Skin sympathetic nerve activity in patients with chronic orthostatic intolerance

Abstract

Background.

Chronic orthostatic intolerance (OI) is characterized by the development of tachycardia and other symptoms when assuming an upright body position.

Objective.

To test the hypothesis that skin sympathetic nerve activity (SKNA) bursts are specific symptomatic biomarkers in patients with chronic OI.

Methods.

We used an electrocardiogram (ECG) monitor with a built-in triaxial accelerometer to simultaneously record SKNA and posture in ambulatory participants. Study 1 compared chronic OI (14 women and 2 men, 35±10 years) with reference control participants (14 women, 31±6 years). Study 2 included 17 chronic OI participants (15 women and 2 men, 39±12 years) not yet treated with ivabradine, pyridostigmine or beta blockers.

Results.

In Study 1, there were 124 episodes (8±4/participant) of postural changes, with 11 (8.9%) episodes associated with symptoms. In comparison, 0 of 104 postural changes (7±3/participant) in control were symptomatic (p=0.0011). In chronic OI participants, the SKNA bursts associated with symptoms had higher burst frequencies, longer burst durations, and larger mean burst areas than bursts during asymptomatic periods. However, SKNA bursts and tachycardia were asymptomatic in control. We analyzed 110 symptomatic episodes in Study 2 (6±5 per participant). Among them, 98 (89.1%) followed least 1 SKNA burst. In comparison, only 41 (37.3%) had HR exceed 100 bpm 1 min prior to symptom onset (p<0.0001).

Conclusions.

SKNA bursts are a highly specific albeit insensitive symptomatic biomarker for chronic OI.

The term postural orthostatic tachycardia syndrome (POTS) was coined by Schondorf and Low¹ to describe a clinical syndrome of tachycardia at rest or during a head-up tilt. The commonly accepted diagnostic criteria include chronic (>3 months) orthostatic intolerance (OI) and an increase in heart rate (HR) of ≥ 30 beats/min (bpm) without orthostatic hypotension when moving from a recumbent to standing position.^2–4 Symptoms of POTS include persistent lightheadedness and fatigue, but gastrointestinal dysmotility has also been frequently observed in POTS patients since the original report.¹ Subsequent studies have confirmed that POTS is not a disease limited to the cardiovascular system. Rather, it is often associated with other systemic illnesses, especially mast cell activation syndrome (MCAS) and hypermobile Ehlers-Danlos Syndrome (hEDS).^{5, 6} While tachycardia is in the name, patients with POTS may not consistently demonstrate ≥30 bpm HR increases during orthostatic tests.^{7, 8} For that reason, Raj et al⁸ noted that some patients with chronic OI have POTS while others have postural symptoms without tachycardia (PSWT). The latter proposal downplays tachycardia as a diagnostic biomarker for this disease, but that proposal has not been accepted by the most recent National Institutes of Health expert panel.³

Chronic OI patients have enhanced noradrenergic tone⁹ while POTS is associated with increased cardiac norepinephrine release.¹⁰ It is possible that sympathetic nerve activity (SNA) patterns may serve as a better biomarker than tachycardia in identifying and defining chronic OI. We have developed and validated a method (neuECG) to simultaneously record electrocardiogram (ECG) and skin sympathetic nerve activity (SKNA) using conventional ECG patch electrodes.^{11, 12} Recently, we discovered that a commercially available Faros 180 ECG monitor (Bittium Corp, Oulu, Finland) can be used for neuECG recording.¹³ This ECG monitor also has a built in accelerometer that can be used to determine body posture, as well as a symptom marker. It is therefore an ideal instrument to study the relationship between HR, SKNA, and postural changes in ambulatory patients with chronic OI. The purpose of the present study was to perform simultaneous ambulatory neuECG and accelerometer recordings in chronic OI and normal healthy participants to test the hypothesis that SKNA bursts are specific biomarkers for chronic OI symptoms.

