after the MI and then they do the frequency domain analysis. And the outcome was all cause mortality. And then they took a look at high frequency, low frequency, very low frequency and ultra low frequency and divided course, according to all cause, and then divided mortality and cardiac or rhythmic. And you can see that the ones with a high Z factor, meaning it's more statistically significant. The ultra low frequency was the one that had the greatest prediction. For all cause mortality, the very low frequency had highest risk prediction for cardiac deaths, whereas high frequency and low frequency not so much. So again, going back to the idea that a lot of these lower frequencies have some very important information, you can see this visually, they've divided this into categories, the ultra low, very low, low frequency and high frequency. And you can see that there's a big difference between those who have high ultra low frequency and those with low ultra low frequencies, the ones that high Ultra tended to survive. And there's less of a difference for high frequency range. And this is the summary of the data, which was done by review. And Brett No. And generally, this positive response was replicated across all the studies all through 1987 to 202,002. Generally large trials with hundreds of patients. Many of them were measured, you know, a week to two weeks after the myocardial infarction. And they had measured multiple, and this was confirmed in various measures including time domain frequency domain. And the key thing is that this persisted this this positive results persisted despite being in the age of thrombolysis lysis.
- You know, thrombolysis is administrated ministration of a drug, which breaks down the plot that causes a heart attack. And so this positive effect pertains, this study, which is the only negative study in this group was attributed to inappropriate breaking down of the groups, one group actually had more beta blockers was generally healthier than the other groups. So I think this was confounded by the results. But what's important to know is that they also found that you didn't really need 24 hours that even short acquisition of pharma variability, maybe two to 15 minutes, or even five to 10 minutes was good enough to predict for mortality.
- But then something changed. And this is the same table, but it's too big. But these are the latest studies that happened in 2005. And something happened after 2005. You started to develop these negative studies that didn't show positive results when it comes to heart rate variability predicting mortality. And what happened around this time in 2005, was that you really had a massive uptick of angioplasty and stents. also incorporating beta blockers, which I had shown before significant reduced ventricular fibrillation, you know, chances of developing ventricular fibrillation.
- And in this study alone, you had about 70% of the patients getting revascularize 94% of them are already on beta blockers. So you started to see this difference. And I wanted to sort of quickly show you the dramatic rise of coronary revascularization that you saw it during this time in the late 2000s. This is a graph that was obtained from the Netherlands from a single hospital you could solve see that the number of angioplasty cysts still increased, but importantly, the stents used to keep open the vessels. The coronary vessels that were clogged in a heart attack, had a dramatic rise. And more importantly, also was that they started to involve they started to do early reproof. Usually, and they emphasize acting urgently when someone has a heart attack, rather than waiting a day or two, they recognize that's really important. As soon as someone comes into the hospital, take them directly to the cath unit, or to the shirt to the operating room to reopen the vessels. And the reason that happened is that, you know, they started to see evidence of quicker interventions, early repute, were we perfusion, and enhancing survivability after our heart attack.
- This is a nice study that shows the early heart rate variability measurements after a PTC intervention. This took about one to 23 patients with a first MI, they took 24 hour Holter monitoring beginning at the hospital admission. And then they followed these markers high frequency low frequency as to SDN AI and AR AI over time. And you can see that right after getting the reperfusion or the intervention, you had a drop in heart rate variability, and then a steady rise in the heart rate variability, and all the measures here. And this was done. You can see this for the early reperfusion individuals, those who got revascularisation 12 hours after symptom onset. However, those who got late reperfusion, if you wait 12 hours or more, you saw no change in this heart rate variability. So the heart rate variability, clearly, you can reverse it if you reverse the reduction, the hard drive, if you do the revascularisation early on.
- And this is an example of that study that was done in this era of immediate reperfusion. And again, I'm not going to go into too much detail, but this was a foreign to 12 patients. They had measured SDN N, and they found out that standard deviation, this SDN was no longer a significant risk factor for mortality. In fact, it a p value of point one had a little bit of a trend, but really not statistically significant. So early reperfusion changed the heart rate variability equation and how we treat patients, it just no longer became useful in this era of an immediate reperfusion. What it was interesting for me to find out was that in this specific study, only 7% 19 out of 412 patients had SDN, less than 50. This indicated that the interventions that were done was really effective in reversing the reduction of heart rate variability, and comparison, your original 1987 and pet study that I showed previously, had about 15.5% of the patients with low SDN. So the interventions that we're doing in cardiology was clearly making difference to recovering and improving heart rate variability. And as a result of this, in this era, we were starting to see that heart rate variability no longer had its specificity or positive predictive value in one year cardiac mortality or rhythmic events or Rate Variability alone, which is shown here. Based on the you know, based on the sensitivity, it really had a very low positive predictive value. And the improvements that we saw in in the delivery of cardiac care really showed up and the epidemiology of heart disease.
