Content uploaded by Paul Gilbert
Author content
All content in this area was uploaded by Paul Gilbert
Content may be subject to copyright.
132
Clinical Neuropsychiatry (2008) 5, 3,
© 2008 Giovanni Fioriti Editore s.r.l.
132-139
SUBMITTED APRIL 2008, ACCEPTED JUNE 2008
A PILOT EXPLORATION OF HEART RATE VARIABILITY AND SALIVARY CORTISOL
RESPONSES TO COMPASSION-FOCUSED IMAGERY
Helen Rockliff, Paul Gilbert, Kirsten McEwan, Stafford Lightman and David Glover
Abstract
This study measured heart-rate variability and cortisol to explore whether Compassion-Focused Imagery (CFI)
could stimulate a soothing affect system. We also explored individual differences (self-reported self-criticism, attachment
style and psychopathology) to CFI. Participants were given a relaxation, compassion-focused and control imagery
task. While some individuals showed an increase in heart rate variability during CFI, others had a decrease. There was
some indication that this was related to peoples self-reports of self-criticism, and attachment style. Those with an
increase in heart rate variability also showed a significant cortisol decrease. Hence, CFI can stimulate a soothing
affect system and attenuate hypothalamic-pituitary-adrenal axis activity in some individuals but those who are more
self-critical, with an insecure attachment style may require therapeutic interventions to benefit from CFI.
Key Words: Compassion Cortisol Depression Imagery Heart Rate Variability Self-Criticism
Declaration of Interest: None
Helen Rockliff, Paul Gilbert, Kirsten McEwan, Mental Health Research Unit, Kingsway Hospital, Derby, UK DE22 3LZ
Professor Stafford Lightman, Henry Wellcome Laboratories for Integrative Neurosciences & Endocrinology, Bristol, UK
David Glover, Manchester Royal Infirmary, Manchester, UK
Corresponding Author
Professor Paul Gilbert, Mental Health Research Unit, Kingsway Hospital, Derby, UK DE22 3LZ
Email: p.gilbert@derby.ac.uk Fax: 01332 624576
Self-criticism is a major vulnerability factor to
psychopathology (Gilbert & Procter 2006, Whelton &
Greenberg 2005, Zuroff et al. 2005). In addition, it can
undermine the success of traditional psychotherapies
such as Cognitive Behavioural therapy (Rector et al.
2000). Recent research has explored the use of
compassion-focused imagery (CFI) as part of a
therapeutic process to help people who are highly self-
critical (Gilbert 2000; Gilbert & Irons 2004, 2005;
Wheatley et al. 2007)
For thousands of years imagination has been used
to stimulate various physiological states (Frederick &
McNeal 1999, Leighton 2003). Sexual, anxious, excited
and calm feelings, along with self-confidence and
preparation for tasks, can be stimulated via the
imagination (Singer 2006). Mental imagery has also
been used as a therapeutic aid to help desensitisation to
aversive stimuli, coping with stress, and for the
promotion of positive states of mind (Arbuthnott et al.
2001, Hall et al. 2006, Holmes et al. 2007, Singer 2006).
Recent research and clinical work has focused on the
value of helping patients develop self-compassion via
various means including imagery (Gilbert 2007; Gilbert
& Irons 2004, Gilbert et al. 2006).
Compassion evolved with the attachment system,
such that signals of care, support and kindness, help to
calm and sooth distressed individuals (Bowlby 1969).
The soothing/calming that results from receiving
kindness and support from others has been linked to a
specific oxytocin-opiate affiliative type of positive
affect regulation system (Carter 1998, Depue &
Morrone-Strupinsky 2005, Wang 2005) and reduces
sensitivity in the amygdala, especially to socially
threatening stimuli (Kirsch et al. 2005). This can be
contrasted with an agentic positive affect regulation
system associated with drive and excitement (Depue
& Morrone-Strupinsky 2005). Imagery and fantasising
are believed to access these different systems much
as other forms of imagery stimulate other affect systems
(e.g., anxious or sexual).
