Article Text

Original article
Preliminary report: modulation of parasympathetic nervous system tone influences oesophageal pain hypersensitivity
  1. Claude Botha1,
  2. Adam D Farmer1,
  3. Matias Nilsson2,
  4. Christina Brock2,
  5. Ana D Gavrila1,
  6. Asbjørn Mohr Drewes2,
  7. Charles H Knowles1,
  8. Qasim Aziz1
  1. 1Centre for Digestive Diseases, Blizard Institute of Cell & Molecular Science, Wingate Institute of Neurogastroenterology, Barts and the London School of Medicine & Dentistry, Queen Mary University of London, London, UK
  2. 2Mech-Sense, Department of Gastroenterology, Aalborg University Hospital, Aalborg, Denmark
  1. Correspondence to Dr Adam D Farmer, Centre for Digestive Diseases, Blizard Institute of Cell & Molecular Science, The Wingate Institute of Neurogastroenterology, Barts and the London School of Medicine and Dentistry, 26 Ashfield Street, Whitechapel, London E1 2AJ, UK; a.farmer{at}qmul.ac.uk

Abstract

Objectives Autonomic nervous system dysfunction has been implicated in visceral hypersensitivity. However, the specific contribution of the parasympathetic nervous system (PNS) is unclear. We aimed to determine whether physiological and pharmacological manipulation of parasympathetic tone influences the development of hypersensitivity in a validated model of acid-induced oesophageal pain.

Design Prior to, and following, a 30-min distal oesophageal infusion of 0.15 M hydrochloric acid, pain thresholds to electrical stimulation were determined in the proximal non-acid exposed oesophagus in healthy subjects. Validated sympathetic (skin conductance response) and parasympathetic (cardiac vagal tone) parameters were measured at baseline and continuously thereafter. In study 1, 55 subjects were randomised in a pragmatic blinded crossover design to receive deep breathing or un-paced breathing during acid infusion. In study 2, 32 subjects were randomised in a blinded, crossover design to receive intravenous atropine or placebo (saline) with deep breathing during acid infusion.

Results Study 1: Deep breathing increased cardiac vagal tone (2.1±2.3 vs −0.3±2.3, p=0.0006) with concomitant withdrawal of skin conductance response (−0.6±4.9 vs 3±4.8, p=0.03) in comparison with un-paced breathing. Deep breathing prevented the development of acid-induced oesophageal hypersensitivity in comparison with sham breathing (p=0.0001). Study 2: Atropine, in comparison with placebo, blocked the attenuating effect of deep breathing on the development of acid-induced oesophageal hypersensitivity (p=0.046).

Conclusions The development of oesophageal hyperalgesia is prevented by physiologically increasing parasympathetic tone. This effect is pharmacologically blocked with atropine, providing evidence that the PNS influences the development of oesophageal pain hypersensitivity.

  • Abdominal Pain
  • Visceral Sensitivity
  • Visceral Hypersensitivity
  • Visceral Nociception
View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known on this subject?

  • Visceral pain hypersensitivity is a characteristic feature of functional GI disorders which is mediated and maintained, in part, by the autonomic nervous system.

  • Deep breathing is thought to increase parasympathetic nervous system tone and is frequently used in many psychological, cognitive and alternative therapies. Such therapies have an evidence base in the treatment of chronic visceral pain syndromes such as irritable bowel syndrome.

  • Accumulating evidence suggests that the parasympathetic branch of the autonomic nervous system may mediate an antihyperalgesic effect through the cholinergic anti-inflammatory pathway.

What are the new findings?

  • We demonstrate that physiologically elevating parasympathetic nervous system tone, using deep breathing, prevents the development of oesophageal pain hypersensitivity in a validated human model.

  • This antihyperalgesic effect of deep breathing is ameliorated by antagonising the rise in parasympathetic nervous system tone with concomitant administration of atropine.

How might it impact on clinical practice in the foreseeable future?

  • These data provide preliminary human evidence for the mechanism of action of psychological therapies that use deep breathing techniques in chronic visceral pain disorders.

  • These data provide the mechanistic rationale for a novel line of investigation of therapies directed at modulating parasympathetic tone in disorders where oesophageal pain hypersensitivity is prominent, such as non-erosive reflux disease.

