Pain Medicine 2012; 13: 215–228

Wiley Periodicals, Inc.

PSYCHOLOGY, PSYCHIATRY & BRAIN

NEUROSCIENCE SECTION

Original Research Articles

The Effect of Deep and Slow Breathing on Pain
Perception, Autonomic Activity, and Mood

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

Processing—An Experimental Study pme_1243 215..228

Volker Busch, MD,* Walter Magerl, MD,† Uwe Kern, DSB intervention, subjects were asked to breathe
MD,‡ Joachim Haas, MD,* Göran Hajak, MD,* and guided by a respiratory feedback task requiring a
Peter Eichhammer, MD* high degree of concentration and constant atten-
tion. In the relaxing DSB intervention, the subjects
*Department of Psychiatry and Psychosomatic relaxed during the breathing training. The skin con-
Medicine, University of Regensburg, Regensburg; ductance levels, indicating sympathetic tone, were
measured during the breathing maneuvers. Thermal
†
Department of Neurophysiology, Centre of detection and pain thresholds for cold and hot
stimuli and profile of mood states were examined
Biomedicine and Medical Technology Mannheim
before and after the breathing sessions.
(CBTM), University of Heidelberg, Mannheim;
Results. The mean detection and pain thresholds
‡
Centre for Pain Management & Palliative Care, showed a significant increase resulting from the
Wiesbaden, Germany relaxing DSB, whereas no significant changes of
these thresholds were found associated with the
Reprint requests to: Volker Busch, MD, Center for Pain attentive DSB. The mean skin conductance levels
and Affective Disorders, Department of Psychiatry, indicating sympathetic activity decreased signifi-
Psychotherapy and Psychosomatic Medicine, cantly during the relaxing DSB intervention but not
University of Regensburg, Universitätsstraße 84, during the attentive DSB. Both breathing interven-
93059 Regensburg, Germany. Tel: 49-941-941-0; Fax: tions showed similar reductions in negative feelings
49-941-941-2925; E-mail: [email protected]. (tension, anger, and depression).
Authors Approval/Disclosure: All authors have Conclusion. Our results suggest that the way of
personally reviewed and given final approval of the breathing decisively influences autonomic and pain
version submitted, and neither the manuscript nor its processing, thereby identifying DSB in concert with
data have been previously published or are currently relaxation as the essential feature in the modulation
under consideration of publication. of sympathetic arousal and pain perception.

Key Words. Breathing; Mood; Pain; Relaxation;

Abstract Respiration; Skin Conductance Level

Objective. Deep and slow breathing (DSB) tech- Introduction

niques, as a component of various relaxation tech-
niques, have been reported as complementary Deep and slow breathing (DSB) techniques are widely
approaches in the treatment of chronic pain syn- used in a variety of diseases encompassing somatic dis-
dromes, but the relevance of relaxation for alleviat- orders such as hypertension and pulmonary diseases [1]
ing pain during a breathing intervention was not as well as psychiatric disorders including anxious and
evaluated so far. depressive syndromes [2,3] or stress-related disorders
[4–10]. With regard to chronic pain syndromes, DSB tech-
Methods. In order to disentangle the effects of niques being part of many physical, mental, and spiritual
relaxation and respiration, we investigated two dif- disciplines such as yoga [11], Qi-Gong [12], or Tai Chi [13]
ferent DSB techniques at the same respiration rates are integrated into multimodal treatment approaches.
and depths on pain perception, autonomic activity, Chalaye and colleagues [14] found higher pain thresholds
and mood in 16 healthy subjects. In the attentive and tolerances in healthy adults after a DSB training.

215
Busch et al.

However, despite these findings, inconsistent results SCID-ICV

about the therapeutical efficacy point to the fact that
breathing interventions may be based on a complex inter- The SCID-ICV is a diagnostic exam used to determine
play of distinct factors not entirely identified until now [15]. DSM-IV Axis I disorders including mood disorders, anxiety
disorder, psychotic disorders, and substance-use disor-
In this context, relaxation may play a pivotal role in ders (original publication [21]; German version [22]; psy-
transforming breathing techniques into an effective chometric properties [23,24]).
method in the therapy of pain and stress-related disor-
ders. Furthermore, DSB in concert with relaxation has BDI
proven to efficiently reduce stress-related biological activ-
ity in healthy volunteers as mirrored by a reduction in the The level of depression was assessed using the “Beck
sympathetic tone [7]. In contrast, a DSB training associ-

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

Depression Inventory—BDI,” covering emotional, behav-
ated with a concentration challenge and guided by respi- ioral, and somatic symptoms (original publication [25];
ratory feedback failed to reduce sympathetic arousal [16]. German version [26]; psychometric properties [27]). The
Moreover, sustained concentration on inhaling and exhal- BDI is a 21-question inventory, which is worldwide among
ing during attentive breathing interventions has demon- the most used self-rating scales to assess the intensity of
strated an increase in sympathetic arousal [17]. depression on the basis of the main symptoms discrimi-
nating between depressives and nondepressives [28].
For this reason, relaxation may constitute a biologically Higher total BDI scores indicate more severe depressive
and clinically effective component of breathing tech- symptoms, BDI scores ⱖ 18 indicate a depressive
niques, additionally influencing mood processing [18]. This disorder.
aspect seems of utmost importance, considering that
breathing management is able to modulate emotional pro-
STAI-X2
cessing in the presence of pain, thereby pointing to the
mutual relationship between pain and mood processing
We further assessed anxiety with the “State and Trait
[19]. However, in analogy to the pain- and stress-related
Anxiety Inventory—STAI-,” a psychological 20-question
effects of DSB, recent studies do not provide any insight
inventory based on a 4-point Likert scale (original publi-
that would clarify the distinct impact of respiration and
cation [29]; German version [30]; psychometric properties
relaxation on emotional processing [20].
[31]). Scores range from 20 to 80 with higher scores
correlating with greater trait or state anxiety. There are no
For this reason, our study aims at elucidating the rel-
cutoffs, as normative valued for trait or state anxiety
evance of relaxation as an independent factor which may
depends on genus and age [29]. We only used the “Trait
mediate the effect of DSB on pain perception, sympa-
anxiety” part (STAI-X2) for our study.
thetic activity, and mood. In detail, we used two different
DSB techniques characterized by identical respiration
rates and depths and distinct from the presence of relax- Study Design and General Information
ation. To the best of our knowledge, this is the first human
study estimating the impact of relaxation during a DSB The study (macrocycle) consisted of two succeeding
technique on pain perception. breathing interventions, each lasting 6 weeks (meso-
cycles). Each of the two mesocycles consisted of 6 weeks
(microcycles) (Figure 1). All measurements were taken
Methods from the same experimenter, who supervised the breath-
ing trainings. A second experimenter, who was blinded for
Subjects the type of intervention as well as for the microcycle,
assessed all further analyses.
Sixteen young and healthy undergraduate students (13
female, 3 male) at the local university (Regensburg, Breathing Interventions
Germany) participated in the study. Exclusion criteria were
the following: any psychiatric disorders or neurological Both interventions were separated by a “wash-out” period
syndromes, any cardiac or respiratory diseases, a history of 6 months to avoid any carryover effects. All subjects
of migraines or other (chronic) pain syndromes, or the use were instructed not to practice at home and were further-
of pain medication or psychotropic drugs. All participants more not allowed to participate in any breathing trainings
underwent a neurological examination and were inter- or meditation programs during the washout period. After
viewed by a psychiatrist (first author), who additionally inclusion, all subjects received written handouts explaining
administered the SCID-1 screening instrument (Structured the course of the study and an instruction manual for DSB,
Clinical Interview for DSM-IV Axis I Disorders, Clinician according to the recommendations from breathing litera-
Version [SCID-ICV]). Moreover, the participants completed ture [32]. Furthermore, they received two experimental
the Beck Depression Inventory (BDI) and the “Trait anxiety” sessions of DSB guided by a biofeedback and breathing
part of the State and Trait Anxiety Inventory (STAI-X2). expert scheduled several days apart from the beginning of
Informed consent was obtained from all volunteers and each of both interventions. The two different breathing
the study was approved by the local ethics committee. interventions were:

