Inhalation/Exhalation Ratio Modulates the Effect of Slow Breathing on Heart Rate Variability and Relaxation

Abstract

Slow breathing is widely applied to improve symptoms of hyperarousal, but it is unknown whether its beneficial effects relate to the reduction in respiration rate per se, or, to a lower inhalation/exhalation (i/e) ratio. The present study examined the effects of four ventilatory patterns on heart rate variability and self-reported dimensions of relaxation. Thirty participants were instructed to breathe at 6 or 12 breaths/min, and with an i/e ratio of 0.42 or 2.33. Participants reported increased relaxation, stress reduction, mindfulness and positive energy when breathing with the low compared to the high i/e ratio. A lower compared to a higher respiration rate was associated only with an increased score on positive energy. A low i/e ratio was also associated with more power in the high frequency component of heart rate variability, but only for the slow breathing pattern. Our results show that i/e ratio is an important modulator for the autonomic and subjective effects of instructed ventilatory patterns.

Full text

Inhalation/Exhalation Ratio Modulates the Effect of Slow
Breathing on Heart Rate Variability and Relaxation
Ilse Van Diest •Karen Verstappen •Andre
´E. Aubert •
Devy Widjaja •Debora Vansteenwegen •Elke Vlemincx
!Springer Science+Business Media New York 2014
Abstract Slow breathing is widely applied to improve
symptoms of hyperarousal, but it is unknown whether its
beneficial effects relate to the reduction in respiration rate
per se, or, to a lower inhalation/exhalation (i/e) ratio. The
present study examined the effects of four ventilatory
patterns on heart rate variability and self-reported dimen-
sions of relaxation. Thirty participants were instructed to
breathe at 6 or 12 breaths/min, and with an i/e ratio of 0.42
or 2.33. Participants reported increased relaxation, stress
reduction, mindfulness and positive energy when breathing
with the low compared to the high i/e ratio. A lower
compared to a higher respiration rate was associated only
with an increased score on positive energy. A low i/e ratio
was also associated with more power in the high frequency
component of heart rate variability, but only for the slow
breathing pattern. Our results show that i/e ratio is an
important modulator for the autonomic and subjective
effects of instructed ventilatory patterns.
Keywords Relaxation !Breathing !Respiration !Heart
rate variability !RSA
Introduction
Throughout history and in very different cultures, people
have searched for behavioral strategies that can be helpful
in reducing symptoms associated with a variety of stress-
related disorders and conditions, such as panic and anxiety,
asthma, chronic pain, hypertension and cardiovascular
complaints. One widely applied strategy, used both as a
stand-alone or as a component of a more encompassing
intervention, involves the voluntary modification of the
breathing pattern (Cappo and Holmes 1984). The cardiac
and respiratory changes elicited by breathing instructions
are often assumed to establish a relaxation response. Psy-
chologically, breathing exercises typically induce an
increased focus on internal sensations and a relative neglect
of external stimuli (Vlemincx et al. 2007). Physiologically,
most breathing exercises aim to reduce sympathetic
activity and to increase parasympathetic (or vagal) activity
(Benson et al. in van Dixhoorn 1998).
Vagal efferent outflow to the heart is commonly studied
non-invasively by means of heart rate variability (HRV)
analysis. The use of HRV measures is often inspired by the
polyvagal theory of Porges (2007). From an evolutionary
perspective, this author links emotional and stress-related
disorders to a decreased vagal activity to the heart, which
could be indexed by a reduced respiratory sinus arrhythmia
(RSA, the rising heartbeat during inhalation and the
decrease in heart rate that follows during exhalation;
Grossman and Taylor 2007). Both peripheral and central
mechanisms as well as their interaction likely contribute to
the phenomenon of RSA. Eckberg (2003) refers to a central
I. Van Diest (&)!K. Verstappen !D. Vansteenwegen !
E. Vlemincx
Faculty of Psychology and Educational Science, KULeuven,
Tiensestraat 102, 3000 Leuven, Belgium
e-mail: Ilse.VanDiest@ppw.kuleuven.be
A. E. Aubert
Department of Cardiology, KULeuven, UZ Gasthuisberg O/N1
704, Herestraat 49, 3000 Leuven, Belgium
D. Widjaja
Department of Electrical Engineering (ESAT) – STADIUS,
KU Leuven, Kasteelpark Arenberg 10, Box 2446, 3001 Leuven,
Belgium
D. Widjaja
Future Health Department, iMinds, Kasteelpark Arenberg 10,
Box 2446, 3001 Leuven, Belgium
123
Appl Psychophysiol Biofeedback
DOI 10.1007/s10484-014-9253-x

oscillator that gates autonomic outflow at respiratory
rhythms. A peripheral contribution of respiratory gating of
vagal outflow to the heart comes from the phasic increases
in intra-thoracic pressures caused by breathing, thereby
phasically stimulating baroreceptors and concomitant
autonomic outflow to the heart, leading to cardiac accel-
eration during inspiration and deceleration during expira-
tion (Berntson et al. 1993).
Some authors consider both the amplitude of the high
frequency band (HF; 0.15–0.40 Hz) of heart rate variability
(in the frequency domain) and RSA (in the time domain) to
be valid indicators of vagal tone and of a good prefrontal
inhibitory capacity allowing for the inhibition of subcorti-
cal sympatho-excitatory stress responses (Thayer and
Brosschot 2005; Thayer and Lane 2009; Thayer et al.
2012). Higher HF power or RSA would correlate with a
better capacity to adapt to the environment and induces a
calm, but alert state (Brown and Gerbarg 2005). Heart rate
variability is also a marker of cardiovascular health and
autonomic homeostatic control (Lehrer et al. 1999; Thayer
and Brosschot 2005). Since both HF-HRV and RSA are
strongly dependent on the respiratory rate and the tidal
volume (Ritz et al. 2001; Sakakibara and Hayano 1996), it
is indeed conceivable that voluntary changes in the
breathing pattern may alter efferent autonomic activity. As
such, respiration may be an interesting interface to
behaviourally change autonomic activity and stress
responses.
