Exploring relationships for visceral and somatic pain with autonomic control and personality

PAINÒ 144 (2009) 236–244

www.elsevier.com/locate/pain

Research papers

Exploring relationships for visceral and somatic pain with autonomic control and personality Peter Paine a, Jessin Kishor b, Sian F. Worthen b, Lloyd J. Gregory a, Qasim Aziz b,* a b

Department of Gastrointestinal Sciences, Hope Hospital, University of Manchester, UK Barts and the London, Queen Mary School of Medicine and Dentistry, University of London, UK

a r t i c l e

i n f o

Article history: Received 1 July 2008 Received in revised form 24 February 2009 Accepted 24 February 2009

Keywords: Visceral pain Personality Autonomic Vagus Brainstem Co-activation Neuroticism

a b s t r a c t The autonomic nervous system (ANS) integrates afferent and motor activity for homeostatic processes including pain. The aim of the study was to compare hitherto poorly characterised relations between brainstem autonomic control and personality in response to visceral and somatic pain. Eighteen healthy subjects (16 females, mean age 34) had recordings during rest and pain of heart rate (HR), cardiac vagal tone (CVT), cardiac sensitivity to baroreflex (CSB), skin conductance level (SC), cardiac sympathetic index (CSI) and mean blood pressure (MBP). Visceral pain was induced by balloon distension in proximal (PB) and distal (DB) oesophagus and somatic pain by nail-bed pressure (NBP). Eight painful stimuli were delivered at each site and unpleasantness and intensity measured. Personality was profiled with the Big Five inventory. (1) Oesophageal intubation evoked ‘‘fight-flight” responses: HR and sympathetic (CSI, SC, MBP) elevation with parasympathetic (CVT) withdrawal (p < 0.05). (2) Pain at all sites evoked novel parasympathetic/sympathetic co-activation with elevated HR but vasodepression (all p < 0.05). (3) Personality traits correlated with slope of distal oesophageal pain-related CVT changes wherein more neurotic-introvert subjects had greater positive pain-related CVT slope change (neuroticism r 0.8, p < 0.05; extroversion r 0.5, p < 0.05). Pain-evoked heart rate increases were mediated by parasympathetic and sympathetic co-activation – a novel finding in humans but recently described in mammals too. Visceral pain-related parasympathetic change correlated with personality. ANS defence responses are nuanced and may relate to personality type for visceral pain. Clinical relevance of these findings warrants further exploration. Ó 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction Visceral pain is characterised as more diffuse and unpleasant than somatic pain and exhibits viscerosomatic ‘‘referral” [12,45– 47]. Perceptual differences are also found between ‘‘superficial” and ‘‘deep” somatic pain [24]. These psychophysical differences reflect lower Ad and C-fibre densities in viscera and also spinal cord viscerosomatic convergence [11]. Viscera additionally receive privileged extrinsic afferent and efferent autonomic innervation that interface with myenteric plexi [7,27]. Furthermore, brain-imaging studies show ‘‘limbic cortex” activation differences for visceral pain [45,47], albeit less evidently when controlled for unpleasantness [17]. Striated proximal oesophageal muscle has greater extrinsic innervation than ‘‘true viscera” distal smooth muscle [2] and may therefore exhibit intermediate responses to true visceral and somatic pain. Some current pain paradigms (such as pain as

* Corresponding author. Present address: Director Wingate Institute of Neurogastroenterology, 26 Ashfield Street, London E1 2AJ, UK. Tel.: +44 (0) 207 882 2630; fax: +44 (0) 207 375 2103. E-mail address: [email protected] (Q. Aziz).

a ‘‘homeostatic emotion” [15]) have tended however to minimize central neuroanatomical differences for all pain types. The autonomic nervous system is a hierarchically controlled, bidirectional, body–brain interface integrating afferent bodily inputs with central motor outputs for homeostatic-emotional processes [26,37]. Appreciation is growing for hierarchical superiority and for a greater range and flexibility of dis/inhibitory processes involving the pre-frontal cortex and parasympathetic nervous systems (PNS) to facilitate defensive and affiliative behaviours [8,9,14,49]. This is compared to ‘‘fight-flight” sympathetic nervous system (SNS) responses traditionally emphasised in pain research. For example the ‘‘polyvagal perspective” [42] highlights mammalian elaboration of brainstem nucleus ambiguus (NA) cardiac vagal control (CVC) to permit expanded behavioural repertoires [41,42]. Moreover it relates CVC to sophisticated cognitive and emotional processes including personality differences [43]. Temperamental differences in animals affect behavioural and ANS responses to pain and threat but are poorly characterised in humans [13,20,32,53]. The influential Gray-McNaughton personality model places defence response systems centrally in determining behavioural activation and inhibition [21,38]. Personality differences in man are known to affect pain sensitivity and

0304-3959/$36.00 Ó 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2009.02.022

P. Paine et al. / PAINÒ 144 (2009) 236–244

neurophysiological responses to stressors under some genetic influence [43,44]. In particular, extroversion and neuroticism personality traits are implicated as important factors in ANS function and pain response [22,43]. In summary, separate strands of evidence link personality differences with individual differences for pain and in turn link personality with differences in ANS function. Little however has drawn these strands together into a unified picture, nor teased out if their inter-relationships differ for visceral and somatic pain. Our hypothesis was that since the ANS is a core integrative body– brain interface, studying patterned ANS pain response differences could provide a unifying substrate linking personality and pain. Selective ANS components can now be measured non-invasively with improved time resolution. Hitherto it was not possible to non-invasively study time-courses of pain evoked cardiac vagal control (CVC) change in humans. Older techniques analyse CVC respiratory frequency modulation and are statistically invalid over shorter than 1 min time windows [5]. CVC due to baroreceptor modulation however can now be measured beat-to-beat [30,34]. The current study for the first time attempts to explore timecourses of selective autonomic responses to pain evoked from distal oesophagus (true viscera); nail-bed (true ‘‘deep” somatic) and proximal oesophagus (possibly intermediate viscero-somatic) and relate them to personality trait variables. Our further hypotheses were that significant psychophysical and autonomic differences occur for visceral and somatic pain; in particular visceral pain might evoke greater vagal response and neuroticism/introversion might relate more to vagal and visceral pain changes. 2. Methods 2.1. Subjects Healthy volunteers took part in this study. Ethics approval was obtained from Central Manchester Local Research Ethics committee (ref 07/Q1407/3). Fully informed consent was obtained. Subjects with a history of current or chronic pain, gastrointestinal, neurological or psychiatric medical problems or taking medication affecting GI, pain or neuropsychological function were excluded.