216
 Effect of Breathing on Pain Perception

2 subsequent Mesocycles (attentive DSB and relaxing DSB; 6 weeks each)

Microcycle1 Microcycle2 Microcycle3 Microcycle4 Microcycle5 Microcycle6

1 week 1 week 1 week 1 week 1 week 1 week

Sup. Training Sup. Training Sup. Training

Sup. Training Sup. Training Sup. Training
Measurement Measurement Measurement

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

Baseline Block1 Break1 Block2 Break2 Block3
2-3min 5min 1min 5min 1min 5min

P P
O Detection and NB Breathing NB Breathing NB Breathing Detection and O
M Pain Thresholds Intervention Intervention Intervention Pain Thresholds M
S (attentive or (attentive or (attentive or S
relaxing DSB) relaxing DSB) relaxing DSB)

Figure 1 Course of the study. Sup. Training = training under supervision; min = minute; POMS = profile of
mood states; DSB = deep and slow breathing; NB = natural breathing.

1. Attentive DSB (aDSB): The subjects were asked to were supervised and guided by a trained biofeedback
breathe according to a respiratory feedback task [33], expert. The subjects were instructed not to hold their
assuming an “externally paced” respiration with the breath, but to breathe slowly and deeply throughout the
help of a given ideal breathing curve representing res- breathing cycle. In both interventions, expiratory/
piration frequency and depth. The ideal breathing curve inspiratory time ratio of each breathing cycle was 60/30%,
and the individual breathing curve were presented followed by a brief pause (10%), similar to the recommen-
together on a monitor. The subjects had to try to fit their dation for DSB in literature [16,38]. To control for interac-
own respiration curve to the ideal curve, which required tion effects between the experimenter and the participant
constant attention and concentration on the breathing and in order to get similar breathing depths and rates, the
task. experimenter provided similar breathing instructions and
2. Relaxing DSB (rDSB): The subjects were told to direct supported the participants by giving the same number of
their awareness on the experience of breathing. They directly addressed verbal suggestions in both groups.
had to look on a spot on a wall with their eyes kept Structured sentences were based upon voice dialogue
open. They were “internally paced” by verbal instruc- recommendations used in deep relaxing breathing therapy
tions of the experimenter in order to provide similar regimens [39]. Total duration of the breathing training in
respiration rates and depths compared with the aDSB each microcycle was 20 minutes, consisting of one base-
intervention. The subjects did not get a visual control of line period of 2–3 minutes, followed by three breathing
their performance, as their breathing activities were not blocks of 5 minutes (interrupted each by short breaks of 1
fed back on a monitor. These aspects provided a type minute). The baseline period served as an adaptation
of breathing that required very little cognitive process- phase to get accustomed to the laboratory situation and
ing and which has been reported to induce a more to allow the experimenter to calibrate the equipment.
meditative state [7,34–36]. Values from baseline and breathing breaks were not used
for further analyses.
In both interventions, the subjects were instructed to keep
a constant, slow, and deep diaphragmal breathing rhythm Measurement of Breathing Parameters
with a respiration rate of 7 cpm (cycles per minute), which
is approximately half of the normal rate, that had been Breathing techniques used in biofeedback and voluntary
recommended in previous studies investigating the effects breathing training paradigms are often designed to control
of DSB on arousal [37]. Furthermore, a respiration depth for thoracic or abdominal respiration movements [40]. In
of 2 cm amplitude/cycle was required. Both interventions studies of voluntarily controlling of breathing activity, the

217
Busch et al.

abdominal control of respiration—with and without visual the breathing training was performed six times in
control—accurately reproduced specified breathing fre- each mesocycle under supervision of the trainer (once
quencies and depths [41]. Therefore, we used an abdomi- per week), the thermal detection and pain thre-
nal respiratory module fixed around the subject’s upper sholds only from the first, fourth, and sixth microcycles
abdomen using a strain gauge belt. To ensure the same were measured, each before and after the breathing
and correct position, the device was mounted 5 cm above intervention.
the umbilicus, directly on the skin. The module was a
two-channel device (resolution 0.2 mm, measurement Measurement of Sympathetic Activity
range 20 cm) producing a digitalized signal of the ana-
logue respiration movements (amplitude in mm/cycle, fre- The skin conductance level (SCL) represents the electrical
quency in cpm). The respiration rates and depths from all conductance of the skin, which varies with its moisture
three blocks of one microcycle and then from the first, level. Any changes in eccrine sweating are related to

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

fourth, and sixth microcycles were averaged. changes of the SCLs [47]. Human sweat glands are con-
trolled by the sympathetic nervous system [48], so skin
Measurement of Detection and Pain Thresholds conductance is used as an indication of psychological or
physiological arousal [49] and has recently been shown to
Thermal detection thresholds and pain thresholds for cold covary with the perception of pain stimuli [50,51]. The SCL
and hot stimuli [42] were measured using a TSA 2001-II was recorded using a constant-voltage device (Biofeed-
thermal sensory testing device (Medoc, Ramat Yishai, back Expert 2000, Schuhfried, Mödling, Austria; distrib-
Israel) according to the quantitative sensory testing (QST) uted by Schwa-Medico, Ehringshausen, Germany),
protocol developed by the German Research Network on according to Venables and colleagues [47]. SCL
Neuropathic Pain [43]. All QST measurements were responses were recorded continuously throughout the
obtained from the right-hand dorsum with a contact area microcycle (range: 0–50 mS; digital resolution: 0.024 mS).
of the thermode of 3 ¥ 3 cm (9 cm2). A strap was affixed Before attaching the Ag/AgCl electrode to the nondomi-
to maintain constant pressure between the hand and the nant hand (on the palmar surface of the middle phalanx of
thermode. Cold detection threshold (CDT) and warm the ring finger), the skin was cleaned with a small dispos-
detection threshold (WDT) were measured first, followed able alcohol pad [52]. Time markers, separating breathing
by cold pain threshold (CPT) and heat pain threshold blocks and breaks, were synchronized with the recording
(HPT). The baseline temperature was 32°C and all thermal of the physiological data on a common time line. The mean
stimuli were applied as ramps with a change of 1°C/ changes of the SCLs during the breathing maneuvers were
second from baseline. Cutoff temperatures were 0 and calculated as the proportion of the SCL values of the first
50°C. The respective thresholds were recorded and minute and the last minute of one breathing block. Mean
stimuli were terminated when the subject pressed a changes of SCL were then averaged from all three blocks
button to signal the detection of the respective sensation. of one microcycle. Raw data (in mS) were used to study the
The mean threshold temperature of three consecutive change of SCL in terms of percentage. The respiration
measurements before and after the first, fourth, and sixth depths and rates and the SCLs during the first, fourth, and
microcycles was calculated, respectively. During the sixth microcycles were recorded.
experiment, the subjects were not able to watch
the computer screen. WDTs and CDTs along with Measurement of Mood States
CPTs and HPTs were z-transformed on the basis of the
following calculation, ensuring that all parameters were The subjects were asked to complete the “Profile of Mood
scaled in units of the standard normal distribu- States” (POMS) before and after the first, fourth, and
tions (0 = mean, 1 = standard deviation [SD]): z-score = sixth microcycles. The POMS is a well-established, factor-
(Xsingle value - Meangroup)/SDgroup (group = 15; across all data analytically derived self-report measure of psychological
points). We adjusted the algebraic sign of z-score values distress characterized by high levels of reliability and validity
for WDT and HPT (*-1) and subsequently pooled CDT (original publication [53]; German version [54]; psychomet-
together with WDT (= thermal detection), as well as CPT ric properties [55,56]), which was used in several investi-
together with HPT (= thermal pain). The z-transformation gations studying anxiety, depression, or pain [57–60]. The
had been carried out in order to be able to compare mean POMS consists of 65 adjectives rated on a 0–4 scale,
change of thresholds and to generate one value for detec- providing six different mood states and one total mood
tion and one value for pain perception, respectively. This score: tension–anxiety (TA), depression–dejection (DD),
transformation was done according to the recommenda- anger–hostility (AH), fatigue (F), vigor (V), confusion–
tions of a recent QST reference article [44]. As a conse- bewilderment (CB), and total mood disturbance (TMD). The
quence, the transformed and pooled data reflected the POMS was assessed prior to and after the breathing
participant’s sensitivity for this parameter. Changes of training of the first, fourth, and sixth microcycles.
z-scores above “0” indicate a gain of function (more sen-
sitive), while changes of z-scores below “0” indicate a loss Laboratory Environment
of function (less sensitive) [45]. In addition, the initial
thresholds before the aDSB and rDSB were compared in The experimental room was sound attenuated and pro-
order to determine if carryover effects were present, vided with a diffuse light during the entire session. Music
according to Wallenstein and colleagues [46]. Although was omitted, because its exposure was found to influence