Although most breathing techniques described in the
literature involve a slowing down of the breathing pattern,
there is quite some variation in the specifics of the applied
techniques, as well as in the underlying rationale and the
complaints they are thought to improve. Many breathing
exercises originate from the yoga tradition. Often, the
instruction of yoga breathing comprises a diaphragmatic
breathing or ‘‘tanden breathing’’ (Lehrer et al. 1999), as
well as a nasal inhalation combined with an exhalation
through the mouth, which is slow and involves a pause
between the inspiration and the expiration (Miles 1964).
According to yoga, the ideal breath rate is situated around
six breaths per minute, with an exhalation that is twice as
long as the inhalation (ratio 1:2). General yogic breathing
is believed to stimulate a good mental health, as well as a
state of calm alertness, mental focus and stress tolerance,
by means of several mechanisms.
Interestingly, with a respiratory rate of about six breaths
a minute, variations in heart rate that follow respiration
shift from the HF to the low frequency band (LF;
0.04–15 Hz; Berntson et al. 1997; Grossman and Kollai
1993). Also, RSA amplitude seems to reach a maximum at
a respiration rate of about 6 breaths minute, likely because
at that rate breathing occurs at a resonance frequency of the
cardiovascular system, caused by the rhythm in heart rate
produced by the baroreflex (spontaneous oscillations in
blood pressure that are reflexively compensated for by
changes in heart rate) in that individual (Giardino et al.
2003; Song and Lehrer 2003). At resonance frequency,
RSA and baroreflex effects mutually stimulate each other,
causing very high oscillations in heart rate, thereby stim-
ulating modulatory effects of both processes, and appar-
ently increasing the power of modulatory processes in the
cardiorespiratory system (Lehrer et al. 2000). Based on this
phenomenon, Lehrer and colleagues have developed a
particular form of HRV biofeedback training, called
‘Resonance frequency training’, in which people regularly
exercise breathing at their particularly resonance frequency
(Lehrer et al. 2000). Growing evidence indicates that the
intervention is associated with beneficial clinical outcomes
in a variety of disorders (hypertension, asthma, depression
and chronic pain) and with an increased sensitivity of the
baroreflex (e.g., Nolan et al. 2010; Wheat and Larkin 2010;
Lehrer et al. 2003,2004).
Breathing retraining is also often applied to reduce
symptoms of panic or anxiety. Typically, it implies
adopting a diaphragmatic, slow, regular and shallow
breathing pattern (Meuret et al. 2004). Meuret et al. (2008)
introduced a new, brief, capnometry-assisted breathing
retraining therapy (BRT), aimed at increasing self-moni-
tored end-tidal pCO
2
(to reach or maintain a pCO
2
range
higher than 35 mmHg) and reducing respiration rate by
means of breathing exercises. Such breathing retraining
was generally found to reduce panic frequency and to
normalise pCO
2
levels (Meuret et al. 2008; Salkovskis
et al. 1986). Han et al. (1996) reported a significant change
in the breathing pattern of patients with hyperventilation
and/or anxiety disorders following a similar breathing
retraining. Mean values of inspiration time (Ti), expiration
time (Te) and tidal volume (Vt) were found to be increased
(Vt/Ti decreased as a result), indicating that a prolonged
breath was compensated for by an increase in tidal volume.
A further analysis showed that the relief of the symptoms
correlated mainly with a slowing down of the breathing
rate.
Although breathing retraining is a popular and relatively
simple stress intervention, a theoretical concern is that the
therapeutic results may be achieved primarily by non-
respiratory mechanisms (Meuret et al. 2003). Hibbert and
Chan (1989) proposed that placebo mechanisms may
explain the beneficial effects of breathing retraining
(Garssen et al. 1992). However, recent evidence does not
support the idea that the reduction in panic symptoms in
patients receiving respiratory training is mediated by non-
specific factors, such as perceived control (Meuret et al.
2010).
Breathing techniques are also applied to treat cardio-
vascular complaints (Grossman et al. 2001; Pitzalis et al.
Appl Psychophysiol Biofeedback
123

1998). E.g., a fda-approved intervention in reducing high
blood pressure in hypertensives involves device-guided
breathing exercises (see http://www.resperate.com). With
differentiated inspiration and expiration ‘‘sounds’’ the user
is guided to lower the respiratory frequency to less than 10
breaths/min with prolonged expiration. Studies on the
efficacy of this treatment (Grossman et al. 2001; Logten-
berg et al. 2007; Rosenthal et al. 2001; Schein et al. 2009;
Viskoper et al. 2003) reported significant reductions in
systolic blood pressure. Van Dixhoorn (1998) instructed
myocardial infarction patients to actively regulate respira-
tion by making the passage of air audible for 5–6 breaths/
min. This was achieved by breathing through slightly
pursed lips and by employing slight movements of the
body that facilitated the respiration. Again, it was found
that the slower breathing pattern implemented in this
intervention was associated with a significant decrease in
heart rate and an increase in RSA (Lehrer et al. 1999; van
Dixhoorn 1998).
Thus, different breathing techniques are described in
literature and applied to cope with a variety of stress-
related conditions. The most common feature of these
breathing exercises is a reduction in respiratory rate, which
is assumed to increase parasympathetic activity. However,
despite the wide application and likely effectiveness of
such breathing techniques, some rather basic questions
remain unsolved.
First, in contrast with the cardiovascular effects, the
subjective effects of instructed slower breathing have been
described rather poorly. Whereas some studies apply some
sort of self-report on anxiety, avoidance, depression
intensity (Clark et al. 1985; Hibbert and Chan 1989) or on
panic, asthma or hypertension complaints, they lack a
measure specifically related to relaxation. Interestingly,
Smith (2001) has listed different affective states of relax-
ation and published inventories to asses 19 relaxation states
(R-States) that can be divided in four categories based on
factor analytic research (Smith et al. 2000): basic relaxa-
tion, mindfulness, positive energy transcendence and
stress. It is largely unknown how instructed breathing
patterns affect these R-States and categories.