2.2. Physiological techniques 2.2.1. Painful oesophageal balloon distension Oesophageal balloons were assembled from 3 cm lengths of silicone tubing (Medasil UK) tied and glued 0.5 cm at each end over perforations 2 cm from the tip of a commercially available nasogastric tube (Pennine Healthcare Ltd, 2.7 mm, 120 cm) creating a 2 cm diameter balloon. The nasogastric tube was swallowed per-orally or trans-nasally according to subject preference and the balloon positioned 34 cm ab orus/36 cm ab nares for distal oesophagus and 22 cm ab orus/24 cm ab nares for proximal oesophagus. No local anaesthetic was used for intubation but passage was eased through the nasopharynx with a water-based lubricant jelly (KY jelly, Johnson & Johnson). The balloon was inflated manually using a syringe at a rate of 2 ml/s until the subject indicated that pain tolerance threshold had been reached. A manual rather than automated inflation technique was used to enable ‘‘real time” titration of balloon volumes to each subject’s pain experience such that pain tolerance was consistently reached each time. This allowed for more precise stimulus-to-stimulus control of volume over short inter-stimulus time periods than a pre-set automated technique which may not account for habituation effects. Subjects were instructed that pain tolerance should be the level beyond first pain sensation at which they could not tolerate any further increase. The balloon was man-

237

ually deflated immediately when pain tolerance threshold was reached as indicated. The subjects were then asked to rate the pain they had just felt on a verbal scale of 1–10 for both intensity and unpleasantness, 10 being most intense/unpleasant pain and 1 being no sensation/not unpleasant. A verbal rather than visual analogue scale was chosen to minimize movement artifacts on autonomic measurement recordings and also because both hands were employed during the study with either nail-bed stimulation or autonomic recordings. 2.2.2. Painful nail-bed stimulation The extension probe of a strain-gauge (Mecmesin, UK) was mounted onto a spring-loaded, manually operated device (Medical Physics Department, Hope Hospital, Salford UK). The extension probe incorporated a blunt-ended thin rod that was applied to the nail bed of the subjects left middle finger using the springloaded device. As per oesophageal balloon distension, the nail-bed pressure was increased until subjects indicated that pain tolerance threshold had been reached. The pressure was then immediately released. The subjects subsequently rated intensity and unpleasantness using the same verbal scale anchors as for oesophageal pain. Once again the manual technique was preferred to the automated in achieving standardization of subjective endpoint (pain tolerance threshold) by flexible stimulus intensity delivery. 2.3. Psychological traits Subjects completed the Big Five Inventory questionnaire [4,28,29], a validated 44 item personality questionnaire. Subscales of neuroticism and extroversion only were used in the subsequent analysis because of a priori data suggesting links between these personality dimensions with autonomic function and visceral pain [22,43] and to minimize the chances of finding spurious associations through multiple comparisons. 2.4. Autonomic measures 2.4.1. Heart rate Skin was firstly prepared by light excoriation to reduce impedance and improve signal (Nuprep, DO Weaver & Co, USA) in areas for standard 3 lead ECG placement (right and left sub-clavicular and cardiac apex). ECG electrodes (Cleartrace, Conmed Corporation, New York) were placed with a small quantity of conducting gel (Spectra 360 electrode Gel, Parker Laboratories USA). ECG was acquired at a rate of >3 kHz using a commercially available biosignals acquisition system (Neuroscope, Medifit, UK). R waves are the first upwards deflection above the electrical baseline on the ECG as part of the QRS complex which represents ventricular depolarisation. The Neuroscope has an in-built R wave detection algorithm, which features accuracy to the nearest millisecond, from which the interbeat (R–R) interval and heart rate are derived. 2.4.2. Parasympathetic efferent (cardiac vagal tone) The Neuroscope has in-built ‘‘voltage controlled oscillators” that detect positive phase shifts in the beat-to-beat R–R interval, a process called ‘‘phase-shift demodulation”. This phase-shift technique is based uniquely among non-invasive measures of cardiac vagal control (CVC) on ‘‘beat-to-beat” baroreceptor reflex physiology [18,27,30,31,34]. Most traditional CVC measures are based on ‘‘breath-to-breath” respiratory modulation, i.e. respiratory sinus arrhythmia (RSA) [1]. The baroreceptor reflex technique therefore has better face validity for smaller/beat-to-beat time windows in the study of CVC than do RSA measures. It also has no built-in assumption of stationary activity (stationarity) compared with that of spectral analysis techniques. The phase-shift technique has been validated for humans using pharmacological blockade [30] and is

238

P. Paine et al. / PAINÒ 144 (2009) 236–244

measured in standardised units on a Linear Vagal Scale (LVS). The Neuroscope technique refers to this measure of CVC as cardiac vagal tone (CVT).

2.4.3. Parasympathetic afferent (Cardiac Baroreflex Sensitivity) The Neuroscope incorporates beat-to-beat R–R and mean blood pressure into an algorithm on a 10-s cycle which calculates cardiac sensitivity to the baroreflex [31]. This is the change in blood pressure for each change in R–R interval (mmHg/ms). The CSB is a putatively indirect measure of parasympathetic afferent sensitivity.

2.4.4. Sympathetic cardiomotor (Cardiac sympathetic index) The edited R–R series was exported into another PC DOS-based program called CMET [1]. This produces a cardiometric called the Toichi Cardiac Sympathetic Index (CSI). This has been validated in humans for 1 min time windows or more as a sympathetic cardiometric in pharmacological blockade studies [51]. It is essentially a ratio of R–R intervals and therefore has no units. It is a relatively novel measure, however other more widely used putative sympathetic cardiometric indices such as the low frequency component of spectral analysis and thereby sympathovagal balance have been largely discredited [19].