Introduction

Visceral pain occurs in both organic and functional GI disorders (FGID) and is a major global cause of disability, healthcare seeking and leading to reduction in quality of life.1 The experience of visceral pain is highly individual, both in health and disease, with a multitude of factors proposed to account for this variability.2 ,3 Among these is dysfunction of the autonomic nervous system (ANS), a bidirectional, hierarchically controlled brain–body nexus that integrates the external environment with the internal milieu. The ANS has been postulated to play a critical role in the modulation of pain through its multiple interactions with the nociceptive system at the level of the periphery, spinal cord, brainstem and forebrain.4 Considering that the parasympathetic nervous system (PNS) is broadly antinociceptive and the sympathetic nervous system (SNS) pro-nociceptive, the implication arises that a balance between them is required for normal pain perception.5 Indeed, accumulating evidence suggests that such an imbalance may represent an important pathophysiological factor in disorders where visceral pain and inflammation are a prominent feature.6 Previously, we have developed and validated a human model of oesophageal pain hypersensitivity where following acid infusion into the distal oesophagus, pain thresholds (PTs) to electrical stimulation in the non-acid exposed proximal oesophagus are reduced due to central sensitisation.7 However, not all individuals who are exposed to this paradigm develop sensitisation, defined as less than 6 mA drop in PT to electrical stimulation, thus suggesting the presence of inter-individual factors that mediate postacidification gut sensitisation.8 These factors are incompletely defined and understood but a plausible postulation centres on the involvement of physiological aspects.9 For instance, it has been demonstrated that oesophageal acidification is associated with a rise in SNS and fall in PNS activity, with subjects who withdrew their PNS tone the most developing heightened sensitivity.9

Despite progress in the identification of the culpable molecular mechanisms that account for development and maintenance of the visceral pain hypersensitivity, translation into efficacious pharmacotherapeutic agents has remained limited, notwithstanding concerns regarding safety in some cases.10 ,11 Therefore, it is not surprising that a number of alternative/cognitive interventions have been used with some success in the management of visceral pain syndromes.12 A common feature of many of these interventions is conscious control of breathing frequency and depth. Deep breathing has been proffered as a method of inducing analgesia, possibly through increasing PNS tone through activation of the baroreceptor reflex.13 Moreover, an increasing body of evidence from animal studies has proposed the importance of the PNS, or as is more commonly referred to in the literature ‘cholinergic tone’, as a critical mediator for the inflammatory–anti-inflammatory balance.14 Taken together, these data suggest that the PNS may have antihyperalgesic and analgesic properties in the human viscera. In the current study, we have tested this hypothesis by performing physiological and pharmacological modulation of PNS tone in a cohort of healthy subjects who sensitise to the aforementioned model of oesophageal pain hypersensitivity.

Materials and methods

Study population

Healthy adult subjects, aged 18–60 years, took part in the study. Subjects were non-smokers with no past medical history and were not taking any medications. Women were all studied in the follicular phase of their menstrual cycle. Subjects were asked to refrain from alcohol consumption for 24 h prior to the study. All subjects were screened for subclinical anxiety and depression using the validated Hospital Anxiety and Depression Scale.15 Written informed consent was obtained after the nature and the purpose of the trial had been explained. All subjects were naive to the experimental protocol and had never previously been subjected to this model of acid perfusion. All protocols were approved by the City and East London Research Ethics Committee, UK (reference 09/H0704/71) and the Ethics Committee for North Jutland, Denmark (reference N-20120065).

Main measurements and definition of terms

Oesophageal PT

Oesophageal sensory testing was performed via a pair of silver–silver chloride bipolar ring electrodes (inter-electrode distance 1 cm) sited 16 cm proximal to the tip of a 3 mm diameter catheter (Unisensor AG, Ch-8544 Attikon, Switzerland). Following identification of the lower oesophageal sphincter using standard station pull-through manometry, electrical stimulation was performed 18 cm proximal to the lower oesophageal sphincter. Electrodes were connected to an electrical stimulator (Model DS7, Digitimer, Welwyn Garden City, UK) and stimuli were delivered at a frequency of 0.5 Hz, using square wave pulses (0.5 s duration), at intensities varying between 0 and 80 mA.7 ,10 The intensity of stimulation was increased in an incremental fashion by 2 mA and each subject was asked to report both the sensory threshold (visual analogue scale (VAS) of 1 out of 10) and when they could not tolerate any further increase (VAS of 7 out of 10), that is, PT. The electrical stimulation was immediately stopped when PT was reached.