Second, it is unclear to which extent features of the
breathing pattern other than respiratory rate contribute to
the effects of slow breathing. In particular, the ratio
between inspiration and expiration (i/e ratio) may be a
relevant breathing parameter to look at. Several studies
have documented that states of relaxation and stress induce
decreases and increases in i/e ratio’s, respectively (e.g.,
Boiten 1998; Gomez et al. 2004; Van Diest et al. 2001). In
addition, the literature suggests that HF-HRV and RSA are
greater when breathing with a low compared to a high i/e
ratio (Porges 2007; Strauss-Blasche et al. 2000). Although
the latter phenomenon does not necessarily reflect an
increased vagal tone (see Song and Lehrer 2003 for an
explanation in terms of a more complete hydrolization of
acetylcholine at the SA-node), findings from Cappo and
Holmes (1984) corroborate the idea that arousal is smaller
with low compared to high i/e ratio’s. In the latter study,
participants breathed at 6 breaths/min during three differ-
ent conditions: one condition with quick inhalation (2 s)
and slow exhalation (8 s), one with slow inhalation (8 s)
and quick exhalation (2 s) and a third condition with equal
inspiratory and expiratory times (5 s). The authors reported
that ‘‘fast/slow breathing’’ reduced electrodermal activity
and subjective arousal during anticipation of an electric
shock (Cappo and Holmes 1984). Together, the above
mentioned findings suggest that the arousal reducing effect
of slow breathing may be partly explained by the low i/e
ratio’s that typically accompany slow breathing patterns.
Therefore, the present study aimed to examine and
disentangle the effects of respiratory rate and i/e ratios on
self-reported relaxation states, self-reported affect, and on
RSA and HF-HRV. Participants were instructed to adopt a
slow or normal respiration rate, either with a high or a low
i/e ratio. Heart rate and respiration were continuously
measured and subjective dimensions of relaxation and
affect were assessed after each instructed breathing
exercise.
Method
Participants
Participants were recruited through the Experiment Man-
agement System (EMS) website of the Faculty of Psy-
chology and Educational Sciences at the University of
Leuven. The sample consisted of 30 undergraduate stu-
dents (age range 1–22 years; 2 men), who participated in
return for course credit. The sample size was based on that
of similar studies in the field (e.g., Song and Lehrer 2003).
Exclusion criteria were: any self-reported cardiac or
respiratory disorder, use of psychopharmaca or presence of
a psychiatric disorder in the past 2 years, pregnancy, and
practicing yoga/meditation/relaxation/mindfulness on a
regular basis. Five participants were excluded from ana-
lysis because they were positive on at least one criterion:
one participant with experience in relaxation and breathing
techniques, two participants taking selective serotonin re-
uptake inhibitors and/or beta-blockers and two participants
reporting asthma. Two other participants were excluded
from analysis because they did not comply with the
breathing instructions. As a result, N was 23 in our final
dataset, which was subjected to statistical analysis.
All participants provided their informed consent and the
study was approved by the local ethical committee.
Appl Psychophysiol Biofeedback
123

Stimuli
Four different breathing videos of 5 min each were pre-
sented to assist the participant in altering his or her
breathing pattern. Each displayed a vertical bar (4 cm x
19,5 cm) filling up while inspiration was expected and
emptying while expiration was expected, along with a tone
varying in pitch (with a high frequency of approximately
630 Hz for inspiration, and a low frequency of approxi-
mately 330 Hz for expiration).
The videos were presented on a Flat Panel Monitor
computer screen in front of the participant, who was
wearing headphones (Philips-SPH2500).
A first breathing video (‘‘1.5_3.5’’) reflected a breathing
rate of 12 breaths/min and an i/e ratio of 0.42 (an inhalation of
1.5 s to an exhalation of 3.5 s). A second video (‘‘3.5_1.5’’)
was created to display a breathing rate of 12 breaths/min,
with an i/e ratio of 2.33 (a 3.5 s inhalation to a 1.5 s exha-
lation). The third (‘‘3_7’’) consisted of a breathing rate of 6
breaths/min with a 3 s inhalation and a 7 s exhalation (i/e
ratio of 0.42). A fourth and final video (‘‘7_3’’) contained a
slow breathing rate of 6 breaths/min with an i/e ratio of 2.33
(a 7 s inhalation to a 3 s exhalation).
Measures
Heart Rate and Breathing Pattern
Respiratory and ECG signals were obtained with the
LifeShirt System
"
(Vivometrics Inc., Ventura, CA); a non-
invasive, ambulatory monitoring system. Data were sam-
pled at 200 Hz for the ECG and at 50 Hz for the respiratory
signals. Respiratory data were continuously collected by
means of respiratory inductive plethysmography (RIP).
Two shielded electrical wires that were sewn into an elastic
LifeShirt garment at the level of the rib cage and abdomen
served as RIP transducers. Tidal volume was defined as the
sum of the rib cage and the abdominal deflections (Konno
and Mead 1967). A Qualitative Diagnostic Calibration of
the Vivologic software was used. This procedure auto-
matically selects a 5 min recording of breaths that are
averaged to a default value of 400 ml. Thereafter, the
software calculates and applies a calibration factor, gen-
erating ml values for all breaths in the respiratory trace. We
did not calibrate the RIP signal with spirometry, so the
reported volumes should not be used as absolute quanti-
tative values. They merely reflect changes in breathing
patterns within an individual, which is appropriate and
sufficient given the within-subject design of the present
study.
Electrical activity of the heart was registered by means
of three Silvertrace ECG sensors, connected with the
LifeShirt system. The RIP transducers and ECG sensors
were connected with the LifeShirt recording system, which
contained a compact flash memory card on which data
were stored.
Self-reports
A short, custom-made health questionnaire was used to
assess the participants’ medical history and experience in
relaxation and breathing techniques.
Furthermore, the participants were asked to rate their
expected emotional state if they would maintain breathing
for a few minutes in accordance with the displayed breathing
video. Pleasantness, arousal and feeling of control were rated
on three custom-made scales, each ranging from -50 to
?50, with -50 referring to the highest unpleasantness, the
lowest arousal and lowest level of control.
Following the actual performance of adopting a partic-
ular breathing pattern for 5 min, participants rated their
emotional state on the affective dimensions of valence,
arousal and dominance using the pencil-and-paper-version
of the nine-point Self-Assessment Manikin (SAM) (Brad-
ley and Lang 1994).
They also completed a Dutch translation of the Smith
Relaxation States Survey (SRSI-3) (Smith 2001); a 38-item
relaxation questionnaire with a 6-point likert scale
(1 =not at all,6=the maximum) used to measure the
experiences people have when practicing different kinds of
relaxation. Because no Dutch version of the SRSI-3 is
available, we made and used an ad-hoc translation. The
data of the SRSI-3 were scored according to the inventory
manual and mean values of the different relaxation states
(R-states) subscales, i.e. basic relaxation, mindfulness,
positive energy, transcendence and stress were calculated.