Porgess Polyvagal theory (2003, 2007) details
how the evolution of the myelinated vagus nerve has
supported interpersonal approach behaviours that
enable social affiliations, caring and sharing. The
myelinated vagus nerve evolved with attachment and
the ability for infants to be calmed by parental caring
behaviours (Depue & Morrone-Strupinsky 2005). This
addition to the autonomic nervous system can inhibit
sympathetically driven threat-defensive behaviours
(e.g. fight/flight) and hypothalamic-pituitary adrenal
(HPA) axis activity, and promote a calm physiological
state, conducive to interpersonal approach and social
133
Heart Rate Variability and Salivary Cortisol Responses to Compassion-Focused Imagery
Clinical Neuropsychiatry (2008) 5, 3
affiliation. In general, the safer people feel, the more
open and flexible they can be in response to their
environment (Gilbert 1993). This is reflected in the
dynamic balancing of the sympathetic and para-
sympathetic nervous systems that give rise to the
variability in heart rate (Porges 2007). Hence, feeling
safe is linked to HRV, and higher HRV is linked to a
greater ability to self-sooth when stressed (Porges 2007).
In contrast, when individuals feel unsafe, they will
tend to rely on more threat focused and stereotyped
defensive behaviours (Dickerson & Kemeny 2004,
Gilbert 1993) characterised by a less flexible balance
of the sympathetic and parasympathetic nervous
systems, with lower tone in the myelinated vagus nerve
(Porges 2007). This relative inflexibility and unbalance
of the autonomic nervous system, associated with lower
measures of HRV, has been associated with both mental
and physical ill health (Appelhans & Luecken 2006;
Thayer & Lane 2007).
This study is a pilot exploration of both the impact
of baseline HRV measures on individual experiences
of CFI, and the acute effects of CFI on HRV and cortisol.
Imagery conditions were designed to stimulate two
different types of positive affect. One type focused on
being the recipient of compassion, while the control
imagery focused on anticipation of reward (making
ones ideal sandwich). Having two types of imagery
targeted to stimulate the two different positive affects
enables exploration of the specificity effects of positi-
ve affect on HRV.
We also explored the influence of self-reported
self-criticism, self-compassion, attachment style, ease
of feeling socially safe and stress, anxiety and
depression on HRV responses to the CFI. We
hypethesised whether self-criticism was associated with
people finding CFI difficult and if this would be
reflected in HRV and cortisol measures.
Method
Participants
Participants were recruited from the University of
Derby (n =184) and completed a self-report screening
questionnaire. Inclusion criteria were: aged between 18
and 35 years; BMI between 19 and 28; non-smoker;
not-currently using medication (except contraception)
or illicit drugs; consuming under 30 units alcohol per
week; not working night shifts; no history of
cardiovascular problems; and no major mental health
problems (47 met these criteria). The final analyses
consisted of 22 participants, exclusions included: A
heart arrhythmia, being unable to sit still and concen-
trate, falling asleep, and imagining pity rather than
compassion.
Procedures
All experimental sessions commenced at 14:00 (to
allow consistent recording of cortisol levels) in the
psychology department at the University of Derby.
Participants were welcomed by the researcher and
seated in a relaxed but upright position with arm and
head supports for the duration of the study. Figure 1
outlines the procedure.
Imagery
Imagery was verbally guided by the researcher.
Each condition (relaxation/baseline, compassion and
control) lasted five and a half minutes. This was to
ensure that 300 seconds of artefact-free ECG data was
available for calculation of HRV metrics (Task Force
1996). During relaxation participants were asked to
allow the tension to drain from each of their muscles
in turn. Previous studies have used a variety of different
instructions for obtaining baseline measures, including
continuous counting, instructions to relax (Osumi et
al. 2006) and no baseline but inclusion of a control
condition (Takahashi et al. 2005).