Definition of sensitisers versus non-sensitisers

Sensitisers were defined as having a postacid infusion reduction in upper oesophageal PT of ≥6 mA at one or more time points, as previously defined by Sharma et al.16 Non-sensitisers were defined as having a no reduction or a reduction of <6 mA in proximal oesophageal PT, at any time point after acidification.

Oesophageal pH

A twin-channel pH catheter (Synectics Medical, Enfield, UK) measured pH in both the proximal oesophagus (the site of electrical stimulation) and in the distal oesophagus (the site of acid infusion) for the duration of each study. Recordings were made using a twin-channel pH box (Synectics Medical).

Psychological factors

At baseline, state and trait anxiety was assessed using the validated Spielberger Anxiety Inventory.17

Autonomic nervous system

SNS: skin conductance responses

Skin conductance is a putative sympathetic ‘emotional sudomotor’ measure response within milliseconds to threatening stimuli.18 Skin on the distal digit pulp of the right index and ring fingers was wiped with water and allowed to dry. In each subject, skin conductance electrodes were then attached and the skin conductance level was zeroed using a commercially available bioamp (Powerlab, AdInstruments, Oxford, UK). The mean skin conductance response was extracted and analysed offline.

PNS: cardiac vagal tone

ECG electrodes (Ambu Blue Sensor P, Denmark) were placed in right and left subclavicular areas and cardiac apex. The ECG was acquired using a commercially available biosignals acquisition system (Neuroscope, Medifit Instruments, Essex, UK).19 In brief, the Neuroscope non-invasively measures brainstem PNS efferent activity, known as cardiac vagal tone, in real time using inbuilt ‘voltage controlled oscillators’ that detect phase shifts in the beat-to-beat RR interval in a process known as ‘phase-shift demodulation’20 (see online supplementary material). Cardiac vagal tone is measured on a validated linear vagal scale, where 0 represents full atropinisation.21 Autonomic parameters were recorded according to internationally agreed recommendations.22

Main exposures

Oesophageal acid infusion

Four 60 mL disposable syringes were preloaded with the hydrochloric acid (HCl) (Stockport Pharmaceuticals, Stockport, UK and Hospital Pharmacy, Aalborg University Hospital, Aalborg, Denmark), and warmed to body temperature prior to infusion. An amount of 0.15 M HCl was infused into the lower oesophagus, 3 cm above the lower oesophageal sphincter, through a 1 mm diameter twin-channel pH catheter (Synectics Medical) at a constant rate of 8 mL/min for 30 min via an infusion pump (KDS Scientific 100, Linton Instrumentation, Pulgrave, UK).

Main interventions

Deep and un-paced (sham) breathing

The deep breathing protocol used in this study was based on a modified procedure described by Thijs et al,23 consisting of breathing at full inspiratory capacity for 4 s, followed by exhaling to forced expiratory vital capacity in 6 s, repeated at a frequency of 0.1 Hz (ie, 6 breaths/min for 1 min every 5th min (ie, 6 cycles in total)) during the 30-min duration of the acid exposure phase. This method was chosen to increase cardiac vagal tone repeatedly during the acid exposure. During sham breathing, an investigator engaged with the subjects in a standardised fashion during which they were asked to focus on their breaths and count the number of their breaths they took during the acid infusion, that is, un-paced, spontaneous breathing, to control for any effect of distraction that could be attributed to deep breathing.24 The randomisation of subjects to the deep breathing or sham/un-paced breathing arms was performed by means of approved statistical software (http://www.randomization.com).

Intravenous atropine and placebo

An amount of 0.5 mg atropine sulfate (International Medication Systems, Slough, UK), a potent anticholinergic through its action as a competitive inhibitor of acetyl-choline at muscarinic receptor sites, was administered via an indwelling 18GA intravenous cannula (Venflon, Becton Dickinson, Oxford, UK) sited in the left ante-cubital vein, as a slow bolus over 2 min. A similar labelled syringe containing 0.9% normal saline was administered as placebo, in an identical fashion, over 2 min. All subjects had continuous monitoring of cardiovascular parameters during and after administration of both atropine and placebo.

Primary outcome measures

Three measurements of PT were taken in the proximal oesophagus and the mean value derived. Measurements were taken prior to acid infusion (T0), then 60 min (T60), 90 min (T90) and 120 min (T120) after completion of the acid infusion.