218
 Effect of Breathing on Pain Perception

physiological response to stress [61,62]. Room tempera- score of the final sample for BDI was 4.27 (⫾3.24), which
ture was kept stable at 20°C [63]. As many substances is significantly below the cutoff of 18 points indicating a
can affect the sympathetic activity measured by SCL, depressive disorder [25]. The mean score of the final
subjects were asked to refrain from drinking alcohol for at sample for anxiety (trait) was 38.70 (⫾8.80), well within the
least 48 hours, from drinking more than two cups of coffee 95% confidence interval for male and female adoles-
on the day of the measurement, and from smoking for at cents (age 15–29; 34.49 ⫾ 5.5 female adolescents,
least 2 hours before the recording [64,65]. The subjects 35.65 ⫾ 5.7 male adolescents) [30]. All remaining sub-
were encouraged to sit comfortably in a chair and rest jects (13 female adolescents, 2 male adolescents) accom-
their hands on their laps. plished both mesocycles of the study. None of the
subjects reported any undesirable side effects.
Statistics
Data Distribution

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

We used three-factorial analyses of variance for repeated
measures (general linear model) to assess the effect of the After the averaging process, mean respiration depths and
breathing intervention on z-transformed detection and rates, SCL and detection and pain thresholds in each of
pain thresholds. Within-subject factors were “intervention” the three microcycles were normally distributed (Shapiro–
(aDSB vs rDSB), “course” (microcycle 1 vs microcycle Wilks Tests for small sample sizes; qq-Plots). The POMS
4 vs microcycle 6; = time across the microcycles), and marginally failed normal distribution. Further analyses of
“session” (before vs after a 20-minute breathing training; the mood states were done with nonparametric tests
= time across one breathing session). Comparisons of (s.statistics).
initial pain thresholds were assessed with paired Student’s
t-tests. Likewise, comparison of respiration parameters Respiration
was assessed with paired Student’s t-tests. Correlations
of mean SCL changes and thermal pain threshold Mean values of respiration rates and depths were similar
changes were tested by using Pearson correlation analy- during both breathing interventions, with a significant
ses. Nonparametric Wilcoxon tests for comparison of the return to baseline during the breaks (Table 1).
mood states were conducted. Effect sizes were calculated
using Cohen’s d to examine the size of the post hoc Detection and Pain Thresholds
differences [66] allowing for correlated design (due to the
repeated measurements) [67]. Results were regarded as No significant differences of the initial detection thresholds
significant with P < 0.05. (CDT: t = 0.36, ns [0.73], WDT: t = 1.02, ns [0.32]) and
pain thresholds (CPT: t = 0.15, ns [0.88], HPT: t = -0.93,
The statistical analyses were performed using SPSS for ns [0.37]) between both intervention could be found.
Windows 17.0 (SPSS Inc., Chicago, IL, USA). Detection thresholds (CDT, WDT) and pain thresholds
(CPT, HPT) before and after breathing interventions for the
Results first, fourth, and sixth microcycles are shown in Table 2.
For statistical analyses, the detection and pain thresholds
Subjects were converted into a standard normal distribution
(z-transformation, as pointed out in the Methods). CPTs/
One male subject did not finish the second breathing HPTs and CDTs/WDTs, respectively, were then pooled in
intervention. His data were not used for further analyses order to build one compound variable for detection and
(per protocol). The mean age of the final sample of 15 one compound variable for pain perception. Analyses of
subjects was 25.1 ⫾ 2.1 years (range 23–30). The mean variance for repeated measures for the z-transformed

Table 1 Respiration parameters

aDSB rDSB
M ⫾ SD M ⫾ SD T P

Respiration depths (cm) Baseline 0.95 ⫾ 0.50 1.01 ⫾ 0.37 -0.45 ns (0.66)
Breathing blocks 2.25 ⫾ 0.70 2.02 ⫾ 0.77 1.53 ns (0.15)
Breaks 0.98 ⫾ 0.43 0.95 ⫾ 0.40 0.52 ns (0.61)
Respiration rates (min-1) Baseline 15.05 ⫾ 2.04 14.16 ⫾ 2.54 1.29 ns (0.22)
Breathing blocks 7.04 ⫾ 0.52 7.65 ⫾ 1.26 -1.10 ns (0.29)
Breaks 15.77 ⫾ 2.59 14.45 ⫾ 2.57 1.85 ns (0.09)

Comparison of the mean (⫾SD) respiration depths and rates for the baseline period, all breathing blocks, and breaks of the first,
fourth, and sixth microcycles in the aDSB and rDSB intervention.
aDSB = attentive deep and slow breathing; rDSB = relaxing deep and slow breathing; M = mean; SD = standard deviation;
T = paired test statistic; P = Significance; ns = nonsignificant.