Methods

The detailed Methods are described in an Online Supplement. The protocol for this study was approved by the Institutional Review Board of Cedars-Sinai Medical Center. Study 1 included 16 consecutive participants with a diagnosis of chronic OI who came for in-person clinical visits (Table 1). We also recruited 14 asymptomatic women of child-bearing age as reference controls. After the results were analyzed, we performed Study 2 in 17 chronic OI participants not treated with sinus node suppressing drugs (ivabradine, beta blockers or pyridostigmine). The data were analyzed to further confirm the findings of Study 1. All participants had ambulatory neuECG recordings and body posture determinations using a triaxial accelerometer for 24 hours. Statistical tests were done using IBM SPSS Statistics 24 (SPSS Inc, Chicago, IL, USA) and R version 4.1.1 (2021-08-10). A two-sided p value of ≤0.05 was considered statistically significant.

Table 1.

Participant Characteristics

	Chronic OI, Group 1	Chronic OI, Group 2	Healthy Control
N	16	17	14
Sex	14 F, 2 M	15 F, 2 M	14 F
Dx *	Chest Pain (3), COVID-19 (2), Dizziness (4), EDS (4), Fibromyalgia (3), Joint Pain (2), MCAS (7), Migraine (8), Orthostatic Hypotension (2), Overactive Bladder (1), Presyncope (3), Syncope (2), Tachycardia (6)	Chest Pain (7), COVID-19 (3), Dizziness (9), EDS (6), Fibromyalgia (5), Joint Pain (4), MCAS (14), Migraine (11), Presyncope (5), Seizure Disorder (1), Sleep Apnea (1), Syncope (2), Tachycardia (6)	None
Meds *	Carvedilol (1), Cromolyn (2), Diltiazem (3), Droxidopa (1), Ivabradine (5), Ketotifen (2), Metoprolol (1), Naltrexone (2), Propranolol (4)	Cromolyn (1), Diphenhydramine (4), Famotidine (1), Fludrocortisone (1), Ketotifen (2), Mirabegron (1)	None
Age (years)	35±10 Range: 21–54	39±12 Range 22–71	31±6 Range: 22–48
Avg HR (bpm)	79±10 Range: 60–98	78±11 Range: 60–107	75±6 Range: 64–85
Avg SKNA (μV)	0.99±0.07 Range: 0.85–1.17	1.03 ±0.13 Range: 0.93–1.46	0.98 ±0.06 Range: 0.91–1.06

Avg = Average; COVID-19 = Coronavirus disease 2019; Dx = diagnosis; EDS = Ehlers-Danlos Syndrome; MCAS = Mast cell activation syndrome; Meds = medications

^*Includes only the most common diagnoses and medications.

The average age, HR, and SKNA were not statistically significant between groups (p=0.123, p=0.350, and p=0.421, respectively).

Results

Study 1: Postural changes, tachycardia and SKNA in chronic OI and control participants

The analyzed recording duration per participant was 1360±210 min and 1387±131 min for chronic OI and control participants, respectively (p=0.67). Participant characteristics are listed in Table 1.

Using an accelerometer to determine postural changes

In Study 1, the validity of the accelerometer posture criteria was investigated during active standing tests in 11 chronic OI participants. Figure 1A shows a recording of an active standing test where the participant was asked to lie supine, sit up (red dotted line), and then stand up. This recording shows that the participant’s HR increased by 24 bpm within 1 min of standing. Large bursts of SKNA were noted prior to the patient achieving an upright posture. Both postural changes and mental stress¹⁴ might have contributed to the SKNA elevation. Figure 1B shows recordings of a chronic OI participant in an ambulatory setting. This recording shows that upon standing, the participant had a 45 bpm HR increase. Both panels show that the postural changes were associated with a large increase of aSKNA (red arrows). For this study, we only analyzed postural change episodes where a movement to an upright posture was preceded by 5 min of uninterrupted lying posture.

Figure 1.

Accelerometer recordings during postural changes. The red dotted lines mark the beginning of a posture transition. A shows the ECG, accelerometer and SKNA recordings during an active standing test. The test procedure calls for an examination chair to raise the patient gradually from a supine to sitting position, after which the patient was asked to stand up. The slow pace of the postural change explains the gradual slope of the X accelerometer readings. B shows an ambulatory recording of an unwitnessed postural change documented by abrupt increases of X and Z accelerometer values. An anticipatory SKNA burst (first red arrow) preceded the changes of body posture.