Study protocols

All subjects were studied in the morning (from 09:00 to 12:00) in a temperature controlled (20–22°C), quiet laboratory. All experiments were conducted with subjects resting on an examination couch at 45°, having fasted for a minimum of 6 h. The catheter and pH probe assembly was passed trans-nasally, without the aid of local anaesthetic, into the oesophagus until the infusion port was 3 cm, and the stimulating electrodes were 18 cm proximal to the lower oesophageal sphincter. The pH sensors were located adjacent to the infusion and stimulation sites. In all experiments, blood pressure, pulse and cardiac vagal tone were measured at baseline and continuously thereafter. Skin conductance response was measured at baseline and continuously thereafter in study 1 but was not measured in study 2.

Study 1: The effect of physiological manipulation of autonomic tone with deep breathing

A total of 55 healthy subjects (31 men, mean age 26 years, range 18–48 years) were recruited from the London study centre (Queen Mary, University of London, London, UK). Baseline PT was determined in the proximal oesophagus, prior to acid infusion (ie, T0). Acid was then infused into the distal oesophagus for 30 min, during which subjects were randomised to undertake either deep breathing or sham/un-paced breathing. PTs were measured in the proximal oesophagus prior to acid infusion and then at 60, 90 and 120 min afterwards (ie, at T60, T90 and T120). Subjects who received sham/un-paced breathing at the first visit who did not sensitise were excluded from the study at that point without being exposed to deep breathing. Subjects were then restudied after a period of 2 weeks, during which they exposed to the breathing intervention to which they had not been exposed during the first visit. Subjects who received sham/un-paced breathing at the second visit who did not sensitise were excluded from the subsequent study analysis. In total, 18 healthy subjects (13 men, mean age 27 years, range 19–45 years) were classified as non-sensitisers (10 subjects after the first visit and eight subjects after the second visit) were excluded from the subsequent analysis (see online supplementary flowchart 1). This was an expected rate of non-sensitisation based on our previous work.10 This pragmatic design was chosen to improve recruitment and subject retention within the study in order to present data on a larger cohort than have been previously reported using this model.

Study 2: The effect of pharmacological modulation of autonomic tone during deep breathing

A total of 32 healthy subjects (17 men, mean age 28 years, range 21–49) were recruited from the Aalborg study centre (Aalborg Hospital, Aalborg, Denmark). Subjects underwent a screening visit, where PT was determined in the proximal oesophagus before and after acid infusion into the distal oesophagus to define sensitisation status. Thirteen subjects (nine men, mean age 30 years, range 21–47) were non-sensitisers and therefore were excluded from the study. Five subjects (three men, mean age 28 years, range 24–38) were excluded from the study due to the following reasons: two subjects failed to complete both visits, one subject did not tolerate oesophageal intubation, one subject became pregnant and one subject was diagnosed with asthma between the screening visit and randomisation. The 14 healthy subjects who sensitised (eight male, mean age 28 years, range 22–49) were randomised, in a blinded crossover design, to deep breathing with intravenous atropine or placebo (0.9% normal saline) during acid infusion. PT in the proximal oesophagus was then measured at T0, T60, T90 and T120. Subjects were then crossed over and the protocol was repeated at least 2 weeks after the first visit. We chose to use deep breathing across both arms of study 2 to assess any additive, or indeed reductive, effect of atropine on PT but also to assess replication of any findings from protocol 1 in an unrelated independent cohort of subjects.

Data analysis

Results have been presented as mean (with SD), median and range dependent on data type, determined by Shapiro–Wilk testing. In both protocols, changes in PT were analysed using linear mixed effects regression models with maximum restricted likelihood (fixed effects: time, interventions, ie, deep breathing/sham breathing; atropine/placebo; random effect=subject) with T0 thresholds accounted for in the model as zero to yield a regression coefficient for intervention effect (with CI). As data were paired, additional analyses were performed with paired t tests and linear regression as appropriate. Analyses were performed using proprietary software (Stata V.10.0, Stata Corporation, Texas, USA). p Value<0.05 was adopted as the statistical criterion.

Results

Study 1: The effect of deep/sham breathing on autonomic tone and PTs

Deep breathing increases cardiac vagal tone and decreases skin conductance responses

Deep breathing produced an increase in cardiac vagal tone (Δ cardiac vagal tone: mean=2.1±2.3) compared with sham breathing which had no significant effect (Δ cardiac vagal tone: mean=−0.3±2.3), p=0.0006 (paired t test) (see figure 1A). Deep breathing produced a marginal withdrawal of skin conductance response (Δ skin conductance response: mean=−0.6±4.9) compared with sham breathing which had no significant effect (Δ skin conductance response: mean=3±4.8), p=0.03 (paired t-test) (see figure 1B).