2.4.5. Sympathetic sudomotor (skin conductance level) This was obtained using commercially available equipment (Powerlab, AD instruments, UK). Skin on the distal digit finger pulps of right index and ring fingers was lightly wiped with water and then dried. Skin conductance electrodes were then attached to these and the skin-impedance level (uS, micro Siemens) was zeroed for that subject. The mean skin-impedance level was subsequently extracted off-line. Skin conductance is a putative sympathetic ‘‘emotional sudomotor” measure responsive within milliseconds to novel/threatening stimuli [10].

2.4.6. Sympathetic vasomotor (mean blood pressure) Digit arterial blood pressure was obtained using a commercial technique whereby a photoplethysmographic cuff was attached to the distal interphalangeal portion of the right middle finger (Portapress, Finapress Medical Systems, Copenhagen). The analogue waveform generated was exported to the Neuroscope which then directly measured beat-to-beat mean blood pressure. The mean blood pressure under some experimental conditions indexes sympathetic vasomotor tone [48], however blood pressure is broadly under mixed SNS and PNS control as is heart rate.

2.5. Protocol Subjects attended for a single visit lasting 3 h. On arrival they completed consent, and then height, weight and age were obtained. They then completed the Big Five personality inventory. They were reclined comfortably at 45 degrees on a bed in a quiet and temperature controlled laboratory. The ANS recording equipment was attached and a 3 min pre-intubation baseline recording was taken at rest. Subjects were then intubated as described above. ANS recordings were acquired continuously for 6 min post-intubation. In counter-balanced randomised order, subjects then received eight stimulations of either distal balloon or nail bed with an inter-stimulus interval of 3 min (180 s). This was then followed after a brief rest by runs of eight stimuli each from the other two sites in turn. If stimuli or responses were contaminated by coughing, sneezing, swallowing or other artefacts they were rejected and subsequently repeated. Subjects received a verbal warning 30 s before each stimulus delivery to try and standardise any effects of anticipation. (See Fig. 1). 2.6. Analysis For effects of intubation on ANS responses comparisons were made using 90-s data windows. Data from the first 90 s (early phase) and last 90 s (late phase) of the 6 min post-intubation ANS recordings were extracted for comparison with the middle 90 s of the resting pre-intubation baseline. The middle 90 s was used to avoid any potential disturbance present in the first or last resting epochs due to proximity to set-up (i.e. electrode attachment) or continuation of the experiment (i.e. impending intubation). (Fig. 1). Where technically possible, effects of pain on pre- and poststimulus ANS levels were determined over a much smaller time window befitting the more phasic nature of the painful stimulus compared with intubation. CSI measures were technically limited to a minimum of 60-s epochs and our SC measure to 30-s epochs pre- and post-pain stimulus. For the remaining measures (CVT, CSB, MBP and HR) it was possible to compare 10-s peri-stimulus time windows. Maximum change for ANS indices was found to occur in the 10- to 20-s post-stimulus epoch therefore this 10-s epoch was compared with the immediate 10-s pre-stimulus epoch (see Fig. 1). For each subject the mean of each series of 8 pre- and post-pain stimulus levels were first calculated and then group means derived. Group comparisons were made with two-sided paired t-tests using Stats-Direct software. Correlation analysis was performed using Pearson’s correlations in SPSS software. To capture the dynamic information available from stimulusto-stimulus change in pain-related CVT responses, these were transformed into a slope statistical summary measure (Excel,

Fig. 1. Study protocol represented as a graphical timeline with study ‘‘events” highlighted in the top half of the graph and analysis epochs shown in the bottom half.

P. Paine et al. / PAINÒ 144 (2009) 236–244

Microsoft). The slope measured was that of the linear regression line through data points in known_y’s and known_x’s. In this case, x was time (minutes) between stimuli and y was the absolute painrelated change in 10-s peri-stimulus CVT for the eight consecutive stimuli. The change used for the slope function was the maximal change outlined above i.e. the difference between the 10 s immediately pre-stimulus to the second 10-s epoch (10–20 s) post-stimulus. The slope was then the vertical difference divided by the horizontal distance between any two points on the line, which was the rate of change along the regression line. For a priori reasons, namely evidence of clinically relevant relationships between CVT and personality under genetic influences [22,43] and to mini-

239

mize the chances of spurious results due to multiple comparisons, assessment of dynamic change relationships to personality were limited to CVT. All ANS measures were normally distributed and did not require logarithmic transformation. 3. Results 3.1. Subject characteristics Eighteen subjects were studied (two males) with a mean age (±SEM) of 35.4 (±2.7).

Fig. 2. For all panels A–F the white bars represent pre-intervention values and grey bars represent post-intubation values. For the first white bar and two subsequent grey bars in all panels A–F the intervention was oesophageal intubation. The first grey bar in each panel therefore represents the first 90 s post-intubation and the second grey bar represents the last 90 s post-intubation of a 6 min post-intubation recording period. For all panels A–F there then follow three pairs of white and grey bars representing distal oesophageal pain responses then nail-bed pain responses then lastly proximal oesophageal pain responses. For panels A, E and F the white bar in these three pairs represents the 10 s epoch immediately pre-stimulus and the grey bar represents the second 10 s epoch post-stimulus at which maximal change occurred. For panel C 30 s pre- and postpain epochs are shown and for panels B and D 90 s pre- and post-stimulus epochs are shown. (A) Heart rate (HR). (B) Cardiac vagal tone (CVT). (C) Skin conductance. (SC) (D) Cardiac Sympathetic Index (CSI). (E) Mean Blood Pressure (MBP). (F) Cardiac sensitivity to baroreflex (CSB). Error bars are SEM. Base = pre-intubation baseline, Int = postintubation, DB = distal balloon, NB = nail-bed pressure, PB = proximal balloon.