HR and aSKNA Elevation During Postural Transitions

For all chronic OI participants in Study 1 who underwent an active standing test, the test increased HR by a maximum of 14±8 bpm. Only one participant had a ≥30 bpm HR increase (POTS) and the remaining 10 participants had PSWT based on the results of that test. In comparison, ambulatory ECG monitoring showed that 12/16 participants had at least one instance of a ≥30 bpm HR increase within the first min and 14/16 within the first 10 min after a postural change. The remaining two participants had maximum HR increases of 26 and 15 bpm after a postural change, respectively.

There were 124 episodes (8±4 per participant) of postural changes in chronic OI participants. Due to a limitation of the accelerometer, these episodes include body position changes from a lying to either a sitting or standing position. Among them, 11 (8.9%) were associated with symptoms. In comparison, 0 of 104 episodes (7±3 per participant) of postural changes in control participants were symptomatic (p=0.0011). Typical examples of the ambulatory recording in chronic OI and control participants are shown in Figures 2A and 2B, respectively. In both cases, body posture changes were associated with abrupt and sustained HR increases. Figures 2C and 2D show the aSKNA and HR during asymptomatic postural changes for all chronic OI and control participants, respectively. There was no significant difference found in the aSKNA and HR during the 5 min lying periods prior to a postural change (p=0.75 and p=0.09, respectively) between the two. The aSKNA increased significantly from the 5 min lying to 1 min upright posture periods for both the chronic OI (0.142 μV/min, p<0.001) and control participants (0.121 μV/min, p<0.001) during postural changes. However, there was no significant difference in the rates of the aSKNA increase between the two groups (p=0.462). HR also increased significantly from the 5 min lying to 1 min upright posture periods for both the chronic OI (13 bpm, p<0.001) and control participants (17 bpm, p<0.001) during postural changes. The rate of the HR increase in control participants was significantly faster than in chronic OI participants (p=0.0024). aSKNA then decreased significantly during the 3 min upright posture period for both the chronic OI (−0.04115 μV/min, p<0.001) and control participants (−0.0236 μV/min, p=0.0275). Although numerically the aSKNA of the chronic OI group decreased faster than the aSKNA of the control group, the differences were not significant (p=0.232). HR also decreased significantly during this period for both the chronic OI (−3 bpm, p<0.001) and control participants (−4 bpm, p<0.001). There was no significant difference in the rates of HR decrease between the chronic OI and the control participants (p=0.26). Ivabradine and beta blocker use did not have a significant effect on either the aSKNA (p=0.418 and p=0.408, respectively) or the HR (p=0.345 and p=0.277, respectively).

Figure 2.

Postural changes during ambulatory recordings in chronic OI and control participants. A and B show typical examples of posture transitions in chronic OI and control participants, respectively. The red dotted lines mark the beginning of a posture transition. The red arrows point to anticipatory SKNA bursts prior to postural changes. The blue segment of the upper panels are enlarged in the lower panels. C and D show the aSKNA and HR of all chronic OI and control participants studied, respectively. Asterisks indicate that there are significant increases of aSKNA and HR during postural changes (from lying to upright) in both groups of participants. However, no differences of aSKNA were observed between these two groups.

Postural Transitions and Symptoms in Chronic OI Participants

All chronic OI participants experienced symptoms during their recording (153 episodes total, 10±10 per participant). The most recorded symptoms were lightheadedness, heart palpitations, nausea, and dizziness. Only 82 of 153 episodes (53.6%) were associated with HR ≥100 bpm. Additionally, 87 of 153 episodes (56.9%) occurred without postural changes, including 28 (18.3%) that occurred during a continuous lying posture and 59 (38.6%) that occurred during a continuous upright posture. Figure 3 shows two typical examples of symptomatic episodes that occurred during continuous lying postures. Figure 3A displays two symptom markers (red dotted lines). At the beginning of the recording excerpt, there were robust SKNA discharges while the participant was upright. A transition to a lying posture is shown, and a symptom marker was pressed 3 min after lying down. Another symptom marker was recorded 8 min after the first one, after the participant had assumed a lying posture for 11 min. The participant reported dizziness and fatigue for the first symptom marker and continued dizziness and heart palpitations for the second symptom marker. Figure 3B shows a symptom marker after a movement from the right lateral recumbent to supine position, though there was no movement to an upright posture. There were robust aSKNA discharges and a HR elevation not produced by a postural change. Intermittent SKNA bursts and HR elevation were also observed in control participants, but none of them were symptomatic.