Chronbach alpha reliabilities for the R-state scales (of
previous versions of the inventory; excluding three new
mindfulness items) range from .60 to .88 (Smith 2001).
In a custom-made post-experimental questionnaire, the
participants rated the extent to which they had complied with
the breathing instructions on a 7-point likert scale, ranging
from 1 =absolutely true to 7 =absolutely untrue. Also,
this post-experimental questionnaire asked them to write
down what they thought to be the hypotheses of the study.
Procedure
Phase 1: Baseline Measurements and Practice Period
Participants were tested individually in a single experi-
mental session that lasted approximately 1 h. After arrival,
the female experimenter (KV) provided a short outline of
the experimental procedure. The participants were told that
their heartbeat would be recorded during the induction of
four different breathing patterns, that they would receive
Appl Psychophysiol Biofeedback
123

rating scales and questionnaires to be filled out at different
times during the experiment and that they were to put on
the Life Shirt after the attachment of three electrodes (on
either side of the participant’s chest and right beneath the
left rib cage). The participants filled out an informed
consent and a short questionnaire referring to their health
background. Next, the electrodes were attached and they
put on the Life Shirt. Following this, the experimenter
asked them to sit quietly for a 7-min baseline assessment.
Afterwards, the participants practiced each of the four
breathing patterns (i.e., 1.5_3.5, followed by 3.5_1.5; 3_7;
and 7_3). Each pattern was practiced for approximately
45 s. The experimenter remained with the participant to
guide him or her throughout this practice session. The
participants were asked to try to follow the instructed
rhythm as closely as possible and were told to control their
breathing rate without forcing themselves to breathe in
deeply. After each short period of practicing a different
breathing pattern, the participants rated on a –50 to ?50
scale how they expected to feel in terms of pleasantness,
arousal and control if they would adopt the practiced
breathing pattern for a couple of minutes.
Phase 2: Breathing Instructions
Subsequently, in order to obtain a baseline measure of the
self-reports, the participants completed the SAM scales and
the SRSI-3. Then, the experimenter instructed the partici-
pants to adjust their breathing by following the instructions
of the breathing video that would be shown. Next, the
experimenter left the room and the breathing video was
started. After this 5 min breathing exercise, the SAM-
scales and the SRSI-3 were administered. This sequence
was repeated for each of the four breathing patterns. Par-
ticipants were counterbalanced across the 24 possible
presentation orders.
Phase 3
After the last breathing video, the electrodes were detached
and participants would change back into their own clothes.
Next, they filled out the short Social Desirability ques-
tionnaire and the post-experimental questionnaire. Fol-
lowing this, the participants received a written debriefing
and had the opportunity to leave their e-mail address to
receive more clarification on the findings of the study.
Data Analyses
Physiological Data Reduction
Respiration. Dedicated Vivologic software (Vivometric
Inc., Ventura, CA) was used to edit raw respiratory data.
Respiratory parameters—inspiration to expiration time (i/e
ratio =Ti/Te), respiration rate (RR), and tidal volume (Vt)
were calculated breath-by-breath and then averaged for
each period of interest (the baseline period and each of the
four instructed breathing periods).
ECG. The Vivologic software was also used to derive
the time between consecutive R-peaks (RR intervals) from
the raw ECG signal, and to detect and linearly interpolate
ectopic beats. Finally, the data were visually inspected for
additional artefacts, but none were found. Heart rate (HR),
RSA (difference between maximum and minimum cardiac
interbeat interval per breath) and power in the high (HF-
HRV) and the low frequency band (LF-HRV) of heart rate
variability were calculated and averaged for each of the
5-min breathing instructions and the baseline measurement.
To reduce skewness, HF-HRV and LF-HRV were log-
transformed before entering the statistical analyses.
Statistical Analyses
A two-way repeated measures ANOVA analysis was per-
formed using the SPSS General Linear Model procedure on
cardiovascular and respiratory parameters and on each of
the self-reports. ‘‘RR’’ (12 and 6) and ‘‘i/e ratio’’ (low and
high) served as two within-subject variables (each with two
levels). Bonferonni corrections were applied for multiple
follow-up comparisons of significant interactions. We will
also report partial e
ˆta-squared effect sizes.
Results
Respiratory Parameters
Table 1lists the mean values for the respiratory parameters.
The main effects of i/e ratio on Ti/Te (F(1, 22) =355.34,
p\.05, g
p
2
=.94, and on respiratory rate (F(1, 22) =
184.12, p\.05, g
p
2
=.89) indicated that participants
changed their breathing pattern according to the instructions.
Participants breathed with larger volumes when
breathing at 6 as compared to 12 breaths a minute (main
effect of RR on Vt (F(1, 22) =142.06, p\.05, g
p
2
=.87).
This effect interacted with the inhalation to exhalation ratio
of the breathing (i/e ratio x RR interaction: F(1, 22) =
4.59, p\.05, g
p
2
=.17). The difference in tidal volume
when adopting a respiration rate of 6 versus 12 was more
pronounced when breathing at a low i/e ratio, compared to
breathing at a high i/e ratio.
Self-reported Relaxation and Affect
Prior to the actual performance of the breathing exercises,
participants expected higher pleasantness and marginally
Appl Psychophysiol Biofeedback
123

significantly also less arousal when breathing at 6 com-
pared to 12 breaths a minute (main effects of RR on
valence, F(1, 22) =6.17, p\.05, g
p
2
=.22; on arousal,
F(1, 22) =4.03, p\.06, g
p
2
=.15). In addition, they
tended to anticipate higher feelings of control with a low
compared to a high i/e ratio (main effects of i/e ratio, F(1,
22) =4.17, p=.05, g
p
2
=.16).
Following the performance of each instructed breathing
pattern (see Table 2), participants reported higher pleas-
antness for the low compared to the high i/e ratio, F(1,
22) =7.41, p\.05, g
p
2
=.25, and with a respiration rate
of 6 compared to 12 breaths a minute, F(1, 22) =4.43,
p\.05, g
p
2
=.17. Also, they reported lower arousal with a
respiration rate of 6 compared to 12 breaths a minute, F(1,
22) =8.17, p\.05, g
p
2
=.27) and higher feelings of
control for the low compared to the high i/e ratio, F(1,
22) =6.89, p\.05, g
p
2
=.24.