P. Paine et al. / PAINÒ 144 (2009) 236–244

240

In two subjects no distal oesophageal balloon responses were obtained because pain tolerance threshold could not be reached at the limit of balloon distension (40 mls). In one further subject technical delays resulted in insufficient time to obtain proximal balloon responses. These subjects were excluded where necessary from comparisons between sites and in correlations. When comparisons or correlations are based on fewer than the full group of 18 subjects this is made explicit in the results section by giving the number of subjects incorporated into the analysis. Therefore, whereas 18 subjects underwent nail-bed stimulation only 15 subjects underwent stimulation at all 3 sites. Seventeen subjects could be compared for both nail-bed stimulation and proximal balloon distension; 16 subjects for nail-bed stimulation and distal balloon distension; and 15 subjects for proximal and distal balloon distension. 3.2. Personality characteristics The personality characteristics for the group were a mean extroversion score (±SEM) of 3.54 (±0.19) and neuroticism score of 2.73 (±0.22). Neuroticism and extroversion scores were negatively correlated for the group (r = 0.67, two-sided p = 0.0024).

Table 2 Psychophysical characteristics of pain at different sites. Site

Subject number

Intensity

Unpleasantness

Nail bed Proximal balloon Distal balloon

18 17 16

8.06 ± 0.17 8.07 ± 0.14 7.85 ± 0.19

7.77 ± 0.25 8.17 ± 0.16 7.95 ± 0.19

more unpleasant than nail-bed pressure in 17 subjects who underwent both (PB 8.2 ± 0.16 vs. NBP 7.68 ± 0.25, two-tailed paired ttest p = 0.04). 4.1.2. Balloon volumes and nail-bed pressures Distal balloon volumes at pain tolerance threshold were 24.03 ± 1.09 mls. Proximal balloon volumes were 19.77 ± 1.35 mls. In 15 subjects who had both proximal and distal balloon distensions the proximal balloon volumes were significantly lower than the distal (PB 19.1 ± 1.4; DB 23.5 ± 0.99; two-tailed paired ttest p = 0.0002, n = 15). DB and PB volumes were positively correlated suggesting that subjects had similar pain tolerances at both sites relative to other individuals (p = 0.0004, r = 0.79). Nail-bed pressures at pain tolerance threshold were 40.51 ± 2.98 N. Nail-bed pressures were not correlated with balloon volumes.

3.3. ANS responses to intubation Intubation caused a sustained elevation in heart rate (Fig. 2 panel A, left-hand side, three bars). In the early phase (first 90 s) postintubation this was associated with both sympathetic activation (Fig. 2 panels C–E left-hand side, first grey bar) and parasympathetic withdrawal (Fig. 2 panel B, left-hand side, first red bar) and a reduction in parasympathetic afferent sensitivity (Fig. 2 panel F, left-hand side, first grey bar). In the late phase post-intubation (last 90 s of the 6 min postintubation recording period) sympathetic indices (Fig. 2 panels C and D left-hand side, second grey bars), with the exception of MBP (Fig. 2 panel E left-hand side, second grey bar), returned to baseline levels. This pattern was also seen for the vagal afferent sensitivity measure, CSB (Fig. 2 panel F left-hand side, second grey bar). A post-intubation sustained elevation in heart rate was associated with a sustained cardiac parasympathetic withdrawal (Fig. 2 panels A and B left-hand side, second grey bar). The detailed findings and significance levels are shown in Table 1. 4. Pain responses 4.1. Psychophysical response characteristics 4.1.1. Behavioural responses The pain intensity and pain unpleasantness ratings for each of the sites are shown in Table 2. The pain intensity did not differ with respect to the sites but proximal balloon pain was rated as

4.2. ANS pain responses All ANS measures had changes in the post-stimulation period compared to the pre-stimulation period. These changes were maximal in the first minute post-stimulus with recovery occurring in the second minute post-stimulus. In particular maximal change occurred in the 10- to 20-s period post-stimulus (for measures where this time resolution was available). An exception to this was proximal balloon pain where the biggest decrease in MBP was in the 20–30 s post-stimulus epoch. In summary, a pain-related heart rate increase occurred in response to stimulation at all three sites. This was associated with sudomotor and cardiomotor sympathetic increases but vasodepression. Intriguingly the pain-related HR increase was also associated with CVT increases. Reductions occurred (significant only for proximal balloon) in CSB. There were some quantitative differences between the sites but qualitatively the patterned ANS responses were the same at all three sites. The details were as follows (see Table 3 and Fig. 2 right-hand side). 4.2.1. Heart rate There were significant increases at all three sites in the second 10 s epoch post-stimulus compared to the immediate pre-stimulus 10 s epoch. The HR increase was significantly greater for proximal balloon than for nail-bed pressure in the 17 subjects who had both (PB +5.7 bpm ± 1.3; NB +3.03 bpm ± 0.7; p = 0.036) (Fig. 2 panel A).

Table 1 ANS responses to intubation compared with baseline. Statistical tests are two-sided paired t-tests. Numbers are mean and SEM. EPI = early post-intubation; LPI = late postintubation; B = baseline. CVT = cardiac vagal tone; CSB = cardiac sensitivity to baroreflex; CSI = cardiac sympathetic index; MBP = mean blood pressure; SC = skin conductance level. Measure

Baseline

Early Pl

Late Pl

B v EPl

EPl v LPl

B v LPl

Heart rate CVT CSB CSI MBP SCR

70.49 ± 2.19 7.91 ± 0.75 7.16 ± 0.67 1.96 ± 0.14 91.69 ± 2.47 3.03 ± 0.76

77.63 ± 2.75 6.32 ± 0.83 5.4 ± 0.7 2.62 ± 0.20 98.18 ± 2.89 7.83 ± 0.77

73.63 ± 2.12 6.30 ± 0.71 6.31 ± 0.98 2.29 ± 0.13 96.10 ± 2.34 3.56 ± 0.87

p = 0.0003 p = 0.06 p = 0.029 p = 0.0032 p = 0.0001 p < 0.0001

p = 0.0056 p = 0.94 p = 0.16 p = 0.053 p = 0.09 p < 0.0001

p = 0.001 p = 0.025 p = 0.34 p = 0.07 p = 0.002 p = 0.4

P. Paine et al. / PAINÒ 144 (2009) 236–244 Table 3 ANS measures pre- and post-stimulus for nail-bed pressure (NB); proximal balloon distension (PB) and distal balloon (DB). Significance tests were performed using twosided paired student t-tests. ANS measure