Figure 3.

Symptomatic episodes may occur without postural changes. The red dotted lines mark the exact times of symptom markers. A shows 2 episodes of symptoms while the patient was in the right lateral recumbent position. B shows active aSKNA and HR elevations while the patient was assuming a lying posture, emphasized by the green arrows. The symptomatic episode in this recording was not induced by a change from lying to upright posture. Rather, large SKNA bursts were observed prior to the symptom marker.

A total of 12 symptomatic episodes occurred within 5 min after a postural transition as defined above. Only 6 of the 16 chronic OI participants had these symptomatic episodes associated with postural transitions (2±1 episodes per participant). Between the symptomatic and asymptomatic postural transitions in these chronic OI participants, there was no significant difference found in the aSKNA during the 5 min lying period prior to a postural change (p=0.9894) and every minute for 3 min after the transition to upright posture (p=0.7793, p=0.8922, and p=0.2586 for the first, second, and third minutes, respectfully). No difference was found in the HR for the 5 min lying period and first minute upright (p=0.4402 and p=0.1689) as well. That said, a significant difference was observed in HR during the second and third minutes upright (p=0.0450 and p=0.0310, respectfully), with the HR higher during the symptomatic episodes in these periods.

Figure 4A shows an example of a symptomatic episode associated with a postural transition in a chronic OI participant. There were bursts of SKNA before the transition and symptom marker, though the HR did not increase ≥30 bpm until after the symptom marker. Figure 4B shows an example of a posture transition in a control participant, fulfilling the criteria for POTS with a ≥30 bpm HR increase but nonetheless asymptomatic. Figure 4C shows a symptomatic episode in a chronic OI participant, neither relating to a postural transition nor displaying a ≥30 bpm HR increase. Large aSKNA bursts were seen in all 3 examples (arrows). The accelerometer readings of the remaining 54 symptomatic episodes (3±5 per participant) were consistent with small sporadic movements rather than postural changes.

Figure 4.

SKNA bursts are seen in both symptomatic and asymptomatic episodes. The red dotted lines mark the exact times of symptom markers. Anticipatory SKNA bursts are noted prior to achieving an upright posture. A is a recording of a symptomatic episode associated with a postural change in a chronic OI participant. A symptom was indicated shortly after the postural change. Note the SKNA burst prior to the posture change and the >30 bpm sustained increase in HR following the posture change and symptom marker. B is a recording of a postural change in a control participant. There is also an SKNA burst and >30 bpm sustained increase in HR associated with the postural change in this recording. However, there was no indication of symptoms. C shows a symptomatic episode not associated with a posture change in a chronic OI participant. There are several bursts in SKNA shown, but there was no increase in HR ≥30 bpm.

SKNA burst analyses of asymptomatic and symptomatic episodes

For chronic OI participants, symptomatic recording periods were defined ±1 minute from a symptom marker. The rest of a participant’s total recording was deemed asymptomatic. Figure 5A shows that the 24 hour aSKNA of a chronic OI participant can be fitted into two Gaussian distributions. The mean and SD of the first Gaussian distribution in this figure are 0.8659 μV and 0.0674 μV, respectively. To calculate burst thresholds, 3 SD were added to the mean of the first Gaussian distribution, resulting in a threshold of 1.0681 μV for this participant (blue dotted line).^{12, 15} Using this threshold, we identified the burst area (red) in asymptomatic (Figure 5B) and symptomatic (Figure 5C) 2 min windows of the same participant. Figure 6A shows the results of SKNA burst analysis between the asymptomatic and symptomatic recording periods for all chronic OI participants studied. The burst frequencies were 0.59±0.14 bursts/min and 1.13±0.29 bursts/min (p<0.001). The burst amplitudes were 1.29±0.14 μV and 1.28±0.15 μV (p=0.565). The burst durations were 17.1±11.0% and 35.2±22.8% (p<0.001). The mean burst areas were 0.061±0.033 μV*sec and 0.124±0.085 μV*sec (p=0.001), respectively.

Figure 5.