Figure 1

The effect of sham/un-paced breathing (shaded black) and deep breathing (unshaded) on cardiac vagal tone (mean±SE) (parasympathetic tone) and skin conductance response (mean±SE) (sympathetic tone). *Statistically significant at p<0.03.

Deep breathing prevents the development of oesophageal hypersensitivity

During acid infusion, pH fell to <2.0 in the distal oesophagus but remained >6.0 in the proximal unexposed oesophagus in all subjects. The most common symptom reported during the acid infusion was nausea. Absolute PT data at T0 and after acid infusion (T60, T90, T120) are shown in table 1. There were no differences in absolute values of PT at T0 in subjects undergoing sham versus deep breathing (mean (SD) 49.3 mA (20.7) vs 46.2 mA (18.3), p=0.54). Deep breathing reduced the development of acid-induced hypersensitivity in the proximal oesophagus (see figure 2). Mixed effects regression showed a coefficient of effect for deep breathing of 9.94 (CI 8.3 to 11.6), p=0.0001. There was no relationship between state or trait anxiety and T0 thresholds or with degree of acid sensitisation at subsequent time points. A significant relationship was found between STAI-T and change in cardiac vagal tone during deep breathing (coefficient 0.45 (CI −1.73 to −0.36), p=0.004) but not sham breathing (coefficient −1.05 (CI −0.54 to +1.44), p=0.36).

Table 1

Absolute values for proximal oesophageal pain thresholds before (T0) and after (T60, T90 and T120) acid infusion with (a) sham and (b) deep breathing

Figure 2

The effect of sham breathing (▪) and deep breathing (●) on the development of central sensitisation, derived from the paired change in pain thresholds (mean±SE of the mean), in the proximal oesophagus at T60, T90 and T120, with mixed effects regression showing a coefficient of effect for deep breathing of 9.94 (CI 8.3 to 11.6), p=0.0001.

Study 2: The effect of pharmacologically influencing autonomic tone during deep breathing

Atropine prevents deep breathing rise in cardiac vagal tone

Deep breathing + placebo produced an increase in cardiac vagal tone (Δ cardiac vagal tone: mean=+2.3±1.8). In contrast, atropine resulted in significant cardiac vagal tone withdrawal (Δ cardiac vagal tone: mean=−3.1±3.3), mean difference=−5.44 (CI −7.7 to −3.1, p=0.0002, paired t test).

Atropine with deep breathing fails to prevent the development of oesophageal hypersensitivity

During acid infusion, pH fell to <2.0 in the distal oesophagus of all subjects but remained >6.0 in the proximal unexposed oesophagus. The most common symptom reported with acid infusion was nausea. Other sensations included a cold sensation in the chest region and heartburn. Absolute threshold data at (T0) and after acid infusion (T60, T90, T120) are shown in table 2. There were no differences in absolute values of PT at T0 in subjects undergoing deep breathing with atropine or placebo (mean (SD) placebo 52.4 mA (15.2) vs atropine 54.4 (15.7), p=0.59). In contrast to protocol 1, there was a slight fall in thresholds with placebo despite deep breathing (−4.2 mA to−4.1 mA across three time points); however, this fall did not meet the criteria for acid-induced sensitisation. This decrease in PT was however more pronounced when deep breathing was combined with atropine (see figure 3). At T60, mean change in PT was −13.1 mA (CI −19.1 to −7.2) with deep breathing/atropine, compared with −4.2 mA (CI −11.3 to +1.6) with deep breathing/placebo. This pattern was repeated at T90 and T120 although the differences converged slightly (T90: mean change in PT was −12.1 mA (CI −17.9 to −6.2) with deep breathing/atropine compared with −4.5 mA (CI −7.7 to −1.2) with deep breathing/placebo; T120: mean change in PT was −9 mA (CI −16.1 to −2.1) with deep breathing/atropine compared with −4.1 mA (CI −10.5 to +2.4) with deep breathing/placebo). Mixed effects regression showed a significant effect for atropine (coefficient −3.5 mA/unit time (CI −6.8 to −0.06), p=0.046). There was no relationship between state or trait anxiety and T0 thresholds (p=0.48) or with degree of acid sensitisation at subsequent time points with atropine or placebo (p=0.41–0.66, linear regression). A small but significant relationship was found between STAI-T and change in cardiac vagal tone during deep breathing (coefficient −0.09 (CI −0.43 to +0.25), p=0.029) but this was abolished with atropine (coefficient 0.45 (CI −1.73 to −0.36), p=0.004).