Site

Pre-stim

Post-stim

Significance

HR(bpm)

NB PB DB

71.02 ± 2.1 71.04 ± 2.4 72.3 ± 2-5

73.9 ± 2.05 76.7 ± 2.01 76.9 ± 2.01

p = 0.0008 p = 0.0005 p = 0,0097

SCR (uS)

NB PB DB

6.86 ± 0,97 6.59 ± 1.04 6.25 ± 0.96

8.35 ± 1.05 7.65 ± 1.07 7.32 ± 1.08

p < 0.0001 p < 0.0001 p = 0.0002

MBP(mmHg)

NB PB DB

99.1 ± 2.5 104.2 ± 3.5 95.8 ± 3 5

95.5 ± 2.2 100.1 ± 3.4 91.3 ± 4.3

p < 0.0001 p = 0.0036 p = 0,0025

CSI

NB PB DB

2.36 ± 0.15 2.46 ± 0.18 2.38 ± 0.19

2.91 ± 0.26 3.36 ± 0.28 3.33 ± 0.31

p = 0.006 p = 0.0004 p = 0.001

CVT (LVS)

NB PB DB

8.11 ± 1.06 8.12 ± 1.08 8.15 ± 1.4

8.5 ± 1.12 8.73 ± 1.22 8.84 ± 1.32

p = 0.03 p = 0.06 p = 0.015

CSB

NB PB DB

7.2 ± 1.04 7.3 ± 12 7.1 ± 1.4

6.9 ± 0.9 6.1 ± 0.8 6.17 ± 0.7

p = 0.7 p = 0.03 p = 0.2

4.2.2. Sympathetic nervous system 4.2.2.1. Skin conductance levels. There was a significant increase in SC in the 30 s post-stimulus compared with the 30 s pre-stimulus at all three sites (Fig. 2 panel C). The increase in SC was greater for nail-bed pressure than for distal balloon distension in 16 subjects who had both (NB +2.6uS ± 0.4; DB 1.7 ± 0.4 p = 0.0028). SC was also greater for nail bed than for proximal balloon distension in 17 subjects who had both (NB 2.8uS ± 0.4; PB 1.7uS ± 0.3 p = 0.012).

241

Table 4 Correlations of the slope of CVT pain-related changes with personality for the three sites of stimulation. Site

Personality

Subject number

r Value

p Value

Distal balloon

Neuroticism Extroversion

16 16

0.8 0.5

0.0001 0.036

Proximal balloon

Neuroticism Extroversion

17 17

0.45 0.13

0.07 0.6

Nail bed

Neuroticism Extroversion

18 18

0.26 0.26

0.3 0.3

levels except for proximal balloon distension (Fig. 2 panel F). Maximal change occurred in the 10 s period post-stimulus. There were no significant differences between the sites. 4.5. Relationships of personality to ANS responses Relationships between pain-related CVT change and personality were explored for all three sites of stimulation using the slope statistical summary measure for pain-related peri-stimulus CVT change. Neuroticism was positively correlated with the slope of pain-related CVT change for post distal balloon distension. In other words, subjects with a higher neuroticism score tend to have a more positive change in CVT with consecutive stimuli whereas those with lower neuroticism scores tend to have a more negative change in CVT with consecutive stimuli (Table 4, Fig. 3A). Extroversion conversely had a negative correlation with the slope of CVT pain-related change for distal balloon (Table 4, Fig. 3B).

4.3. MBP There was a significant post-stimulus drop in MBP after stimulation at all three sites. This was maximal in the second 10 s epoch post-stimulus for distal balloon and nail bed but in the third 10 s epoch for proximal balloon (Fig. 2 panel E). There were no significant differences between the three sites for the maximal drop in blood pressure. 4.4. CSI There was a significant increase in CSI in the 90 s post-stimulus compared with the 90 s pre-stimulus at all three sites (Fig. 2 panel D) but there were no significant differences between the sites. 4.4.1. Parasympathetic nervous system 4.4.1.1. Cardiac vagal tone. For distal balloon and nail bed there was a significant increase in CVT in the 90 s post-stimulus compared with pre-stimulus CVT (Fig. 2 panel B). There was an increase in CVT for proximal balloon which was just short of significance. The maximal increase in CVT for all three sites was seen during the second 10s epoch post-stimulus although the inter-individual variability of the changes in this time frame was such that these did not reach significance levels at the group level. The change in CVT was significantly greater for distal balloon than for nail bed in the 10 s time frame for 16 subjects who had both (DB mean change +1.37 ± 0.7; NB mean change – 0.06 ± 0.5, p = 0.02). 4.4.1.2. Cardiac sensitivity to baroreflex. There were pain-related decreases in CSB at all three sites but these did not reach significant

Fig. 3. Relationships of distal balloon (DB) distension pain-related changes in CVT to (A) neuroticism and (B) extroversion.

242

P. Paine et al. / PAINÒ 144 (2009) 236–244

There were no significant correlations for CVT and personality at the other two sites of painful stimulation (Table 4).