219
Busch et al.

thermal thresholds showed significant effects for the

aDSB = attentive deep and slow breathing; rDSB = relaxing deep and slow breathing; CDT = cold detection threshold; WDT = warm detection threshold; CPT = cold pain threshold;
(⫾2.87)
(⫾3.35)
(⫾3.22)
(⫾4.04)

(⫾3.72)
(⫾3.39)
factor “session” and the interaction of “intervention*

M (⫾SD)
session” for both detection and pain thresholds. The
factor “course” and its interactions did not explain a

37.86
39.67
40.19
39.76

41.36
40.23

Detection and pain thresholds of the first, fourth, and sixth microcycles before (pre) and after (post) the breathing maneuvers in the aDSB and rDSB intervention (in °C).
Post significant portion of variance in this model (Table 3). Post
hoc analyses allocated the overall increase of pain (-0.27,
P = < 0.001, Cohen’s d = 1.01) as well as detection (0.73,
(⫾2.97) P = < 0.001; Cohen’s d = 0.88) thresholds under the con-
(⫾3.41)
(⫾4.13)
(⫾3.71)
(⫾3.79)
(⫾4.12)
dition of the rDSB, whereas no change of the overall pain
M (⫾SD)

(0.04, P = ns [0.62]; Cohen’s d = -0.09) and detection

(-0.12, P = ns [0.20]; Cohen’s d = 0.23) thresholds
38.16
39.71
39.59
39.01
38.96
40.54
HPT

HPT = heat pain threshold; SD = standard deviation; M = mean; SD = standard deviation; M1 = microcycle 1; M4 = microcycle 4, M6 = microcycle 6.
Pre

occurred in the aDSB (Figure 2).

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

SCL
(⫾3.89)
(⫾4.98)
(⫾5.59)
(⫾3.26)
(⫾4.77)
(⫾4.21)

The SCL decreased significantly during the rDSB interven-

M (⫾SD)

tion in the first, fourth, and sixth microcycles. In contrast,

25.16
23.98
21.79
24.24
21.98
20.90

there was no significant change of SCL in the aDSB

Post

intervention in any of the microcycles, rather tending

toward an increase of sympathetic arousal. The mean
overall change of SCL (D%) revealed a highly significant
(⫾3.88)

(⫾5.67)
(⫾2.99)
(⫾4.62)
(⫾3.89)
(⫾5.11)

decrease by 18% in the rDSB intervention (T = 3.88,

P = 0.002; Cohen’s d = 1.35) and a mean overall nonsig-
M (⫾SD)

nificant increase of 1% in the attentive breathing interven-

24.85
23.27
21.44
24.67
23.40
22.71

tion (T = 0.85, P = ns [0.41]; Cohen’s d = -0.23)

CPT

Pre

(Figure 2).

Correlation Analyses
(⫾0.52)
(⫾0.95)
(⫾0.54)
(⫾0.60)
(⫾0.53)
(⫾0.54)

The overall changes of the z-transformed pain thresholds

M (⫾SD)

and the overall changes of SCL were inversely correlated

33.45
33.39
33.79
33.69
33.56
33.76

with regard to the rDSB (r = -0.40; P < 0.05) but not to the
Post

aDSB (r = -0.30; P = ns [0.11]). The overall changes of

z-transformed detection thresholds and the overall
changes of SCL were not significantly correlated with
(⫾0.48)
(⫾0.73)
(⫾0.53)
(⫾0.50)
(⫾0.30)
(⫾0.41)

regard to the rDSB (r = -0.14; P = ns [0.48]) or to the

aDSB (r = -0.18; P = ns [0.35]).
M (⫾SD)

33.51
33.22
33.64
33.35
33.41
33.55
WDT

Profile of Mood States

Pre

Feelings of tension, depression, and anger were signifi-

cantly reduced after attentive and relaxed breathing exer-
(⫾0.39)
(⫾0.58)
(⫾0.95)
(⫾0.69)
(⫾0.90)
(⫾0.87)

cises, but without any significant differences between the

types of breathing intervention. Feelings of vigor, fatigue,
M (⫾SD)

and confusion did not change significantly, neither after

Detection and pain thresholds

30.96
30.74
30.18
30.43
30.08
30.01
Post

the aDSB, nor after the rDSB. Total mood disturbances

decreased significantly after both interventions, but again
without a significant difference between the types of
breathing intervention (Table 4).
(⫾0.44)
(⫾0.60)
(⫾0.91)
(⫾0.49)
(⫾0.71)
(⫾0.90)

Discussion
M (⫾SD)

31.03
30.74
30.40
30.97
30.70
30.69
CDT

Despite their frequent clinical use, the specific symptom-

Pre

related effects of DSB techniques have been barely eluci-

dated so far. We focused on the impact of two distinct
DSB techniques on mood, sympathetic arousal, and
M1
M4
M6
M1
M4
M6

especially pain perception in 15 healthy subjects. More

precisely, both breathing interventions were characterized
Table 2

by similar breathing instructions, verbal suggestions, as

(In °C)

aDSB

rDSB

well as similar breathing depths and rates. However, in

one of these interventions, the subjects performed an

220
 Effect of Breathing on Pain Perception

Table 3 ANOVA for repeated measurements for detection and pain thresholds

Detection Thresholds Pain Thresholds

Main Factors and Interactions df hypothesis df error F P F P

Intervention 1 29 3.55 ns (0.07) 0.68 ns (0.41)

Course 2 28 0.15 ns (0.98) 0.54 ns (0.59)
Session 1 29 16.08 <0.001 8.72 <0.01
Intervention * course 2 28 0.46 ns (0.63) 0.17 ns (0.84)
Intervention * session 1 29 19.02 <0.001 6.75 0.01
Course * session 2 28 0.04 ns (0.96) 2.35 ns (0.11)

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

Intervention * course * session 2 28 0.90 ns (0.42) 1.92 ns (0.17)

Analyses of variance for repeated measurements for z-transformed detection and pain thresholds. Main effects and interactions are
shown for a three-factorial model with “intervention” (attentive deep and slow breathing vs relaxing deep and slow breathing),
“course” (microcycle 1 vs microcycle 4 vs microcycle 6), and “session” (before vs after a breathing training) as within subject factors.
ANOVA = analysis of variance; F = ANOVA test statistic; P = significance; df = degrees of freedom of hypothesis and error;
ns = nonsignificant. Bold denotes significant results.

aDSB associated with a concentration task requiring per- in a psychomotor task motor speed [71] and psychomotor
sistent attentional focusing, whereas in the other interven- vigilance was even improved, even in novice meditators
tion, a DSB mode particularly aiming at pure relaxation [72].
without mental effort was chosen.
One could argue that the subjects may not be distracted
The most striking finding in the present study was a sig- equally during both interventions and that a different
nificant increase of pain thresholds in our subjects only amount of distraction between both groups may have
after the rDSB condition in all of the three microcycles, additionally contributed to the effects of the breathing
thereby indicating an attenuation of pain perception trainings on detection and pain thresholds. However, we
(becoming less sensitive). In contrast, the aDSB mode did take the view that distraction may not significantly sepa-
not alter pain and detection thresholds. rate both interventions, as some of the subjects may be
more distracted by fitting their respiration curve onto an
In keeping with our findings of increased pain thresholds ideal curve during the first intervention; some others may
after the rDSB, recent work underscores the potential be similarly distracted by dwelling on thoughts during the
utility of a DSB training in reducing pain intensity ratings second intervention.
as compared with a natural or rapid breathing mode
[14,18,68]. Intriguingly, due to the study design, the afore- Interestingly, detection as well as pain thresholds
mentioned investigations could not differentiate between increased following the rDSB, suggesting that this type of
the effect of relaxation and the effect of respiration on pain breathing intervention may have induced a general loss of
perception. In detail, a control group was missing, which somatosensory perception. In support of our observation,
had to breathe at similar respiration depths and rates, a multitude of studies report significant decreases in
however without the possibility of relaxation. For this late somatosensory event-related potentials (SERP) in
reason, our study suggests relaxation as an essential pre- response to nociceptive stimulation during hypnotic anal-
requisite of a DSB technique in efficiently modulating pain gesia [73,74]. De Pascalis et al. found increases in
perception. sensory and pain thresholds mirrored by a reduction of
certain components of the cortical SERP across different
A nonspecific psychomotor effect following the DSB inter- hypnosis conditions [75]. In this context, it is tempting to
vention may have been associated with relaxation, thus speculate that relaxation may exert its effect on soma-
providing a nonspecific increase in detection and pain tosensory perception by inhibiting thalamocortical activity
thresholds, as fatigue and a reduction in alertness both via a frontal cortex feedback loop [73].
were found to slightly increase mean reaction times [69].
However, in our study, fatigue was equivalent in both Moreover, our study revealed a significant decrease of
groups and did not change significantly during the DSB SCLs in all of the three microcycles indicating a slowing
interventions. In addition, it has been shown recently that down of sympathetic activity clearly restricted to the rDSB
mental fatigue did not affect the temporal preparation time condition. The changes of pain thresholds and SCLs were
in a choice reaction time task [70]. Moreover, such inversely correlated in the rDSB condition.
changes, if any, are unlikely to explain a threshold differ-
ence of approximately 1°C in our study, as this would As the changes of SCL and detection and pain thres-
afford reduced reaction times of almost 1 second. Inter- holds were only weakly correlated, we would strongly
estingly, following two yoga-based relaxation techniques, suggest discussing any potential (causal or not causal)