Symptomatic episodes are associated with increased SKNA bursts. A shows that the distribution of aSKNA over a 24 hour period can be fitted into two Gaussian distributions. The burst threshold was selected to be the mean plus 3 SD of the first Gaussian distribution. B and C show SKNA bursts during asymptomatic and symptomatic episodes, respectively. Note that SKNA bursts during symptomatic episodes have a longer duration than asymptomatic bursts. The red dotted line shows the exact time of the symptom marker, which occurred without changing from a lying to upright position, as indicated by the X acceleration below 400 mG.

Figure 6.

SKNA burst analysis in all chronic OI and control participants. A shows a comparison of SKNA burst frequency, amplitude, duration, and mean area between asymptomatic (asymp.) and symptomatic (symp.) periods in chronic OI participants. B shows a comparison of SKNA burst frequency, amplitude, duration, and mean area between control and chronic OI participants throughout their entire 24 hour ambulatory recordings. There were no statistically significant differences observed between any of these comparisons.

SKNA burst analyses of chronic OI and control participants

For chronic OI and control participants, the baseline SKNAs were 0.916±0.049 μV and 0.906±0.037 μV (p=0.514), and the burst threshold SKNAs were 1.092±0.110 μV and 1.059±0.095 μV (p=0.391), respectively. Though the means of both measures were numerically higher in POTS participants, no statistically significant differences were found. Figure 6B shows the SKNA burst frequency, amplitude, duration, and mean area between chronic OI and control participant groups throughout their entire recordings. For the chronic OI and control participant groups, the SKNA burst frequencies were 0.59±0.14 bursts/min and 0.62±0.12 bursts/min (p=0.544), the burst amplitudes were 1.29±0.14 μV and 1.25±0.15 μV (p=0.410), the burst durations were 17.4±11.6% and 20.1±12.7% (p=0.545), and the mean burst areas were 0.063±0.035 μV*sec and 0.066±0.038 μV*sec (p=0.774), respectively.

Study 2

The analyzed recording duration per participant was 1372±211 min in these 17 participants. Their clinical characteristics are listed in Table 1. Among them, 13 completed an active standing test in clinic. The test increased HR by a maximum of 16±8 bpm. No participant had a ≥30 bpm HR increase during their active standing test. In comparison, ambulatory ECG monitoring of the same 17 patients showed that 14 (82.4%) had at least one instance of a ≥30 bpm HR increase within the first min and first 10 min after a postural change. These 14 patients would fulfill the diagnostic criteria of POTS.³

There were 180 episodes (11±8 per participant) of postural changes. 11 (6.1%) were associated with symptoms. Among 110 symptomatic episodes (6±5 per participant), 98 (89.1%) had at least 1 SKNA burst, but only 41 (37.3%) had HR exceed 100 bpm 1 min prior to the symptom marker (p<0.0001). Of these episodes, 38 (34.5%) displayed both an SKNA burst and HR over 100 bpm. Only 10 (9.1%) symptomatic episodes had neither an SKNA burst nor tachycardia 1 min prior to the symptom marker. The symptoms occurred during a continuous upright posture in 78 (70.9%), during a continuous lying posture in 9 (8.2%), and during a postural transition in 11 episodes (10%). The remaining 12 (10.9%) episodes occurred during sporadic movement that could not be associated with a postural change or continuous posture.

We performed a comparison between the results of Studies 1 and 2. We found that 82/153 (53.6%) symptoms were associated with tachycardia (>100 bpm) in Study 1 versus 41/110 (37.3%, p= 0.0121) in Study 2. However, there were no statistically significant differences in the maximum increase of HR between Study 1 (14±8 bpm) and Study 2 (16±8 bpm, p=0.6278) during active standing tests. The presence of SKNA bursts before symptom markers was noted in 132/153 (86.3%) symptoms in Study 1 versus 98/110 (89.1%, p=0.5736) in Study 2. More symptomatic episodes occurred during a continuous upright posture than a lying posture in both studies (38.6% versus 18.3% in Study 1 and 70.9% versus 8.2% in Study 2). A total of 11/153 (7.2%) symptoms were associated with a clean postural change in Study 1 versus 10/110 (9.1%) in Study 2 (p=0.6472). Active standing tests were done in 11 patients in Study 1 and 13 in Study 2. One patient displayed a reduction of BP of 21 mmHg upon standing in one of their 2 clinic visits. None of the other patients had a >20 mmHg reduction of BP upon standing.

