1. Introduction
Rheumatoid Arthritis (RA) and Systemic Lupus Erythematosus (SLE) are chronic autoimmune diseases with systemic inflammation. RA affects 0.5–1.0% of adults in industrialised countries and the prevalence of SLE in adults ranges from 30 to 150 per 100,000, both diseases being more common in women [
1,
2]. Many therapeutic agents exist for treating RA and SLE. However, SLE-treatment still has high failure rates and toxicity, and SLE is associated with premature mortality [
2]. Likewise, patients with RA have higher mortality compared to the general population [
3], and despite different treatment options there are still patients with RA who continue to have signs and symptoms suggestive of inflammatory disease activity [
4]. In addition, pain often remains problematic even in low levels of inflammation [
5] and many patients experience side effects from the medicine, which can lead to withdrawal of the treatment [
6]. These circumstances call for novel treatment strategies in the management of RA and SLE.
Studies have investigated the autonomic nervous system and found that patients with RA and SLE have an imbalance of their autonomic nervous system [
7,
8]. A systematic literature review by Adlan et al. found that 60% of patients with RA have an autonomic nervous system dysfunction—especially reduced parasympathetic activity and altered heart rate variability (HRV) [
8]. HRV is a method used to evaluate cardiac autonomic nervous system function [
8], especially vagus nerve tone [
9]. A high HRV indicates good health and vitality [
10], and reduced HRV, which is seen in patients with RA and SLE, is associated with increased mortality [
7,
8]. Autonomic imbalance precedes the development of RA in individuals at risk of developing arthritis [
11], and seropositivity, disease activity, and pro-inflammatory cytokine levels are predictive of autonomic dysfunction [
12]. Patients with more vagal activity generally respond better to anti-tumour necrosis factor therapy, and in healthy subjects decreased parasympathetic activity is also correlated with increased inflammatory status [
6,
13]. Studies also confirm the presence of autonomic dysfunction measured by HRV in patients with SLE [
14,
15].The link between the autonomic nervous system and the immune system has been investigated previously to explain the autonomic dysfunction. In this context, the vagus nerve has been investigated as the possible missing link [
16].
The vagus nerve is part of the parasympathetic nervous system and controls visceral functions. However, recent data have suggested an anti-inflammatory role of the vagus nerve as well, which is believed to be mediated through three different pathways: (a) The hypothalamic-pituitary-adrenal axis, (b) the cholinergic anti-inflammatory pathway, and (c) the splenic sympathetic anti-inflammatory pathway [
16]. The outcomes are inhibition of the pro-inflammatory cytokine tumor necrosis factor-
or increased cortisol, which results in an anti-inflammatory effect [
16]. Knowing that patients with RA and SLE have an autonomic dysfunction, especially decreased parasympathetic function, an interesting perspective could be to investigate if stimulation of the vagus nerve could exert an anti-inflammatory effect and thereby improve disease activity in these patients.
One possible way of upregulating the vagus nerve tonus could be through vagus nerve stimulation (VNS). VNS is currently being used in the treatment of refractory epilepsy, migraine, cluster headache, and depression [
17,
18]. VNS can be both invasive and non-invasive. Invasive techniques involve implantable devices with a cuff electrode around the vagus nerve on the neck and non-invasive techniques include transcutaneous VNS (tVNS) of the auricular or the cervical branch of the vagus nerve. A study by Koopman et al. found that 21 days of invasive VNS through an implantable vagus nerve stimulator in 17 patients with RA inhibited tumor necrosis factor production and decreased the disease severity significantly [
17]. However, when using invasive techniques such as surgical implantation of a VNS device there is always a risk associated with the general anaesthesia and a risk of postoperative complications [
19]. Physiological methods of stimulating the vagus nerve also exist, such as deep breathing (DB), yoga, or meditation [
20,
21,
22]. Therefore, if physiological methods were to be effective, they would often be preferred, due to fewer adverse events, no financial costs, and their transferability into clinical practice.
A way to modulate the vagus nerve without invasive procedures and equipment is through DB, where the vagus nerve is stimulated through the baroreflex: When blood pressure increases, the arterial baroreceptors are activated. This results in activation of the vagus nerve, which signals to the sinoatrial node in the heart, resulting in decreased heart rate. The same mechanisms are activated during DB, due to changes in the intrathoracic pressure, thereby affecting blood pressure and ultimately resulting in increased vagus nerve tone and increased HRV [
23].