221
Busch et al.

0,20
Microcycle 1 Microcycle 4 Microcycle 6 overall
aDSB
Skin conductance levels

0,10
rDSB
0,00
Δ Change (%)

p=ns(.41)
-0,10

-0,20

-0,30
***

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

-0,40 p=.02

-0,50
0,20
Gain of function
0,00
Loss of function
Detection thresholds

-0,20
p=ns(.20)
z-transformed

-0,40
Δ Change

-0,60

-0,80

-1,00

-1,20
***
-1,40 p<.001
0,20

0,10
Gain of function
0,00
p=ns(.62)
Pain thresholds
z-transformed

Loss of function
-0,10
Δ Change

-0,20

-0,30

-0,40
***
p<.001
-0,50

-0,60

Figure 2 Mean changes of skin conductance level and detection and pain thresholds. Mean changes
(⫾standard deviation) and post hoc analyses of skin conductance levels and of the z-transformed detection
and pain thresholds from the first, fourth, and sixth microcycles and overall for both breathing interventions.
aDSB = attentive deep and slow breathing; rDSB = relaxing deep and slow breathing; ***P < 0.001;
ns = nonsignificant.

222
 Effect of Breathing on Pain Perception

relationships carefully. However, our results are in agree-

Nonparametric comparisons (Wilcoxon tests) of the profile of mood states and the total mood disturbances before (pre) and after (post) the aDSB and rDSB intervention. Values are
(0.83)
(0.68)
(0.33)
(0.83)
(0.17)
(0.51)
(0.55)

aDSB = attentive deep and slow breathing; rDSB = relaxing deep and slow breathing; M = mean; SD = standard deviation; Z = test statistic; P = significance; ns = nonsignificant.
ment with other studies investigating the effects of deep

aDSB vs rDSB
Comparison relaxation techniques on the autonomic nerve system. As

ns
ns
ns
ns
ns
ns
ns
P
has already been demonstrated, the basal electrodermal
activity was significantly reduced during meditation [76],

-0.21
-0.41
-0.98
-0.22
-1.36
-0.66
-0.59
mindfulness-based stress reduction [77], or an integrative
Z body–mind training [78]. In contrast, the aDSB mode
along with a concentration task did not reduce, but rather

ns (0.53)
ns (0.19)
ns (0.06)
<0.001 tended to increase the sympathetic arousal in our sample.
<0.01 As a potential mechanism underlying this phenomenon,
<0.01

<0.01
the attentional demands during the aDSB may have main-
P

tained vegetative arousal (finally resulting in maintenance

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

or even increase of SCLs). This hypothesis is in agreement
-3.41
-2.94
-2.93
-0.63
-1.32
-1.88
-2.73
with a study conducted by Cappo and Holmes, demon-
strating that the effort involved in accomplishing a chal-
Z

lenging attentive breathing exercise did not reduce, but

(⫾12.42)

even heightened subjects’ sympathetic arousal during the

(⫾1.60)
(⫾3.13)
(⫾1.42)
(⫾5.25)
(⫾3.07)
(⫾2.51)

practice period [16]. In support of this finding, control of

respiration combined with an attention tracing condition
M (⫾SD)

did not reduce the subjects’ stress responses [79]. More-

2.22
2.93
1.80
14.91
6.49
4.47
3.04
Post

over, in a clinical population characterized by high trait

anxiety scores, rDSB without concentration on a pacing
tone resulted in a greater reduction of skin conductance
(⫾19.91)

responses than conventional aDSB, providing greater

(⫾2.82)
(⫾6.12)
(⫾1.19)
(⫾3.99)
(⫾3.41)
(⫾2.59)

effort on the breathing challenge [80].

M (⫾SD)

Based on these data, it is tempting to conclude that the

6.53
6.13
4.42
15.96
7.69
5.04
13.98
rDSB

required concentration component in our aDSB group

Pre

may have attenuated the sympathetic decrease. There-

fore, a DSB pattern does not decrease stress responses
ns (0.20)
ns (0.99)
ns (0.44)

inevitably, but may strongly be dependent on the magni-

<0.001

tude of relaxation. In general, biological evidence points to

<0.01
<0.01

<0.05

a strong interplay between autonomic functioning and

P

pain perception. In this context, painful cold pressor tests

have been shown to elicit sympathetic activity [81]. More-
-3.41
-3.13
-2.67
-1.29
+0.01
-0.77
-1.99

over, either the elevation [82] or the hyperreactivity [83,84]

Z

of the basal sympathetic tone was a finding frequently

replicated in patients suffering from chronic pain syn-
(⫾10.54)

dromes such as fibromyalgia. Elevated SCL was associ-

(⫾2.50)
(⫾2.09)
(⫾1.40)
(⫾4.79)
(⫾3.85)
(⫾2.04)

ated with increased anxiety and muscle pain in these

patients [85]. Reduced autonomic responsiveness and
M (⫾SD)

pain perception after DSB exercises in patients suffering

means of the first, fourth, and sixth microcycles (overall).
2.40
1.82
1.67
15.56
6.40
4.44
0.90
Post

from fibromyalgia may result from a complex modulation

of sympathetic arousal and pain perception [86] as
defined by a downregulation in stress activity [87].
(⫾21.44)
(⫾3.66)
(⫾5.86)
(⫾3.34)
(⫾4.54)
(⫾4.21)
(⫾2.98)

With regard to the mood states, we found a significant

reduction of tension, anger, and depressive feelings after
M (⫾SD)
Profile of mood states

both breathing interventions, indicating a more general

aDSB

6.62
5.42
3.62
16.42
6.24
4.71
9.87

reduction of the stress level. As has been shown, several

Pre

studies parallel to our findings have demonstrated a

reduction of negative feelings due to breathing exercises
Total mood disturbances

both in patients with chronic pain and healthy controls

[2–4,18,35,88–92]. In contrast to our findings that only
rDSB is able to selectively modify autonomic response
and pain perception, mood processing was affected in a
similar manner irrespective of the breathing maneuver. As
Depression

Confusion

a matter of fact, self-reports of mood or stress levels do

Table 4

Tension

Fatigue

not necessarily conform to autonomic responses in stress

Anger
Vigor

reduction tasks [77]. Moreover, mood in itself was not

sufficient to explain the changes of pain and nociceptive

223
Busch et al.

processing in patients with chronic back pain [93] or major ings. However, we are aware that a potential decrease in
depression [94], indicating that affective and sensory pain experimental pain perception due to DSB interventions in
processing may follow different courses. Interestingly, pain patients does not inevitably mean a significant alle-
DSB without a relaxation component was shown to be viation of their clinical pain. Further studies should be
ineffective in reducing pain levels, although most of the provided focusing on the effects of DSB techniques on
subjects felt it was useful and increased the patient’s patients with different pain syndromes and including
feeling of rapport [95]. Therefore, different factors apart different pain modalities.
from relaxation may drive mood improvement in breathing
techniques. One possible explanation may be based on Conclusions
psychological reasons, grounded in the subject’s expec-
tation, that a breathing intervention, commonly known as Taken together, our results suggest that the way of breath-
an effective stress reduction strategy, is able to attenuate ing decisively influences autonomic and pain processing.