- We saw, for instance, the rates of heart disease death, plummet significantly, actually, from 1970s onwards. In 1970, for instance, there was about 670,000 deaths per year in 2010, without 412,000 deaths, so a reduction of 39% Despite a rise in the overall United States population, and then the case fatality rates for those individuals who are hospitalized with acute MI, those who are less than 65 years of age, the rates of mortality in the hospitalization dropped from 60% to less than 2%. And those who are aged greater than 65, dropping 38% to 7%. So heart disease is a success story for medicine.
- And I didn't want to ignore the other condition where for which Heart Rate Variability has been demonstrated to be useful, and that is autonomic neuropathy and diabetes. This was the sort of the landmark study done in 1990. There was 25 diabetic patients compared to 11 Healthy controls. The diabetic patients had advanced cases, symptoms and signs suggestive of autonomic neuropathy, which included signs of impotence, bowel, bladder disturbance, abnormal sweating, they tended to have diabetes for a long duration, peripheral neuropathy, and then they measured their autonomic function Using the autonomic testing that I briefly briefly described earlier. And then based on this testing, they were just able to distinguish those 25 patients into two groups, those with vagal, neuropathy and those with and without vagal neuropathy, and they evaluate the standard deviation of successive difference for the RR interval. And this is heart rates, you can see heart rates for the healthy individuals with respect to time of day. And you can see that basically, the heart rate is faster because the RA will shore shorter during the day and then at nighttime gets slower for the healthy individuals. And then for the diabetic patients either with with or without bagel neuropathy. It has a lower our interval, but it's distinction is not that great. But when you use heart rate variability, you see the heart variability, the distinction between heart rate and heart rate variability is much greater. And the heart rate variability really, were able to detect those individuals with autonomic symptoms, but not this label of vagal neuropathy according to the autonomic nervous system pretty well.
- So even for those people who were labeled as without bagel neuropathy had low heart rate, heart rate variability. So just to give a little bit of my interpretation of the heart rate variability clinical data, I would say that heart rate variability did really consistently well in predicting mortality before the age of coronary early coronary revascularization. And that optimization. And me, it was surprising to see these consistent results, particularly given the market variability that you see in heart rate variability, and I'll talk about this later. But there's a lot of variability in heart rate variability between persons and also within individual, because the individual can see signs of change in heart rate variability through the course of the day. The greater heart rate of frailty seen in acute MI patients treated monitor indicates that the treatments were effective. But it also indicated that heart rate variability may still be useful. Because Heart Rate Variability was showing actually, that our interventions were really making a difference. And this is why we're seeing, you know, dramatic reductions in heart disease deaths from acute MIS. And so this is not to discount heart rate variability of measure, it just indicates that we were doing such a good job, the heart rate variability was was increasing. And it wasn't good enough to distinguish those who were going to be at risk later on of developing death for mortality. There's something really important about very low frequency and ultra low frequency, I think this is an understudied area.
- Then the other understair area is the temporal modulation of high frequency or RMSD, or low frequency over the course of the days or weeks or even months. And this may be useful, but it's not, you know, it's under study, many of the high frequencies and low frequency data that was reported in these large randomized control trials for averages over the course of the day, and it really didn't respect the changes that you saw in high frequency, low frequency through the course of the day. And the final thing is that even in the error of acute coronary revascularization, heart rate variability may still be utilized well for subacute, late phases of post MI. And the reason I say that is that the two studies that were positive in the angioplasty or that new revascularisation era had obtained the heart rate variability much later in basically in the sub acute phase of the post mi time. And they measured 70 days, even up to 120 plus or minus 121 days.
- What this indicates is that there's a large tail, so probably a year or even a year after and six weeks after, those are the ones that heart rate variability. turnouts actually show some promising ability to predict depth. And then finally, heart rate variability may be exquisitely sensitive to the autonomic pathology as documented by the diabetic autonomic neuropathy example