Mean values on the relaxation dimensions are displayed
in Table 2. The participants scored higher on basic relax-
ation, mindfulness, positive energy and stress reduction
after adjustment to the breathing patterns with a low as
compared to a high i/e ratio (significant effect of i/e ratio
on basic relaxation (F(1, 22) =13.06, p\.05, g
p
2
=.37),
mindfulness (F(1, 22) =5.93, p\.05, g
p
2
=.22), positive
energy (F(1, 22) =5.77, p\.05, g
p
2
=.21) and stress
reduction (F(1, 22) =13.40, p\.05, g
p
2
=.38). Breath-
ing at 6 as compared to 12 breaths a minute resulted only in
higher self-reported levels of positive energy, (F(1,
22) =8.70, p\.05, g
p
2
=.28). The participants’ feelings
of transcendence (as sensing the deep mystery of things
Table 1 Means and (SDs) of cardiac and respiratory parameters during baseline and each of the 5-min breathing exercises
Baseline 12 breaths/min 6 breaths/min
Low i/e ratio High i/e ratio Low i/e ratio High i/e ratio
RR 14.88 (4.22) 12.27 (0.90) 12.48 (0.64) 7.32 (1.90) 7.69 (2.08)
Ti/Te 0.67 (0.11) 0.53 (0.07) 1.42 (0.19) 0.49 (0.06) 1.44 (0.26)
Vt 532 (216) 854 (325) 874 (269) 1,383 (424) 1,262 (372)
HR 79.11 (11.10) 75.73 (10.72) 74.66 (9.25) 77.90 (12.37) 73.64 (10.26)
RSA 94 (53) 137 (58) 131 (55) 275 (92) 234 (76)
HF-HRV 6.60 (0.82) 7.50 (0.91) 7.42 (0.81) 7.13 (0.82) 6.31 (0.77)
LF-HRV 6.66 (0.93) 6.28 (0.70) 6.37 (0.67) 8.93 (0.69) 9.08 (0.49)
RR respiration rate (in breaths/minute), Ti/Te ratio between duration of inhalation and exhalation (i/e ratio), Vt tidal volume (in ml), HR heart rate
(in beats/minute), RSA respiratory sinus arrhythmia (difference between maximum and minimum cardiac interbeat interval, in ms), HF-HRV log-
transformed power in the high frequency band of heart rate variability, LF-HRV log-transformed power in the low frequency band of heart rate
variability
Table 2 Means and SDs of valence, arousal, control and five dimensions of relaxation dimensions (SRSI-3 questionnaire) to baseline and each
of the 5-min breathing exercises
Baseline 12 breaths/min 6 breaths/min
Low i/e ratio High i/e ratio Low i/e ratio High i/e ratio
SAM
Arousal 3.43 (1.85) 3.09 (1.54) 4.17 (1.88) 2.52 (1.90) 3.52 (1.53)
CControl 5.52 (1.78) 5.91 (1.62) 4.87 (1.74) 5.96 (1.85) 4.91 (1.98)
Valence 6.91 (1.24) 5.96 (1.15) 5.22 (1.31) 6.52 (1.56) 5.39 (1.70)
SRSI-3
Basic relaxation 3.00 (0.80) 3.51 (0.98) 3.06 (0.98) 3.76 (0.91) 3.08 (0.70)
Mindfulness 3.22 (0.65) 2.93 (0.59) 2.76 (0.55) 3.09 (0.61) 2.89 (0.49)
Positive energy 2.48 (0.73) 2.49 (0.79) 2.34 (0.69) 2.66 (0.72) 2.48 (0.72)
Transcendence 2.01 (0.62) 2.01 (0.62) 2.01 (0.62) 2.01 (0.62) 2.01 (0.62)
Stress 1.63 (0.74) 1.58 (0.70) 2.12 (0.92) 1.55 (0.58) 1.92 (0.81)
Higher scores on arousal, control and valence indicate more arousal, more control and higher pleasantness, respectively; the Self-Assessment
Manikin (SAM Scale) is from Bradley and Lang (1994). The Smith Relaxation States Survey (SRSI-3) is from Smith (2001)
Appl Psychophysiol Biofeedback
123

and accepting this) did not differ according to the breathing
patterns.
Cardiovascular Parameters
The mean values of cardiovascular parameters are listed in
Table 1.
Heart Rate
A lower i/e ratio (as compared to a high i/e ratio) resulted
in a higher heart rate (significant main effect of i/e ratio on
HR (F(1, 22) =4.77, p\.05, g
p
2
=18). There was no
effect of RR on heart rate (F(1, 22) =0.44, p=.52).
RSA
In addition to the main effects of RR, F(1, 22) =211.14,
p\.05, g
p
2
=.91, and of i/e ratio, F(1, 22) =7.77,
p\.05, g
p
2
=.26, also the RR x i/e ratio interaction was
significant for RSA, F(1, 22) =7.24, p\.05, g
p
2
=.25.
As expected, RSA was higher when breathing at 6 com-
pared to 12 breaths a minute. Also, follow up comparisons
of the interaction showed that the low i/e ratio was asso-
ciated with a significantly higher RSA when breathing at 6
breaths a minute, F(1, 22) =9.30, p\.025, but not when
breathing at 12 breaths a minute, F(1, 22) =0.64, n.s.
HF-HRV
HF-HRV was significantly lower at 6 than at 12 breaths a
minute, F(1, 22) =39.64, p\.05, g
p
2
=.64, and with the
high compared to the low i/e ratio minute, F(1,
22) =25.46, p\.05, g
p
2
=.54). However, following up
on the RR x i/e ratio interaction, F(1, 22) =18.58,
p\.05, g
p
2
=.46, our findings showed that the low i/e
ratio was associated with a significantly higher HF-HRV
when breathing at 6 breaths a minute, F(1, 22) =36.12,
p\.025, but not when breathing at 12 breaths a minute,
F(1, 22) =0.59, n.s..
LF-HRV
LF-HRV was higher when breathing at 6 compared to 12
breaths a minute, F(1, 22) =862,48, p\.05, g
p
2
=.98).
No other effects were present for LF-HRV.