Several studies have investigated both yoga and breathing exercises in healthy participants and found an increase in HRV-parameters [
24,
25,
26]. In a cross-over clinical trial with 25 healthy participants, Sharpe et al. found that different breathing interventions (2 × 10 min) increased the time domain HRV-parameters [
24]. In 21 healthy participants, Tavares et al. found an increase in the time- and frequency domain HRV-parameters after guided breathing for 10 min [
25]. In a cross-sectional study with 14 yoga practitioners and 14 non-yoga practitioners, Muralikrishnan et al. found that the time- and frequency domain HRV-parameters increased significantly more in yoga practitioners during one minute of controlled deep breathing [
26]. Despite several studies having found increases in HRV-parameters after breathing exercises, others have found a lower parasympathetic nerve activity after controlled breathing in healthy subjects [
27]. Sasaki et al. reported that five minutes of controlled breathing in 20 healthy subjects led to a reduced parasympathetic nerve activity evaluated through the frequency domain parameters [
27]. To our knowledge, only one study investigated DB in patients with RA and SLE and found an increase in HRV [
22]. Likewise, a study by Juel et al. found that the combination of deep slow breathing and tVNS could increase cardiac vagal tone in patients with chronic pancreatitis [
20]. The field is therefore relatively unexplored, and no studies have examined for how long the DB exercises must be performed to modulate HRV in healthy participants and in patients with RA and SLE and the reliability of the method.
The aims of this study were to examine the dose-response of DB on HRV in healthy participants by investigating DB in 5, 15, and 30 min and to examine the reliability of the effect of DB on HRV in patients with RA and SLE. We hypothesize that the HRV increases when the duration of the DB is longer, and that a consistent increase in HRV is observed across days. Thus, the contribution of this paper is two-fold: (1) an investigation of the dose-response relationship of DB and the effect on HRV in healthy participants, and (2) the test-retest reliability of DB in patients with RA or SLE to investigate if similar changes are observed across days.
4. Discussion
In this study the optimal dose of DB in healthy participants and the reliability of the effect of DB on HRV in patients with RA and SLE were investigated for the first time. No significant change was found in the measurements obtained before the intervention, indicating that HRV-parameters do not change while resting. An increase in HRV-parameters was found in healthy participants already after 15 min of DB, but more certainly after 30 min of DB. In patients with RA and SLE, 30 min of DB increased all HRV-parameters, however, the effect on PNN50 had a slower onset. The findings indicated reliability between the two intervention days, and the effect remained throughout the post-measurements, indicating that the effects outlasted the VNS period.
Looking at the baseline measurements, the patients with RA and SLE had lower SDNN (31 milliseconds) and RMSSD (24 milliseconds) compared to normal values reported for healthy adults (SDNN: 50 milliseconds and RMSSD: 42 milliseconds) [
34]. The healthy participants in this study had a baseline SDNN (57 milliseconds) and RMSSD (43 milliseconds) comparable to normal values in healthy adults [
34]. This supports the fact that patients with RA and SLE have a lower tone of the autonomic nervous system measured through HRV, and other studies confirm similar results of depressed HRV [
7,
8,
15].
Thirty minutes of DB increased HRV-parameters in both healthy participants and in patients with RA and SLE. Previous studies in healthy participants (see
Table 6 for an overview) also found an increase in HRV after yoga and breathing exercises [
24,
25,
26,
35]. However, a study in healthy participants by Sasaki et al. found that an HRV-parameter of parasympathetic nerve activity decreased during controlled breathing [
27]. It is difficult to compare results across the different studies since they had different overall aims. In addition, different measures of vagal tone have been used as well as differences in breathing parameters such as inspiration–expiration ratio and duration of the DB intervention, but generally DB interventions have been associated with increased vagal tone and parasympathetic activity. Moreover, other factors can influence HRV such as (1) the length of the recording period (longer recordings are associated with increased HRV), (2) removal of artefacts (e.g., extrasystoles) and (3) age, gender, heart rate, health status, and comorbidities [
9,
36]. In patients with RA and SLE, the findings in the current study are in agreement with a similar study using the same intervention and HRV-parameters [
22]. Similar increases in SDNN and PNN50 were observed, but the absolute RMSSD values were slightly lower in the current study, which could be attributed to inter-individual differences.