Table 2

Absolute values for proximal oesophageal pain thresholds before (T0) and after (T60, T90 and T120) acid infusion with deep breathing and (a) placebo or (b) atropine

Figure 3

The effect of atropine/deep breathing (▪) and placebo/deep breathing (●) on the development of central sensitisation, in comparison with the screening visit (♦), derived from the paired change in pain thresholds (mean±SE of the mean), in the proximal oesophagus at T60, T90 and T120. Mixed effects regression showed a significant effect for atropine (coefficient −3.5 mA/unit time (CI −6.8 to −0.06), p=0.046).

Discussion

Our results provide evidence that the PNS is critical in the development of experimentally-induced oesophageal pain hypersensitivity. Physiologically increasing PNS tone with deep breathing prevents the development of oesophageal pain hypersensitivity, yet this effect is largely antagonised by atropine. Hitherto, the role of the PNS in mediating visceral antihyperalgesia has been largely confined to animal studies. We sought to extend these findings to humans through physiological and pharmacological manipulation of PNS in the development of oesophageal hyperalgesia. In this model we used, oesophageal hyperalgesia in the non-acid exposed proximal oesophagus occurs due to central sensitisation at the spinal dorsal horn.25 Its development is influenced by endogenous pain regulatory systems,26 whose dysfunction plays a role in the maintenance of chronic visceral pain, through its functional and dynamic interaction with the ANS.27 For the first time in a model of human oesophageal pain hypersensitivity, we have demonstrated that deep breathing increases PNS tone and ameliorates the development of central sensitisation. Although there is observational evidence for the effectiveness of deep breathing techniques in management of visceral pain, its mechanism of action, to date, has remained unexplored.12 In the context of chronic somatic pain syndromes, deep breathing has been observed to reduce pain and increase daily functioning in patients with fibromyalgia.28 Similarly in acute somatic pain, Friesner et al29 have demonstrated deep breathing-induced analgesia during thoracic drain removal, although this study did not control for distraction. In the context of these data, an interesting finding of our study was that when atropine and deep breathing were coadministered, sensitisation was only partially prevented. This potentially suggests a degree of associated distraction, which arguably accompanies all paced breathing techniques, may have influenced the analgesic effect of deep breathing even though we assiduously controlled for this effect in our study.

In animal models, the effects of vagotomy, with or without supplementary vagal nerve stimulation, in modulating nociception was first reported in excess of 20 years ago.30 Borovikova et al,31 in a series of elegant studies, observed that vagal nerve stimulation suppresses the systemic anti-inflammatory response in a rodent model, leading to the concept of the cholinergic anti-inflammatory pathway. The cholinergic anti-inflammatory pathway has been proposed to be a functional connection between the immune and central nervous systems, providing an interface which can modulate inflammation and pain signalling.32 While our study was not designed to describe the precise neurobiological mechanisms by which increasing PNS tone imparts an analgesic effect, data derived from vagal nerve stimulation studies in animal models have facilitated important insights demonstrating that such mechanisms may manifest from the periphery to central neurotransmission and supraspinal areas.3336 Interestingly, preliminary human data have demonstrated changes in the visceral pain neuromatrix following electrical vagal nerve stimulation using functional neuroimaging techniques.37

There are a number of important therapeutic implications of our findings. An immediately attractive suggestion is whether the direct measurement of cardiac vagal tone during methods that use deep breathing techniques may facilitate the objective interrogation of the success of such measures in changing PNS tone with the a priori aim of inducing analgesia. Equally, it is possible that episodic utilisation of deep breathing techniques during acute inflammatory episode, such as during gastro-oesophageal reflux events, may produce both symptomatic relief and prevention of central sensitisation, hyperalgesia and chronicity of symptoms. Finally, transcutaneous electrical vagal nerve stimulation has been shown to increase somatic PT and reduce pain ratings, in the absence of any demonstrable cardiovascular side effects, and therefore this novel non-invasive technology may offer a treatment option in patients with oesophageal hypersensitivity associated with reflux disease. Such methods may also have utility in patients with IBD and FGID.38