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

the feelings of tension and anger. Alternatively, considering Based on an experimental study design, we could extend
the close structural connection between respiratory most recent work identifying DSB together with relaxation
regions and neurons within the amygdala complex as the essential feature in the modulation of sympathetic
[96,97], breathing may more directly modulate mood via arousal and pain perception. Our finding of a similar
biological mechanisms. Irrespective of the mode of action, decrease of sympathetic activity together with an attenu-
our data lend further support to the notion that breath- ation of pain perception in all three microcycles may
ing interventions are effective in influencing affective suggest that a rDSB intervention is easy to learn and may
processing. facilitate an inhibitory influence on pain processing. In
contrast, changes in affect processing seem to depend on
Both interventions were executed successively and not different biological factors as both breathing modes as
counterbalanced. We are aware of the fact that this may used in our study lead to similar effects in mood improve-
be a limiting factor for the interpretation of our results. ment. Consequently, our findings point to a more individu-
However, neither during the first nor during the second alized use of DSB guided by clinical core features
intervention we found an effect of breathing trainings on especially with regard to pain-related diseases. Disentan-
detection and pain thresholds, SCL, or mood states over gling the symptom specific relevance of distinct DSB com-
the course of the study (= time across the microcycles). ponents may pave the way to a further increase in its use
Therefore, we assume that all the more after the end of and optimize the therapeutic value of breathing tech-
one mesocycle no relevant long-term effects had an niques representing a broadly used approach in a variety
impact on the following mesocycle a half year later. of diseases such as chronic pain, which are characterized
Moreover, although the experimenter provided similar by obvious limitations in drug treatment.
respiration rates and depths during both interventions in
our subjects, the attentive vs the relaxing breathing tech- Acknowledgments
nique was quite different in the way of performance.
Hence, we do not think that our subjects were essen- The authors thank Ms. Mariya Kozhuharova for her assis-
tially capable to use their experience from the aDSB for tance carrying out the study.
the rDSB. Anyway, hypothesizing subtle training effects,
we think that a period of a half year between both inter- Conflicts of Interest/Grants or Supports
ventions is long enough to “wash-out” these small car-
ryover effects. Finally, we ascertained that there was no None of the authors declares a financial or other conflict of
breathing training the weeks prior to the second inter- interest. No grants or any other forms of support were
vention, as the participants were instructed not to prac- given for this study.
tice at home (which was not possible due to the
absence of adequate sophisticated technical equip- References
ment) and it was not allowed to participate in breathing 1 Gilbert C. Clinical applications of breathing regulation.
trainings or meditation programs. Beyond anxiety management. Behav Modif 2003;27:
692–709.
Some of the issues of our present design were chosen,
because they are geared toward patients with chronic 2 Brown RP, Gerbarg PL. Sudarshan Kriya Yogic
pain. For example, we used a breathing training of 20 breathing in the treatment of stress, anxiety, and
minutes (separated into single 5-minute blocks) over the depression. Part II—Clinical applications and guide-
course of several weeks in our study, because this dura- lines. J Altern Complement Med 2005;11:711–7.
tion refers to a conventional breathing session in the
scope of a therapeutical application [16,38]. We think 3 Kim SD, Kim HS. Effects of a relaxation breathing
that our result may provide an encouraging rationale for exercise on anxiety, depression, and leukocyte in
the investigation of the effect on DSB techniques in hemopoietic stem cell transplantation patients. Cancer
patients with pain syndromes with a similar breathing Nurs 2005;28:79–83.
regimen. The effect sizes of detection (1.01) and pain
(0.88) threshold changes due to the relaxed DSB were 4 Han JN, Stegen K, De Valck C, Clement J, Van de
high suggesting a profound clinical relevance of our find- Woestijne KP. Influence of breathing therapy on com-

224
 Effect of Breathing on Pain Perception

plaints, anxiety and breathing pattern in patients 18 Zautra AJ, Fasman R, Davis MC, Craig AD. The effects
with hyperventilation syndrome and anxiety disorders. of slow breathing on affective responses to pain
J Psychosom Res 1996;41:481–93. stimuli: An experimental study. Pain 2010;149:12–8.

5 Jerath R, Edry JW, Barnes VA, Jerath V. Physiology of 19 McDonnell L, Bowden ML. Breathing management:
long pranayamic breathing: Neural respiratory ele- A simple stress and pain reduction strategy for use
ments may provide a mechanism that explains how on a pediatric service. Issues Compr Pediatr Nurs
slow deep breathing shifts the autonomic nervous 1989;12:339–44.
system. Med Hypotheses 2006;67:566–71.
20 Arch JJ, Craske MG. Mechanisms of mindfulness:
6 Mourya M, Mahajan AS, Singh NP, Jain AK. Effect of Emotion regulation following a focused breathing
slow- and fast-breathing exercises on autonomic func- induction. Behav Res Ther 2006;44:1849–58.