Discussion
Instructed breathing patterns are widely applied to treat a
variety of complaints and conditions. Most often, they
involve a voluntary slowing down of the breathing
frequency. Although voluntary reductions in respiration
rate typically induce a relative shift towards expiration, or,
a lower i/e ratio, the literature remains unclear about
whether the beneficial effects of adopting a slow breathing
pattern may be explained by a concomitant reduction in i/e
ratio. Moreover, the subjective effects of the instructed
breathing patterns have been described rather poorly.
Therefore, the present study aimed to investigate the
effects of altering both respiration rate and i/e ratio on self-
reported affect and relaxation, and on RSA and HF-HRV.
Generally, our findings show that i/e ratio is the more
important determinant for self-reported effects of relaxa-
tion as obtained by instructed breathing. Although partic-
ipants did not expect such effect prior to performing the
breathing exercises, they reported higher pleasantness and
more feelings of control for the breathing patterns with a
low compared to a high i/e ratio. In addition, participants
reported more relaxation, more positive energy, less stress,
and higher mindfulness when adopting a breathing pattern
with a low i/e ratio as compared to a high i/e ratio. In
contrast, effects of respiration rate were observed only for
positive energy. A possible explanation for the association
between i/e ratio and self-reported relaxation is that
inspiration is an active process requiring inspiratory muscle
activity, whereas expiration is a passive process charac-
terized by relaxation of the inspiratory muscles. An alter-
native explanation may be that the high i/e ratio applied in
the present study is a rather artificial breathing pattern that
is unlikely to occur during spontaneous breathing in natural
circumstances. Participants may have experienced more
difficulties with complying with the breathing patterns with
a high i/e ratio, resulting in more stress and lower relaxa-
tion. Future research may want to investigate whether
variations in a more natural range of i/e ratios would yield
similar effects to the ones presently observed.
In line with several other studies (e.g., Giardino et al.
2003; Song and Lehrer 2003), we observed a higher RSA
when the participants breathed at a lower as compared to a
higher respiratory rate. HF-HRV showed a reversed pattern
of results: it was greater when adopting a respiratory rate of
12 as compared to 6 breaths/min. This is due to the fact
that participants breathed at a rate outside the HF band
(0.15–0.40 Hz) when breathing at 6 breaths/min (Berntson
et al. 1997; Grossman and Kollai 1993). Our finding of an
increased HRV in the LF when breathing at 6 compared to
12 breaths/min is consistent with this and is likely due to
the fact that RSA at a respiratory rate of 6 breaths/min
resonates with the cardiac loop of the baroreflex (Vaschillo
et al. 2006).
Nonetheless, the fact that there is still a considerable
power in the HF-HRV when breathing at 6 breaths/min is
very intriguing. Visual inspection of each participant’s
power density spectrum of HRV (Fourier Analysis) in the
Appl Psychophysiol Biofeedback
123

6 breaths/min condition revealed clear harmonics of the
0.1 Hz peak in most participants. That is, there was not
only a clear peak at the (instructed) respiratory frequency
(0.1 Hz), but also a smaller one at 0.2 Hz and sometimes
also yet a smaller one at 0.4 Hz. Interestingly, the power
spectrum of the respiratory volume signal showed similar
harmonics, which may be due to the fact that the respira-
tory signal was not a true sinus signal and/or to participants
not following the imposed breathing perfectly. As a con-
sequence, the resulting tachogram is not perfectly sinu-
soidal either, and fourier analysis applied to such
tachogram may easily produce power spectra that consist
of the fundamental frequency component, or first har-
monic, together with its harmonics (Beauchamp 1973;
Ramirez 1985). As the present study did not include beat-
to-beat blood pressure recordings, it remains unknown
whether the harmonics in the power spectra of heart rate
variability and the respiratory signal co-occur with har-
monics in the baroreflex. Also, whether the harmonics in
heart rate variability reflect vagal activity or not, is unclear
and subject to further research.
The present findings confirm the relatively scarce
reports in the literature on an increase in RSA and HF-
HRV when breathing with a low i/e ratio (Porges 2007;
Strauss-Blasche et al. 2000). However, this effect of i/e
ratio on RSA and on HF-HRV interacted with respiration
rate in our study. A low compared to a high i/e ratio
resulted in a significantly higher HF-HRV when partici-
pants were breathing at 6, but not at 12 breaths/min. As of
now, we can only speculate on why a lower i/e ratio
yielded more power in the HF band when breathing at
6 breaths/min. One explanation may be that it is related to
the higher tidal volumes in the low compared to the high i/e
ratio condition (at 6 breaths/min, see Table 1).
The respiratory findings indicate that the participants
generally managed to change their inspiratory and expira-
tory times according to the instructed pacing. As can be
expected (e.g., Han et al. 1996), a stronger reduction in
breathing frequency (6 as compared to 12 breaths/min) was
compensated for by an increase in tidal volume. Although
we instructed persons not to breathe very deeply when
following the paced breathing patterns, minute ventilation
(respiration rate x tidal volume) during instructed breathing
was high compared to spontaneous breathing during
baseline. This is a common phenomenon, rather typical for
untrained persons. Our participants may have been hyper-
ventilating during instructed breathing, which may explain
the rather small effects observed on self-reported relaxation
and changes in affect for each of the breathing patterns. It
is recommended that future research takes special care to
avoid hyperventilation during instructed breathing, e.g., by
implementing a more rigorous training phase, and/or by
implementing ETCO2 biofeedback.
Contrary to what one would expect, heart rate was
increased when breathing at a low i/e ratio as compared to
a high i/e ratio. This finding is new and was not observed
by Cappo and Holmes (1984) in their study. Visual
inspection of the respiratory signals in our study indicated
that when breathing with a high i/e ratio, most participants
approached their inspiratory volume relatively early on in
the inspiratory cycle. As such, inspiratory flow returns to
zero relatively long before expiration is actually initiated,
producing breathing patterns with rather ‘flattened’
peaks—sometimes similar to what one would observe with
a post-inspiratory pause. The opposit pattern (sharp peaks
and a tendency to post-expiratory pauses) was typical for
the breathing patterns with a low i/e ratio. Therefore, the
former breathing pattern (high i/e ratio) may have been
associated with relatively longer episodes of high intra-
thoracic pressure, thereby increasing stroke volume and
mean arterial blood pressure, which in turn may decrease
heart rate via the baroreflex.