Our data require a degree of circumspection, particularly with respect to our relatively small sample of healthy subjects, although comparable with other studies of this type and fulfilling our sample size calculation (see online supplementary material sample size calculation). It must be appreciated that study 1 was not a traditional crossover design in that sensitisation status was defined during the sham/un-paced visit, rather than at a screening visit as in study 2. Thus, subjects who received the sham/un-paced intervention first, and were found not to sensitise, were excluded and did not attend for the second visit. However, while a limiting factor in the interpretation of the data, this represented a pragmatic approach to enhance recruitment by limiting the study visits in what was an invasive protocol. In study 2, we chose to include a screening visit as we did not wish to expose healthy subjects to potentially an unnecessary drug intervention on safety and ethical grounds as well as allowing us to perform an intention to treat analysis. Nevertheless, despite these two approaches, we have demonstrated a degree of internal validity in that deep breathing reproducibly increased cardiac vagal tone and alleviated the development of central sensitisation in two unrelated cohorts across two study centres despite the aforementioned differences in study design. Additionally, there are inherent limitations to all human pharmacological studies of atropine as its human pharmacodynamics are dose dependent with low (c.2 μg/kg) and high dose (>15 μg/kg) atropine being considered to be partially and completely vagolytic, respectively.39 In our study, largely due to regulatory concerns over cardiovascular safety, we chose a standard dose of 0.5 mg of atropine, which equates to approximately 7 μg/kg. Given that we observed only partial sensitisation in the atropine/deep breathing group, it is possible that the dose we chose was partially, rather than completely, vagolytic. Therefore, it may be possible that by increasing the dose of atropine to vagolytic concentrations, that is, in excess of 15 μg/kg, complete blockade of the PNS effect of deep breathing may occur thereby allowing sensitisation to occur. Atropine is also a competitive antagonist for the muscarinic acetylcholine receptor at the level of smooth muscle, which could therefore influence oesophageal peristalsis and heighten the tone of the lower oesophageal sphincter. However, as we did not use mechanical/balloon distension and quantified the distribution of acid perfusion in the oesophagus, it is unlikely that this exerted any significant degree of confounding on our results. In addition, the pharmacological effect of atropine can occur at a number of levels from the periphery to the spinal cord and in the central nervous system and thus it is therefore a challenge to delineate the exact point of action of deep breathing using this antagonist. Although we have demonstrated an association between an increase in PNS tone and failure of oesophageal sensitisation, this does not establish causality. Similarly, while we attempted to control for any carryover effect, a limitation of many studies employing a crossover design, by adopting a pragmatic interval of at least 2 weeks between visit, it is possible that such an effect may have influenced the data.40 Finally, whether these results are directly applicable to patients with syndromes characterised by chronic oesophageal pain, such as functional chest pain or functional heartburn, is at this stage unclear.

In conclusion, our findings represent the first human studies addressing the pivotal role of the PNS in mediating acid-induced oesophageal pain hypersensitivity. We have shown that the development of acid-induced oesophageal pain hypersensitivity is altered by physiologically and pharmacologically influencing PNS tone. These findings strongly indicate that the PNS plays a central role in the modulation of central sensitisation. Further study is now warranted to investigate the potential of therapeutically manipulating PNS tone in the management of chronic visceral pain syndromes.

Acknowledgments

We would like to thank Dr Richard Hooper, Centre of Health Sciences, Queen Mary University of London, for his advice on the statistical aspects of the study.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • CB and ADF are joint first authors.

    AMD, CHK and QA are joint senior authors.

  • This work was presented in abstract form at the British Society of Gastroenterology Annual Meeting 2012.

  • Contributors CB: acquisition of data; analysis and interpretation of data; critical revision of the manuscript for important intellectual content; statistical analysis. ADF: manuscript preparation; statistical analysis; critical revision of the manuscript for important intellectual content. MN, CB and AG: acquisition of data; critical revision of the manuscript for important intellectual content. CHK: study concept and design; statistical analysis; critical revision of the manuscript for important intellectual content; project supervision; obtained funding. AD: technical or material support; study supervision; critical revision of the manuscript for important intellectual content. QA: pioneered study concept and design, technical or material support; obtained funding as principal applicant; critical revision of the manuscript for important intellectual content; project supervision.

  • Funding CB was funded by a Medical Research Council project grant: G0701706.

  • Ethics approval East London and City Ethics/North Jutland.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.