5. Discussion Our study represents a novel pluralistic comparative exploration in humans, simultaneously for visceral and somatic pain, with psychophysical responses; the time-course of novel selective noninvasive ANS responses and the relationship of these to personality. The most striking finding was a pain-related heart-rate elevation associated with co-activation of sympathetic (SNS) and parasympathetic nervous systems (PNS) but vasodepression. This occurred at all three sites of painful stimulation. This was in marked contrast to the intubation patterned ANS response where the elevated HR was associated with SNS activation, but sustained PNS withdrawal and a vasopressor response. The intubation patterned ANS response was in part what might be expected in a typical ‘‘fight-flight” defence reaction. It is striking to note however that, less typically for classic fight/flight, the PNS withdrawal significantly outlasted the SNS activation. This prolonged PNS withdrawal could represent cardiac ‘‘disinhibition” [49] as the main determinant of the sustained post-intubation HR elevation since SNS activity had returned to near pre-intubation levels in the late post-intubation phase. Roles for neural control of patterned ANS defence responses have variously been assigned to a hierarchical network including pre-frontal and cingulate cortices, amygdala, the septo-hippocampal system, medial hypothalamus and peri-aqueductal grey (PAG). Individual functional differences in this network have been postulated to underlie temperamental differences [38]. The best characterised of these defence/pain systems in mediating patterned ANS responses and facilitating particular behaviours are the medial hypothalamus [35] and PAG [3]. These studies were performed in mammals using top-down lesion/stimulation strategies. The ventrolateral PAG receives visceral sensory input via the nucleus of the solitary tract (NTS) and also receives pain inputs from the spinal dorsal horn. However, only spinal inputs appear to innervate the lateral PAG. Dorsolateral PAG in contrast may be entirely innervated from higher CNS structures. Dorsolateral and lateral PAG stimulation produce classic fight-flight ANS responses, whilst lesions prevent fight-flight behaviour [3]. These fight/flight behaviours are associated with characteristic ANS changes similar to those for intubation, so it is possible that observed intubation patterned ANS responses in our study were mediated via dorsolateral or lateral PAG. The sustained nature of PNS withdrawal may point towards higher, e.g. pre-frontal, influences via the dorsolateral PAG rather than simple reflex spinal effects on lateral PAG. Conversely, the ventrolateral PAG (VLPAG) produces a freeze response characterised by hypotension and bradycardia with marked increased PNS activity and SNS withdrawal. The pain-related ANS responses observed in our study did not fit neatly into these dichotomous fight-flight or freeze ANS response patterns but had some features of both, albeit tending towards the ventrolateral PAG pattern (e.g. vasodepression). PNS and SNS co-activation may be characteristic of a different type of freeze behaviour sometimes called ‘‘effroi”/‘‘fright”, tonic-immobility or hypervigilance [8,9]. In this behaviour SNS activation is initially constrained by PNS activation, but prepared if subsequently necessary for disinhibition of SNS activation from PNS activity for fight/flight. Recent work by Carrive and Walker has suggested that VLPAG outputs can be ‘‘blended” with defence arousal system outputs to produce just this kind of pattern i.e. neither fight-flight nor ‘‘flaccid-freeze” but a more ‘‘tonic-freeze” pattern [54]. Paton et al. used a predominantly bottom-up investigative approach in a working rat heart-brainstem preparation to study

ANS responses to pain and their brainstem origins. They were intrigued to find a pain-related heart rate elevation associated with cardiac vagal and SNS co-activation. They have characterised this further using pharmacological blockade studies and found on occasion that the PNS activity served to constrain the tachycardia but on others paradoxically facilitated it [39,40]. Clearly, the ways in which PNS and SNS interact as part of flexible adaptive response is more varied than simple reciprocal activity [6]. Our study may be the first to report patterned PNS/SNS co-activation plus tachycardia and vasodepression for pain in humans. We found little evidence for considerable distinction between visceral and somatic pain in psychophysical or ANS responses at the group level. It is possible that lack of major differences in qualitative ANS response patterns, as might be expected from lateral/ ventrolateral PAG studies, may be because nail-bed pressure represents ‘‘deep” somatic as opposed to superficial somatic pain (such as heat). Recent evidence suggests that such distinctions are less tenable in an integrated ‘‘homeostatic afferent processing network” [15]. We did not confirm findings from previous psychophysical analysis of unpleasantness rating differences between distal oesophageal and somatic pains [46]. We speculate that this could be accounted for if the deeper somatic pain of nail-bed compression, which we employed, is rated more unpleasant than skin heat pain used previously, though this would require formal comparison. We did however find a significant unpleasantness difference between proximal oesophagus and somatic pains despite similar intensity levels. Some subjects reported associated ‘‘choking” sensations with proximal oesophageal pain which may account for this. Studies by Strigo et al. compared brain activation differences between oesophageal and skin heat pain [45,47] and there have been numerous other brain-imaging studies of distension-evoked visceral pain in healthy and clinical populations [25]. For example Kwan et al. found different urge and pain-related fore-brain activations to rectal balloon distension in patients with irritable bowel syndrome compared with controls [33]. These included insula and anterior cingulate cortices which are both implicated in central control of the ANS [27]. There are however inconsistencies as to cortical structures activated between imaging studies and uncertainties concerning their functional significance [25]. Whilst previous studies have separately studied relationships between psychophysics [46], somatic pain [16], personality [43] and visceral pain [22,23,50,52] to autonomic function in both healthy and clinical populations, there are none we know of that have studied these concurrently in adults using the novel ANS measures we have employed. The general pattern detected in previous studies has been of vagal withdrawal and sympathetic activation during acute/experimental pain whereas in chronic/ clinical pain, hyperalgesia has been associated with low ‘‘vagal tone” and/or sympathetic hyperactivity. Another novel feature of our study was dissection of the time course of the selective ANS responding to pain using novel SNS and PNS measures. Maximal change generally occurred in the second 10 s epoch after commencement of the pain stimulus. The stimuli themselves however took several seconds each to perform; they may not therefore be directly comparable to punctate millisecond stimuli on the one hand or more tonic stimuli on the other. The time course of pain responses confirms that it is physiologically more meaningful to study ANS change on a beat-to-beat time scale rather than 1 min blocks as has been the case for non-invasive human PNS studies using traditional RSA-based measures. The other main finding of our study was a relationship between personality and pain-related CVT changes for distal oesophagus. In particular, more neurotic individuals exhibited increases in CVT change with subsequent painful stimuli whereas more emo-