No significant differences in heart rate were found
between the two different respiration rates of 6 and
12 breaths/min; which is in accordance with the results of
Song and Lehrer (2003) and other studies (Grossman et al.
1991; Hayano et al. 1994; Pitzalis et al. 1998).
In summary, the present results strongly suggest that
voluntary changes in i/e ratio are an important determinant
of self-reported states of relaxation, and of RSA and power
in the HF-band when breathing at 6 breaths/min. Our
results suggest that beneficial effects of slow breathing
described in the literature may be primarily due to con-
comitant changes in i/e ratio. Inconsistencies in clinical or
physiological outcomes of slow breathing or HRV bio-
feedback likely result from a lack of control of i/e ratios.
Breathing retraining and respiratory biofeedback-treat-
ments may benefit from a more careful consideration of the
i/e ratios they apply.
Acknowledgments This research was supported by research grants
of FWO Flanders to I. Van Diest, and by IWT PhD grant to D.
Widjaja.
References
Beauchamp, K. G. (1973). Signal processing using analog and digital
techniques. London: Allen & Unwin.
Berntson, G. G., Bigger, J. T., Eckberg, D. L., Grossman, P., Kaufmann,
P. G., Malik, M., et al. (1997). Heart rate variability: Origins,
methods, and interpretive caveats. Psychophysiology, 34, 623–648.
Berntson, G. G., Cacciopo, J. T., & Quigley, K. S. (1993). Respiratory
sinus arrhythmia: Autonomic origins, physiological mechanisms,
and psychophysiological implications. Psychophysiology, 30,
183–196.
Boiten, F. A. (1998). The effects of emotional behaviour on
components of the respiratory cycle. Biological Psychology,
49, 29–51.
Appl Psychophysiol Biofeedback
123

Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The self-
assessment manikin and the semantic differential. Journal of
Behavior Therapy and Experimental Psychiatry, 25, 49–59.
Brown, R. P., & Gerbarg, P. L. (2005). Sudarshan Kriya yogic
breathing in the treatment of stress, anxiety, and depression: Part
I—Neurophysiologic model. The Journal of Alternative and
Complementary Medicine, 11(1), 189–201.
Cappo, B. M., & Holmes, D. S. (1984). The utility of prolonged
respiratory exhalation for reducing physiological and psycho-
logical arousal in non-threatening and threatening situations.
Journal of Psychosomatic Research, 28(4), 265–273.
Clark, D. M., Salkovskis, P. M., & Chalkley, A. J. (1985). Respiratory
control as a treatment for panic attacks. Journal of Behavior
Therapy and Experminental Psychiatry, 16, 23–30.
Eckberg, D. L. (2003). The human respiratory gate. Journal of
Physiology, 548(2), 339–352.
Garssen, B., de Ruiter, C., & van Dyck, R. (1992). Breathing
retraining: A rational placebo? Clinical Psychology Review, 12,
141–153.
Giardino, N. D., Glenny, R. W., Borson, S., & Chan, L. (2003).
Respiratory sinus arrhythmia is associated with efficiency of
pulmonary gas exchange in healthy humans. American Journal
of Physiology – Heart and Circulatory Physiology, 284,
1585–1591.
Gomez, P., Stahel, W. A., & Danuser, B. (2004). Respiratory
responses during affective picture viewing. Biological Psychol-
ogy, 67, 359–373.
Grossman, E., Grossman, A., Schein, M. H., Zimlichman, R., &
Gavish, B. (2001). Breathing-control lowers blood pressure.
Journal of Human Hypertension, 15, 263–269.
Grossman, P., Karemaker, J., & Wieling, W. (1991). Prediction of
tonic parasympathetic cardiac control using respiratory sinus
arrhythmia: The need for respiratory control. Psychophysiology,
28, 201–216.
Grossman, P., & Kollai, M. (1993). Respiratory sinus arrhythmia,
cardiac vagal tone, and respiration: Within- and between-
individual relations. Psychophysiology, 30(5), 486–495.
Grossman, P., & Taylor, E. W. (2007). Toward understanding
respiratory sinus arrhythmia: Relations to cardiac vagal tone,
evolution and bio-behavioral functions. Biological Psychology,
74, 263–285.
Han, J. N., Stegen, K., De Valck, C., Cle
´ment, J., & Van de
Woestijne, K. P. (1996). Influence of breathing therapy on
complaints, anxiety and breathing pattern in patients with
hyperventilation syndrome and anxiety disorders. Journal of
Psychosomatic Research, 41(5), 481–493.
Hayano, J., Mukai, S., Sakakibara, M., Okada, A., Takata, K., &
Fujinami, T. (1994). Effects of respiratory interval on vagal
modulation of heart rate. American Journal of Physiology, 267,
33–40.
Hibbert, G. A., & Chan, M. (1989). Respiratory control: Its
contribution to the treatment of panic attacks: A controlled
study. The British Journal of Psychiatry, 154, 232–236.
Konno, K., & Mead, J. (1967). Measurement of the separate volume
changes of rib cage and abdomen during breathing. Journal of
Applied Physiology, 22, 407–422.
Lehrer, P. M., Sasaki, Y., & Saito, Y. (1999). Zazen and cardiac
variability. Psychosomatic Medicine, 61, 812–821.
Lehrer, P. M., Vaschillo, E., & Vaschillo, B. (2000). Resonant
frequency biofeedback training to increase cardiac variability:
Rationale and manual for training. Applied Psychophysiology
and Biofeedback, 25(3), 177–191.
Lehrer, P. M., Vaschillo, E., Vaschillo, B., Lu, S.-E., Eckberg, D. L.,
Edelberg, R., et al. (2003). Heart rate variability biofeedback
increases baroreflex gain and peak expiratory flow. Psychoso-
matic Medicine, 65, 796–805.
Lehrer, P. M., Vaschillo, E., Vaschillo, B., Lu, S.-E., Scardella, A.,
Siddique, M., et al. (2004). Biofeedback treatment for asthma.
Chest, 126(2), 352–361.
Logtenberg, S. J., Kleefstra, N., Houweling, S. T., Groenier, K. H., &
Bilo, H. J. (2007). Effect of device-guided breathing exercises on
blood pressure in hypertensive patients with type 2 diabetes
mellitus: A randomized controlled trial. Journal of Hypertension,
25, 241–246.