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

tions in patients with essential hypertension. J Altern
Complement Med 2009;15:711–7. 21 First M, Spitzer R, Gibbon M, Williams J. Structured
Clinical Interview for DSM-IV Axis I Disorders, Clinician
7 Pal GK, Velkumary S, Madanmohan M. Effect of short- Version (SCID-CV). Washington, DC: American
term practice of breathing exercises on autonomic Psychiatric Press, Inc.; 1996.
functions in normal human volunteers. Indian J Med
Res 2004;120:115–21. 22 Wittchen HU, Zaudig M, Fydrich T. SKID—
Strukturiertes Klinisches Interview Für DSM-IV. Göttin-
8 Raghuraj P, Telles S. Effect of yoga-based and forced gen: Hogrefe-Verlag; 1997.
uninostril breathing on the autonomic nervous system.
Percept Mot Skills 2003;96:79–80. 23 Lobbestael J, Leurgans M, Arntz A. Inter-rater reliabil-
ity of the Structured Clinical Interview for DSM-IV Axis
9 Sydorchuk LP, Tryniak MH. Effect of the special I Disorders (SCID I) and Axis II Disorders (SCID II). Clin
breathing exercises on the autonomic regulation of the Psychol Psychother 2011;18:75–9.
functional state of respiration and muscular systems.
Lik Sprava 2005;3:44–7. 24 Shear MK, Greeno C, Kang J, et al. Diagnosis of
nonpsychotic patients in community clinics. Am J
10 Nardi AE, Freire RC, Zin WA. Panic disorder and Psychiatry 2000;157:581–7.
control of breathing. Respir Physiol Neurobiol
2009;167:133–43. 25 Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J.
An inventory for measuring depression. Arch Gen
11 Sovik R. The science of breathing—The yogic view. Psychiatry 1961;4:561–71.
Prog Brain Res 2000;122:491–505.
26 Hautzinger M. The Beck Depression Inventory in
12 Li TY, Yeh ML. The application of qi-gong therapy to clinical practice. Nervenarzt 1991;62:689–96.
health care. Hu Li Za Zhi 2005;52:65–70.
27 Beck AT, Steer RA, Garbin MG. Psychometric prop-
13 Li JX, Hong Y, Chan KM. Tai chi: Physiological char- erties of the Beck Depression Inventory: Twenty-five
acteristics and beneficial effects on health. Br J Sports years of evaluation. Clin Psychol Rev 1988;8:77–100.
Med 2001;35:148–56.
28 Richter P, Werner J, Heerlein A, Kraus A, Sauer H. On
14 Chalaye P, Goffaux P, Lafrenaye S, Marchand S. Res- the validity of the Beck Depression Inventory. A review.
piratory effects on experimental heat pain and cardiac Psychopathology 1998;31:160–8.
activity. Pain Med 2009;10:1334–40.
29 Spielberger CD, Gorsuch RL, Lushene RE. Manual for
15 Cahalin LP, Hernandez ED, Matsuo Y. Diaphragmatic the State-Trait Anxiety Inventory. Palo Alto, CA: Con-
breathing training: Further investigation needed. Phys sulting Psychologists Press; 1970.
Ther 2005;85:369–70, author reply 70-3.
30 Laux L, Glanzmann P, Schaffner P, Spielberger C.
16 Cappo BM, Holmes DS. The utility of prolonged res- STAI—State Trait Angst Inventar. Göttingen: Hogrefe-
piratory exhalation for reducing physiological and psy- Verlag; 1981.
chological arousal in non-threatening and threatening
situations. J Psychosom Res 1984;28:265–73. 31 Tilton SR. Review of the State-Trait Anxiety Inventory
(STAI). News Notes 2008;48:1–3.
17 Beary JF, Benson H. A simple psychophysiologic
technique which elicits the hypometabolic changes of 32 Scheibenbogen O, Prieler J. Biofeedback Aided Psy-
the relaxation response. Psychosom Med 1974;36: chotherapy (BAP). New York: Kluwer Academic/
115–20. Plenum Publishers; 2002.

225
Busch et al.

33 Marx R, Scheibenbogen O. Visualisierung der Respi- 46 Wallenstein S, Fisher AC. The analysis of the two-
ration als verhaltenstherapeutische Intervention period repeated measurements crossover design
bei Angsterkrankungen. Foto-Medico—Internationaler with application to clinical trials. Biometrics 1977;33:
Arbeitskreis für Medizinische Bilddokumentation 261–9.
1999;9:7–9.
47 Venables PH, Martin I. The relation of palmar sweat
34 Brown RP, Gerbarg PL. Yoga breathing, meditation, gland activity to level of skin potential and conduc-
and longevity. Ann N Y Acad Sci 2009;1172:54–62. tance. Psychophysiology 1967;3:302–11.

35 Paul G, Elam B, Verhulst SJ. A longitudinal study 48 Fuhrer MJ. Stimulus site effects on skin conductance
of students’ perceptions of using deep breathing responses from the volar surfaces. Psychophysiology
1974;11:365–71.

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

meditation to reduce testing stresses. Teach Learn
Med 2007;19:287–92.
49 Germana J. Effects on behavioral responding on skin
conductance level. Psychol Rep 1969;24:599–605.
36 Wolkove N, Kreisman H, Darragh D, Cohen C, Frank
H. Effect of transcendental meditation on breathing
50 Breimhorst M, Sandrock S, Fechir M, et al. Do inten-
and respiratory control. J Appl Physiol 1984;56:607–
sity ratings and skin conductance responses reliably
12.
discriminate between different stimulus intensities in
experimentally induced pain&quest; J Pain 2011;12:
37 Harris VA, Katkin ES, Lick JR, Habberfield T. Paced 61–70.
respiration as a technique for the modification of
autonomic response to stress. Psychophysiology 51 Storm H. Changes in skin conductance as a tool to
1976;13:386–91. monitor nociceptive stimulation and pain. Curr Opin
Anaesthesiol 2008;21:796–804.
38 Strauss-Blasche G, Moser M, Voica M, et al. Relative
timing of inspiration and expiration affects respiratory 52 Venables PH, Christie I. Electrodermal activity. In:
sinus arrhythmia. Clin Exp Pharmacol Physiol Martin I, Venables PH, eds. Techniques in Psycho-
2000;27:601–6. physiology. Chichester: Wiley; 1973:3–67.

39 Ehrmann W. Handbuch der Atemtherapie. Ahlerstedt: 53 McNair PM, Lorr M, Droppleman LF. POMS Manual:
Param-Verlag; 2004. Profile of Mood States. San Diego, CA: Educational
and Industrial Testing Service; 1992.
40 Wang SZ, Li S, Xu XY, et al. Effect of slow abdominal
breathing combined with biofeedback on blood pres- 54 McNair D, Lorr M, Droppleman L. Profile of Mood
sure and heart rate variability in prehypertension. States POMS—Ein Verfahren zur Messung von Stim-
J Altern Complement Med 2010;16:1039–45. mungszuständen. In: CIPS, ed. Internationale Skalen
für Psychiatrie. Weinheim: Beltz Test Gesellschaft;
41 Minyaev V, Petushkov M. Voluntary control of thoracic 1981:1–5.
and abdominal respiratory movements. Hum Physiol
2005;31:44–8. 55 Lira FT, Fagan TJ. The profile of mood states: Norma-
tive data on a delinquent population. Psychol Rep
1978;42:640–2.
42 Fruhstorfer H, Lindblom U, Schmidt WC. Method for
quantitative estimation of thermal thresholds in
56 Pollock V, Cho DW, Reker D, Volavka J. Profile of
patients. J Neurol Neurosurg Psychiatry 1976;39:
Mood States: The factors and their physiological cor-
1071–5.
relates. J Nerv Ment Dis 1979;167:612–4.
43 Rolke R, Baron R, Maier C, et al. Quantitative sensory 57 Cui J, Matsushima E, Aso K, Masuda A, Makita K.
testing in the German Research Network on Neuro- Psychological features and coping styles in patients
pathic Pain (DFNS): Standardized protocol and refer- with chronic pain. Psychiatry Clin Neurosci 2009;63:
ence values. Pain 2006;123:231–43. 147–52.

44 Magerl W, Krumova EK, Baron R, et al. Reference 58 Jungquist CR, O’Brien C, Matteson-Rusby S, et al.
data for quantitative sensory testing (QST): Refined The efficacy of cognitive-behavioral therapy for insom-
stratification for age and a novel method for statistical nia in patients with chronic pain. Sleep Med 2010;11:
comparison of group data. Pain 2010;151:598–605. 302–9.

45 Glass GV, Stanley JC. Statistical Methods in Educa- 59 Malouff JM, Schutte NS, Ramerth W. Evaluation of a
tion and Psychology. Boston, MA: Allyn & Bacon; short form of the POMS-Depression scale. J Clin
1970. Psychol 1985;41:389–91.