P. Paine et al. / PAINÒ 144 (2009) 236–244

tionally stable individuals had reduced or more negative CVT changes with subsequent stimuli. Relationships between autonomic and personality variables in the literature are inconsistent. This probably in part reflects multiple differences in methodologies of ANS and psychological characterisation as well as experimental paradigm and subject characteristics. Moreover, ANS measures and personality variables are multiply determined and therefore a simple one-to-one correspondence between physiology and psychology is unlikely [36]. Nonetheless these findings confirm a broad consensus linking parasympathetic HRV measures, neuroticism and emotionality under apparent pleiotropic genetic influences [43]. The ‘‘special” relationship of true visceral pain to neuroticism here is intriguing in light of an overrepresentation of high-scoring neurotics in functional visceral pain syndromes in whom links between neuroticism and vagal abnormalities have also been found [22]. We speculate that this may reflect the putatively ‘‘privileged” relations of the viscera to the autonomic nervous system [7,27]; differences in visceral brainstem processing [3,35] and subsequent elaboration within the ‘‘emotional brain” [45,47]. Our hypothesis that the ANS provides a substrate for linkage between personality and pain responses, being an integrative brain– body interface, has been supported. Although overall patterned ANS responses for visceral and somatic pain were qualitatively similar (but notably different from those to oesophageal intubation), parasympathetic responses were influenced by personality for visceral pain. These novel findings suggest that the ANS provides capacity for reflexive bodily responses to noxious stimuli which at the same time are nuanced according to the nature of a threat (tonic freeze vs fight-flight) and also according to pre-set individual differences in response tendency (personality). Further dissection of these findings with similar methodology holds promise for an improved mechanistic understanding of how the ‘‘rubber hits the road” in bodily translation of psychological state and trait processes such as personality and pain. In conclusion, we provide evidence of novel pain-related sympathetic and parasympathetic co-activation with vasodepression which was similar for visceral and somatic pain. Furthermore we found a relationship between personality and visceral pain-related changes in cardiac vagal tone. These findings may have relevance for the clinical overlap in somatic and visceral pain conditions and also for understanding individual differences in pain response. Acknowledgments This research was funded by Cancer Research UK. There are no conflicts of interest. Professor Aziz is funded by an MRC career establishment grant. References [1] Allen JJ, Chambers AS, Towers DN. The many metrics of cardiac chronotropy: a pragmatic primer and a brief comparison of metrics. Biol Psychol 2007;74:243–62. [2] Aziz Q, Thompson DG. Brain-gut axis in health and disease. Gastroenterology 1998;114:559–78. [3] Bandler R, Keay KA, Floyd N, Price J. Central circuits mediating patterned autonomic activity during active vs. passive emotional coping. Brain Res Bull 2000;53:95–104. [4] Benet-Martinez V, John OP. Los Cinco Grandes across cultures and ethnic groups: multitrait multimethod analyses of the Big Five in Spanish and English. J Pers Soc Psychol 1998;75:729–50. [5] Berntson GG, Bigger J, Eckberg DL, Grossman P, Kaufmann PG, Malik M, Nagaraja HN, Porges SW, Saul J, Stone PH, van der Molen MW. Heart rate variability: origins, methods, and interpretive caveats. Psychophysiology 1997;34:623–48. [6] Berntson GG, Cacioppo JT, Quigley KS. Autonomic determinism: the modes of autonomic control, the doctrine of autonomic space, and the laws of autonomic constraint. Psychol Rev 1991;98:459–87. [7] Berthoud HR, Blackshaw LA, Brookes SJ, Grundy D. Neuroanatomy of extrinsic afferents supplying the gastrointestinal tract. Neurogastroenterol Motil 2004;16:28–33.