Meuret, A. E., Rosenfield, D., Seidel, A., Bhaskara, L., & Hofmann,
S. G. (2010). Respiratory and cognitive mediators of treatment
change in panic disorder: Evidence for intervention specificity.
Journal of Consulting and Clinical Psychology, 78(5), 691–704.
Meuret, A. E., Wilhelm, F. H., Ritz, T., & Roth, W. T. (2003).
Breathing training for treating panic disorder. Useful interven-
tion or impediment? Behavior Modification, 27(5), 731–754.
Meuret, A. E., Wilhelm, F. H., Ritz, T., & Roth, W. T. (2008).
Feedback of end-tidal pCO
2
as a therapeutic approach for panic
disorder. Journal of Psychiatric Research, 42, 560–568.
Meuret, A. E., Wilhelm, F. H., & Roth, W. T. (2004). Respiratory
feedback for treating panic disorder. Journal of Clinical
Psychology, 60(2), 197–207.
Miles, W. R. (1964). Oxygen consumption during three yoga-type
breathing patterns. Journal of Applied Physiology, 19, 75–82.
Nolan, R. P., Floras, J. S., Harvey, P. J., Kamath, M. V., Picton, P. E.,
Chessex, C., et al. (2010). Behavioral neurocardiac training in
hypertension: A randomized, controlled trial. Hypertension,
55(4), 1033–1039.
Pitzalis, M. V., Mastropasqua, F., Massari, F., Passantino, A.,
Colombo, R., Mannarini, A., et al. (1998). Effect of respiratory
rate on the relationship between RR interval and systolic blood
pressure fluctuations: A frequency-dependent phenomenon.
Cardiovascular Research, 38, 332–339.
Porges, S. W. (2007). The polyvagal perspective. Biological Psy-
chology, 74, 116–143.
Ramirez, R. W. (1985). The FFT. Fundamentals and concepts.
Englewood Cliffs: Prentice-Hall.
Ritz, T., Tho
¨ns, M., & Dahme, B. (2001). Modulation of respiratory
sinus arrhythmia by respiration rate and volume: Stability across
posture and volume variations. Psychophysiology, 38, 858–862.
Rosenthal, T., Alter, A., Peleg, E., & Gavish, B. (2001). Device-
guided breathing exercises reduce blood pressure: Ambulatory
and home measurements. American Journal of Hypertension, 14,
74–76.
Sakakibara, M., & Hayano, J. (1996). Effect of slowed respiration on
cardiac parasympathetic response to threat. Psychosomatic
Medicine, 58, 32–37.
Salkovskis, P. M., Jones, D. R. O., & Clark, D. M. (1986). Respiratory
control in the treatment of panic attacks: Replication and
extension with concurrent measurement of behaviour and pCO
2
.
British Journal of Psychiatry, 148, 526–532.
Schein, M. H., Gavish, B., Baevsky, T., Kaufman, M., Levine, S.,
Nessing, A., et al. (2009). Treating hypertension in type II
diabetic patients with device-guided breathing: A randomized
controlled trial. Journal of Human Hypertension, 23, 325–331.
Smith, J. C. (2001). Advances in ABC relaxation: Application and
inventories. New York, NY: Springer Publishing Company.
Smith, J. C., Wedell, A. B., Kolotylo, C. J., Lewis, J. E., Byers, K. Y.,
& Segin, C. M. (2000). ABC relaxation theory and the factor
structure of relaxation states, recalled relaxation activities,
dispositions, and motivations. Psychological Reports, 86(3),
1201–1208.
Song, H.-S., & Lehrer, P. M. (2003). The effects of specific
respiratory rates on heart rate and heart rate variability. Applied
Psychophysiology and Biofeedback, 28(1), 13–23.
Strauss-Blasche, G., Moser, M., Voica, M., McLeod, D. R., Klammer,
N., & Marktl, W. (2000). Relative timing of inspiration and
Appl Psychophysiol Biofeedback
123

expiration affects respiratory sinus arrhythmia. Clinical and
Experimental Pharmacology and Physiology, 27(8), 601–606.
Thayer, J. F., A
˚hs, F., Fredrikson, M., Sollers, J. J., & Wager, T. D.
(2012). A meta-analysis of heart rate variability and neuroim-
aging studies: Implications for heart rate variability as a marker
of stress and health. Neuroscience and Biobehavioral Reviews,
36, 747–756.
Thayer, J. F., & Brosschot, J. F. (2005). Psychosomatics and
psychopathology: Looking up and down from the brain.
Psychoneuroendocrinology, 30, 1050–1058.
Thayer, J. F., & Lane, R. D. (2009). Claude Bernard and the heart–
brain connection: Further elaboration of a model of neurovis-
ceral integration. Neuroscience and Biobehavioral Reviews,
33(2), 81–88.
Van Diest, I., Winters, W., Devriese, S., Vercamst, E., Han, J. N., Van
de Woestijne, K. P., et al. (2001). Hyperventilation beyond fight/
flight: Respiratory responses during emotional imagery. Psycho-
physiology, 38, 961–968.
van Dixhoorn, J. (1998). Cardiorespiratory effects of breathing and
relaxation instruction in myocardial infarction patients. Biolog-
ical Psychology, 49(1–2), 123–135.
Vaschillo, E., Vaschillo, B., & Lehrer, P. M. (2006). Characteristics
of resonance in heart rate variability stimulated by biofeedback.
Applied Psychophysiology and Biofeedback, 31(2), 129–142.
Viskoper, R., Shapira, I., Priluck, R., Mindlin, R., Chornia, L., Laszt,
A., et al. (2003). Nonpharmacologic treatment of resistant
hypertensives by device-guided slow breathing exercises. Amer-
ican Journal of Hypertension, 16, 484–487.
Vlemincx, E., Van Diest, I., De Peuter, S., & Van den Bergh, O.
(2007). Relaxatie : Een ‘cold case’ hervat. Psychologie en
Gezondheid, 35(1), 17–26.
Wheat, A. L., & Larkin, K. T. (2010). Biofeedback of heart rate
variability and related physiology: A critical review. Applied
Psychophysiology and Biofeedback, 35(3), 229–242.
Appl Psychophysiol Biofeedback
123