226
 Effect of Breathing on Pain Perception

60 Casten RJ, Parmelee PA, Kleban MH, Lawton MP, changes to noxious stimuli and 2. Transfer learning to
Katz IR. The relationships among anxiety, depression, reduce chronic low back pain. Int J Clin Exp Hypn
and pain in a geriatric institutionalized sample. Pain 1998;46:92–132.
1995;61:271–6.
74 Zachariae R, Bjerring P. Laser-induced pain-related
61 Knight WE, Rickard Ph DN. Relaxing music prevents brain potentials and sensory pain ratings in high and
stress-induced increases in subjective anxiety, systolic low hypnotizable subjects during hypnotic sugges-
blood pressure, and heart rate in healthy males and tions of relaxation, dissociated imagery, focused anal-
females. J Music Ther 2001;38:254–72. gesia, and placebo. Int J Clin Exp Hypn 1994;42:
56–80.
62 Iwanaga M, Moroki Y. Subjective and physiological

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

responses to music stimuli controlled over activity and 75 De Pascalis V, Magurano MR, Bellusci A. Pain percep-
preference. J Music Ther 1999;36:26–38. tion, somatosensory event-related potentials and skin
conductance responses to painful stimuli in high, mid,
63 Boucsein W. Elektrodermal Activity. Berlin: Springer; and low hypnotizable subjects: Effects of differential
1980. pain reduction strategies. Pain 1999;83:499–508.

64 Lyvers M, Miyata Y. Effects of cigarette smoking on 76 Travis F. Autonomic and EEG patterns distinguish
electrodermal orienting reflexes to stimulus change transcending from other experiences during Trans-
and stimulus significance. Psychophysiology 1993;30: cendental Meditation practice. Int J Psychophysiol
231–6. 2001;42:1–9.

65 Zahn TP, Rapoport JL. Autonomic nervous system 77 Lush E, Salmon P, Floyd A, et al. Mindfulness medi-
effects of acute doses of caffeine in caffeine users and tation for symptom reduction in fibromyalgia: Psycho-
abstainers. Int J Psychophysiol 1987;5:33–41. physiological correlates. J Clin Psychol Med Settings
2009;16:200–7.
66 Cohen J. Statistical Power Analysis for the Behavioral
Sciences. Hillsdale, NJ: Lawrence Earlbaum Associ- 78 Tang YY, Ma Y, Fan Y, et al. Central and autonomic
ates; 1988. nervous system interaction is altered by short-term
meditation. Proc Natl Acad Sci U S A 2009;106:8865–
67 Morris SB, DeShon RP. Combining effect size esti- 70.
mates in meta-analysis with repeated measures
and independent-groups designs. Psychol Methods 79 Holmes DS, McCaul KD, Solomon S. Control of res-
2002;7:105–25. piration as a means of controlling responses to threat.
J Pers Soc Psychol 1978;36:198–204.
68 Grant JA, Rainville P. Pain sensitivity and analgesic
effects of mindful states in Zen meditators: A cross- 80 Clark ME, Hirschman R. Effects of paced respiration
sectional study. Psychosom Med 2009;71:106–14. on anxiety reduction in a clinical population. Biofeed-
back Self Regul 1990;15:273–84.
69 Appelle S, Oswald LE. Simple reaction time as a func-
tion of alertness and prior mental activity. Percept Mot 81 Victor RG, Leimbach WN Jr, Seals DR, Wallin BG,
Skills 1974;38:1263–8. Mark AL. Effects of the cold pressor test on muscle
sympathetic nerve activity in humans. Hypertension
70 Langner R, Steinborn MB, Chatterjee A, Sturm W, 1987;9:429–36.
Willmes K. Mental fatigue and temporal preparation
in simple reaction-time performance. Acta Psychol 82 Martinez-Lavin M, Hermosillo AG. Autonomic nervous
(Amst) 2010;133:64–72. system dysfunction may explain the multisystem
features of fibromyalgia. Semin Arthritis Rheum
71 Subramanya P, Telles S. Performance on psycho- 2000;29:197–9.
motor tasks following two yoga-based relaxation
techniques. Percept Mot Skills 2009;109:563–76. 83 Qiao ZG, Vaeroy H, Morkrid L. Electrodermal and
microcirculatory activity in patients with fibromyalgia
72 Kaul P, Passafiume J, Sargent CR, O’Hara BF. Medi- during baseline, acoustic stimulation and cold pressor
tation acutely improves psychomotor vigilance, and tests. J Rheumatol 1991;18:1383–9.
may decrease sleep need. Behav Brain Funct
2010;6:1–9. 84 Cohen H, Neumann L, Shore M, et al. Autonomic
dysfunction in patients with fibromyalgia: Application
73 Crawford HJ, Knebel T, Kaplan L, et al. Hypnotic of power spectral analysis of heart rate variability.
analgesia: 1. Somatosensory event-related potential Semin Arthritis Rheum 2000;29:217–27.

227
Busch et al.

85 Ozgocmen S, Ozyurt H, Sogut S, Akyol O. Current 92 Kubzansky LD, Wright RJ, Cohen S, et al. Breathing
concepts in the pathophysiology of fibromyalgia: easy: A prospective study of optimism and pulmonary
The potential role of oxidative stress and nitric oxide. function in the normative aging study. Ann Behav Med
Rheumatol Int 2006;26:585–97. 2002;24:345–53.

86 Hassett AL, Radvanski DC, Vaschillo EG, et al. A pilot 93 Dickens C, Jayson M, Sutton C, Creed F. The rela-
study of the efficacy of heart rate variability (HRV) tionship between pain and depression in a trial using
biofeedback in patients with fibromyalgia. Appl Psy- paroxetine in sufferers of chronic low back pain. Psy-
chophysiol Biofeedback 2007;32:1–10. chosomatics 2000;41:490–9.

87 Craig AD. A new view of pain as a homeostatic 94 Kundermann B, Hemmeter-Spernal J, Huber MT,
emotion. Trends Neurosci 2003;26:303–7. Krieg JC, Lautenbacher S. Effects of total sleep dep-

Downloaded from https://academic.oup.com/painmedicine/article/13/2/215/1936333 by guest on 12 June 2024

rivation in major depression: Overnight improvement
88 Harvey JR. The effect of yogic breathing exercises on of mood is accompanied by increased pain sensitivity
mood. J Am Soc Psychosom Dent Med 1983;30:39– and augmented pain complaints. Psychosom Med
48. 2008;70:92–101.

89 Kim KS, Lee SW, Choe MA, et al. Effects of abdominal 95 Downey LV, Zun LS. The effects of deep breath-
breathing training using biofeedback on stress, ing training on pain management in the emer-
immune response and quality of life in patients with a gency department. South Med J 2009;102:
mastectomy for breast cancer. Taehan Kanho Hakhoe 688–92.
Chi 2005;35:1295–303.
96 Fulwiler CE, Saper CB. Subnuclear organization of the
90 Tweeddale PM, Rowbottom I, McHardy GJ. Breathing efferent connections of the parabrachial nucleus in the
retraining: Effect on anxiety and depression scores in rat. Brain Res 1984;319:229–59.
behavioural breathlessness. J Psychosom Res
1994;38:11–21. 97 Yasui Y, Tsumori T, Oka T, Yokota S. Amygdaloid
axon terminals are in contact with trigeminal premotor
91 Gaines T, Barry LM. The effect of a self-monitored neurons in the parvicellular reticular formation of
relaxation breathing exercise on male adolescent the rat medulla oblongata. Brain Res 2004;1016:129–
aggressive behavior. Adolescence 2008;43:291–302. 34.

228