243

[8] Bracha HS. Freeze, flight, fight, fright, faint: adaptationist perspectives on the acute stress response spectrum. CNS Spectr 2004;9:679–85. [9] Bracha HS, Ralston TC, Matsukawa JM, Williams AE, Bracha AS. Does ‘‘fight or flight” need updating? Psychosomatics 2004;45:448–9. [10] Cacioppo JT, Tassinary LG, Berntson GG, editors. Handbook of psychophysiology. New York, NY: Cambridge University Press; 2000. 1039, Xiii. [11] Cervero F, Laird JM. From acute to chronic pain: mechanisms and hypotheses. Prog Brain Res 1996;110:3–15. [12] Cervero F, Laird JM. Visceral pain. Lancet 1999;353:2145–8. [13] Cohen H, Zohar J, Matar M. The relevance of differential response to trauma in an animal model of posttraumatic stress disorder. Biol Psychiatry 2003;53:463–73. [14] Corr PJ. Reinforcement sensitivity theory and personality. Neurosci Biobehav Rev 2004;28:317–32. [15] Craig AD. A new view of pain as a homeostatic emotion. Trends Neurosci 2003;26:303–7. [16] Diatchenko L, Anderson AD, Slade GD, Fillingim RB, Shabalina SA, Higgins TJ, Sama S, Belfer I, Goldman D, Max MB, Weir BS, Maixner W. Three major haplotypes of the beta2 adrenergic receptor define psychological profile, blood pressure, and the risk for development of a common musculoskeletal pain disorder. Am J Med Genet B Neuropsychiatr Genet 2006;141:449–62. [17] Dunckley P, Wise RG, Aziz Q, Painter D, Brooks J, Tracey I, Chang L. Cortical processing of visceral and somatic stimulation: differentiating pain intensity from unpleasantness. Neuroscience 2005;133:533–42. [18] Eckberg DL. Temporal response patterns of the human sinus node to brief carotid baroreceptor stimuli. J Physiol 1976;258:769–82. [19] Eckberg DL. Sympathovagal balance. A critical appraisal. Circulation 1997;96:3224–32. [20] Frank E, Salchner P, Aldag J, Salome N, Singewald N, Landgraf R, Wigger A. Genetic predisposition to anxiety-related behaviour determines coping style, neuroendocrine responses, and neuronal activation during social defeat. Behav Neurosci 2006;120:60–71. [21] Gray JA, McNaughton N. The Neuropsychology of anxiety: an enquiry into the functions of the septo-hippocampal system. 2nd ed. Oxford: Oxford University press; 2000. [22] Haug TT, Svebak S, Hausken T, Wilhelmsen I, Berstad A, Ursin H. Low vagal activity as mediating mechanism for the relationship between personality factors and gastric symptoms in functional dyspepsia. Psychosom Med 1994;56:181–6. [23] Heitkemper M, Jarrett M, Cain KC, Burr R, Levy RL, Feld A, Hertig V. Autonomic nervous system function in women with irritable bowel syndrome. Dig Dis Sci 2001;46:1276–84. [24] Henderson LA, Bandler R, Gandevia SC, Macefield VG. Distinct forebrain activity patterns during deep versus superficial pain. Pain 2006;120:286–96. [25] Hobson AR, Aziz Q. Brain imaging and functional gastrointestinal disorders: has it helped our understanding? Gut 2004;53:1198–206. [26] Holstege G, Bandler R, Saper CB. The emotional motor system. Prog Brain Res 1996;107:3–6. [27] Janig W. The integrative action of the autonomic nervous system: neurobiology of homeostasis. New York: Cambridge University Press; 2006. [28] John, OP, Donahue, EM, Kentle RL. The ‘‘Big Five” inventory – versions 4a and 54. Berkeley: University of California, Berkeley, Institute of Personality and Social Research; 1991. [29] John OP, Srivastava S. The Big-Five trait taxonomy: history, measurement and theoretical perspectives. In: Pervin L, John OP, editors. Handbook of personality: theory and research. New York: Guilford; 1999. [30] Julu PO. A linear scale for measuring vagal tone in man. J Auton Pharmacol 1992;12:109–15. [31] Julu PO, Cooper VL, Hansen S, Hainsworth R. Cardiovascular regulation in the period preceding vasovagal syncope in conscious humans. J Physiol 2003;549:299–311. [32] Keller N, Diedrich A, Appalsamy M, Miller L, Caron M, McDonald M, Shelton R, Blakely R, Robertson D. Norepinephrine transporter-deficient mice respond to anxiety producing and fearful environments with bradycardia and hypotension. Neuroscience 2006;139:931–46. [33] Kwan CL, Diamant NE, Pope G, Mikula K, Mikulis DJ, Davis KD. Abnormal forebrain activity in functional bowel disorder patients with chronic pain. Neurology 2005;65:1268–77. [34] Little CJ, Julu PO, Hansen S, Reid SW. Real-time measurement of cardiac vagal tone in conscious dogs. Am J Physiol 1999;276:H758–65. [35] Lumb BM. Hypothalamic and midbrain circuitry that distinguishes between escapable and inescapable pain. News Physiol Sci 2004;19:22–6. [36] Matthews G. Neuroticism from the top down: psychophysiology and negative emotionality. In: Stelmack R, editor. On the psychobiology of personality. Elsevier; 2004. [37] Mayer EA, Naliboff BD, Chang L, Coutinho SVV. Stress and irritable bowel syndrome. Am J Physiol Gastrointest Liver Physiol 2001;280:G519–24. [38] McNaughton N, Corr PJ. A two-dimensional neuropsychology of defense: fear/ anxiety and defensive distance. Neurosci Biobehav Rev 2004;28:285–305. [39] Paton JF, Boscan P, Pickering AE, Nalivaiko E. The yin and yang of cardiac autonomic control: vago-sympathetic interactions revisited. Brain Res Brain Res Rev 2005;49:555–65. [40] Paton JF, Nalivaiko E, Boscan P, Pickering AE. Reflexly evoked coactivation of cardiac vagal and sympathetic motor outflows: observations and functional implications. Clin Exp Pharmacol Physiol 2006;33:1245–50.

244

P. Paine et al. / PAINÒ 144 (2009) 236–244

[41] Porges SW. A phylogenetic journey through the vague and ambiguous Xth cranial nerve: a commentary on contemporary heart rate variability research. Biol Psychol 2007;74:301–7. [42] Porges SW. The polyvagal perspective. Biol Psychol 2007;74:116–43. [43] Riese H, Rosmalen JG, Ormel J, Van Roon AM, Oldehinkel AJ, Rijsdijk FV. The genetic relationship between neuroticism and autonomic function in female twins. Psychol Med 2007;37:257–67. [44] Stelmack R. On personality and arousal: a historical perspective on Eysenck and Zuckermann. In: Stelmack R, editor. On the psychobiology of personality. Elsevier; 2004. [45] Strigo IA, Albanese MC, Bushnell MC, Duncan GH. Visceral and cutaneous pain representation in parasylvian cortex. Neurosci Lett 2005;384:54–9. [46] Strigo IA, Bushnell MC, Boivin M, Duncan GH. Psychophysical analysis of visceral and cutaneous pain in human subjects. Pain 2002;97:235–46. [47] Strigo IA, Duncan GH, Boivin M, Bushnell MC. Differentiation of visceral and cutaneous pain in the human brain. J Neurophysiol 2003;89:3294–303. [48] Sun M, Guyenet P. Medulospinal sympathoexcitatory neurones in normotensive and spontaneously hypertensive rats. Am J Physiol 1986;250:R910–7.

[49] Thayer JF, Friedman BH. Stop that! Inhibition, sensitization, and their neurovisceral concomitants. Scand J Psychol 2002;43:123–30. [50] Tillisch K, Mayer EA, Labus JS, Stains J, Chang L, Naliboff BD. Sex specific alterations in autonomic function among patients with irritable bowel syndrome. Gut 2005;54:1396–401. [51] Toichi M, Sugiura T, Murai T, Sengoku A. A new method of assessing cardiac autonomic function and its comparison with spectral analysis and coefficient of variation of R–R interval. J Auton Nerv Syst 1997;62: 79–84. [52] Tougas G, Spaziani R, Hollerbach S, Djuric V, Pang C, Upton AR, Fallen EL, Kamath MV. Cardiac autonomic function and oesophageal acid sensitivity in patients with non-cardiac chest pain. Gut 2001;49:706–12. [53] Valance D, Boissy A, Despres G, Constantin P, Leterrier C. Emotional reactivity modulates autonomic responses to an acoustic challenge in quail. Physiol Behav 2007;90:165–71. [54] Walker P, Carrive P. Role of ventrolateral periaqueductal gray neurons in the behavioral and cardiovascular responses to contextual conditioned fear and poststress recovery. Neuroscience 2003;116:897–912.