Discussion

We found that the symptoms of chronic OI are more often associated with SKNA bursts than tachycardia. While symptoms of chronic OI primarily occur during postural changes or standing, a notable number of the symptoms occurred while lying flat. In comparison, control participants also had significant increases of SKNA and HR during postural changes. However, they remained asymptomatic.

Using an accelerometer to determine postural changes

The validity of using triaxial accelerometers to measure human movement is well documented in literature.¹⁶ The present study shows that the triaxial accelerometer, alongside ECG monitors, has the potential to be utilized as a useful diagnostic and research tool for patients with chronic OI.

The Diagnosis of POTS

The current diagnostic criteria for POTS include chronic OI and a ≥30 bpm sustained increase in HR with orthostatic tests.^{2, 3} However, neither active nor passive orthostatic testing is highly reproducible, and many patients with chronic OI do not have ≥30 bpm HR increases during orthostatic tests.^{7, 8} Raj et al proposed that patients with chronic OI can be subdivided into POTS and PSWT.⁸ The implication is that tachycardia is not an essential feature of this illness. In Study 1, we found only one chronic OI participant out of 11 had a HR increase ≥30 bpm during their clinical active standing test, but ambulatory ECG monitoring showed that ≥30 bpm HR increases were observed in 14/16 participants. Also notable is that the postural symptoms may occur before HR increases by 30 bpm. Furthermore, it is common for asymptomatic control patients to experience both SKNA elevation and ≥30 bpm HR increases during postural changes. These findings suggest that tachycardia is neither a sensitive nor a specific biomarker for chronic OI.

Sympathetic nerve activity in chronic OI and POTS

Chronic OI patients had higher baseline muscle SNA but lower muscle SNA response during head up tilt tests.⁹ In comparison, Fu et al¹⁷ showed that head-up tilt tests increased the muscle SNA more in POTS patients than in controls. Due to the technical difficulties, none of these microneurography studies were done in ambulatory patients. Even less is known about the association between skin SNA and POTS, as the skin and muscle SNA are independently controlled.^{18, 19} In this study, we did not find a quantitative difference of aSKNA between control and chronic OI participants during postural changes. However, the burst frequency, duration and mean burst area during symptomatic episodes were much higher than during asymptomatic episodes in chronic OI participants. Very few episodes of symptoms occurred without SKNA bursts 1 min prior to the symptom marker. These data indicate that SKNA burst(s) may be a highly specific albeit insensitive biomarker for chronic OI symptoms

Mechanisms of symptoms

Because SKNA bursts were invariably asymptomatic in control participants, the burst alone is not the cause of symptoms. Interoception is the sense of the physiological condition of the body.²⁰ Previous studies have suggested that interoception may play an important role in OI,^{21, 22} but we do not have data in the present study to determine if abnormal interoception causes OI symptoms. Cerebral hypoperfusion plays an important role in POTS.²³ It is possible that cerebral hypoperfusion associated with postural changes might have produced both the symptoms and the SKNA bursts. Small fiber neuropathy is common in patients with POTS.²⁴ It is possible that the association between SKNA bursts and symptoms are in part due to the pathophysiology associated with small fiber neuropathy. However, that association was not investigated in the present study.

Limitations of the study

Our study population is highly heterogenous. In this study, some participants had a history of coronavirus disease 2019 (COVID), and others had various associated comorbid conditions such as EDS and MCAS. While Study 2 patients were not treated with ivabradine, pyridostigmine, or beta blockers, they may have been taking other drugs that can indirectly affect HR or BP. Therefore, our results may not be applicable to all patients with POTS or chronic OI. Another limitation is that a single triaxial accelerometer cannot differentiate the sitting and standing body positions.²⁵

Conclusions

Postural changes increase aSKNA and HR in both control and chronic OI participants, but only patients with chronic OI have symptoms during postural changes. Among patients with chronic OI, the symptoms are strongly associated with SKNA bursts but only weakly associated with tachycardia. These findings suggest that SKNA bursts may be an insensitive but highly specific biomarker for chronic OI.