To our knowledge, no other studies examined the dose-response relationship of DB on HRV in healthy participants. One study found an increase in HRV during breathing at a frequency of 5–7 breaths/minute for 2 min [
10] and another study during 10 min of guided breathing exercises [
25]. However, both studies measured HRV during the breathing exercises, which is debatable since respiration itself can be a confounding factor in HRV evaluation [
37]. The influence of respiration on HRV is a phenomenon known as respiratory sinus arrhythmia (RSA), and it is characterized by shortening of the RR-interval during inspiration and prolongation of the RR-interval during expiration [
38]. In this work HRV was measured after the intervention to avoid the direct influence of RSA.
tVNS is another possible way of stimulating the vagus nerve non-invasively. A study in healthy participants found that 15 min of auricular tVNS increased HRV [
39]. Brock et al. investigated cervical tVNS for 120 s, 3 times a day, for 5 consecutive days in patients with psoriatic arthritis and found a reduction in disease activity and a 20% reduction of CRP [
40]. A pilot study of tVNS in patients with RA showed similar results [
41]. These studies of tVNS could indicate that the required amount of stimulation time is smaller when using tVNS compared to DB. Compliance may also be better ensured when using tVNS, since 30 min of DB demands complete concentration.
The invasive alternative to DB and tVNS is an implantable vagus nerve stimulator, which was investigated in a study by Koopman et al. [
6]. They found that an implantable vagus nerve stimulator could inhibit tumour necrosis factor and interleukin-6 production and improve disease severity in patients with RA, indicating a possible correlation between the immune system and the autonomic nervous system [
6]. However, an implantable vagus nerve stimulator has known side effects such as hoarseness, dysphonia, and coughing [
6], so physiological methods such as DB have some clear advantages such as accessibility, fewer side effects, and no costs.
The present study found increased HRV-parameters after DB in patients with RA and SLE, indicating a higher parasympathetic tone, which may have implications for the management and treatment of RA and SLE. It has been reported that autonomic imbalance precedes the development of RA in individuals at risk of developing arthritis [
11], and patients with more vagal activity respond better to anti-tumour necrosis factor treatment [
13]. Thus, DB could potentially be used to make RA and SLE patients respond better to their medication and hence improve the management of the diseases or potentially prevent or postpone development of the diseases. However, these are just speculations and should be tested in future studies. It is not known if the vagus nerve and its anti-inflammatory properties are activated with resulting decreased levels of pro-inflammatory cytokines, e.g., tumour necrosis factor-
.
Limitations and Future Perspectives
Five, 15, and 30 min of DB were investigated in healthy participants, while only 30 min of DB was investigated in patients. Consequently, it is unknown whether doses in between are effective, and future studies could investigate the dose-response more thoroughly both in patients and healthy participants by changing the interventional dose, e.g., 20 min and 25 min. In continuation of this, it is also important to consider how often and how long the DB should be performed, both concerning compliance and to maintain an adequate effect. It is still uncertain if the effect is sustained beyond 30 min, which is why further examination of the washout effect should be carried out. That could be done simply by adding additional measurements after 30 min.
This research area is still relatively unexplored, and taking the above-mentioned into consideration, future studies could be designed as longitudinal interventional studies, preferably as a randomized controlled trial with the control group performing un-paced (sham) breathing. The interventional effect could be examined in relation to clinical parameters such as disease activity scores and biochemical measures such as CRP and cytokine levels.
HRV is an indicator of cardiac vagal tone, which does not necessarily reflect the anti-inflammatory properties of the vagus nerve. It is not known whether the vagus nerve might be organ specific, meaning that vagal influence on the heart may not represent vagal input to inflammatory organs such as the spleen, lungs, gut, or liver as well [
42]. Therefore, clinical, and biochemical measures could contribute to clarify the potential anti-inflammatory effect of DB. Moreover, future studies should include patients with remission, low-grade disease activity, and high-grade disease activity to make conclusions applicable for a wider spectrum of the patient groups. In addition, the intervention should be tested on a larger number of SLE patients to determine the efficacy of the intervention in this patient group alone since the majority of patients in the current study were diagnosed with RA, which was due to prevalence of RA and SLE patients.
In summary, the future research direction could be: (1) to investigate the parameters of the intervention to maximize the effect, (2) investigate the effect on disease activity and biochemical measures in randomized clinical trials, (3) investigate the exact physiological mechanisms associated with the DB intervention, and (4) investigate how adherence to the DB training can be maximized and made motivating and engaging to perform for the patients.