CFI
The CFI imagery asked participants to imagine
compassion for them coming from an external source.
The researcher suggested that some people find it
Figure 1. Procedure
Continuous ECG recorded
PLQXWHV
5HOD[DWLRQ
,PDJHU\
PLQXWHV
([SODLQ
FRPSDVVLRQ
,PDJHU \
PLQXWHV
&RPSDVVLRQ
,PDJHU\
PLQXWHV
&RPSDVVLRQ
,PDJHU\
PLQXWHV
&RQWURO
,PDJHU\
PLQXWHV
&RQVHQW6HOI
UHSRUWVFDOHV
6DPSOH
&RUWLVRO$
PLQXWHV
$IWHU
PHDVXUHV
PLQXWHV
,QWURGXFH
([SODLQ
SURFHGXUH
6DPSOH
&RUWLVRO%
'HEULHILQJ
134
Helen Rockliff et al.
Clinical Neuropsychiatry (2008) 5, 3
helpful to create a mental picture of this compassionate
other (human or non-human). It was stressed that the
image was to help them feel they were the recipient of
compassion, and it was feeling that compassion that
the research was exploring.
Participants were verbally prompted every 60
seconds by the researcher with various statements such
as: Allow yourself to feel that you are the recipient of
great compassion; allow yourself to feel the loving-
kindness that is there for you.
Control Imagery
As noted, the positive affect system underpinning
appetitive and resource acquiring activity is different
to the soothing/contentment system (Depue & Morrone-
Strupinsky 2005). Therefore to compare different types
of positive affect inducing imagery, our control imagery
focused on preparing ones favourite sandwich.
Participants were prompted every 60 seconds with
statements like Imagine entering the shop of your
choice and taking some time to browse and look around
at the selection of food before you and Imagine
preparing your bread and beginning to build your sand-
wich. The full script is available on request.
Measures
Laboratory Measures
A three lead ECG was recorded using Ag-AgCl
disposable electrodes with Biopac PRO Lab software
version 3.6.7 and MP30 Hardware (Linton Instru-
mentation). Data was sampled at 1Khz to ensure
millisecond accuracy of the Inter-Beat Intervals (IBIs).
IBIs were determined off line using the Biopac R-wave
detection algorithm for 3 x 300 second periods
corresponding to baseline and each imagery condition.
All IBIs were manually checked by a researcher to
ensure their accuracy, and to correct missed and ectopic
beats due to the reported effects from such errors
(Berntson et al. 2005). Artefacts were corrected
according to the Task Force (1996) recommendations.
Heart Rate Variability Metrics
Using the artefact corrected ECG data, inter-beat
intervals were measured and from these heart rate was
calculated. CMetx software (Allen et al. 2007) was used
to calculate the SDNN, the standard deviation of inter-
beat intervals, a measure recommended by the Task
Force (1996). This is a global measure of HRV,
reflecting the sum of all of cyclic components
contributing to the variation of inter-beat intervals. This
was chosen as an appropriate measure for pilot work,
which would give good indication as to the likely merit
of further exploration using other measures of HRV.
To obtain measures of HRV reactivity to CFI the
raw HRV metrics corresponding to each condition
(baseline, compassion and control) were used to
calculate participants change in HRV values. These
change values were calculated by subtracting each
participants control imagery HRV value from their CFI
HRV value. (The control imagery HRV value was used
as it attempts to keep constant the possible effects of
mental effort and visualisation on HRV). Since HRV
shows less acute reactivity in individuals with low HRV
(Porges 2007), we adjusted the raw HRV change values
in proportion to the participants baseline value by
expressing HRV change values as a percentage of the
baseline HRV value. These variables, which we will
refer to as HRV change, make the magnitude of HRV
response to imagery comparable between individuals.
Cortisol
Saliva samples were collected using salivettes
(Sarstedt Ltd). The final cortisol sample was optimally
timed to allow for a delay in possible HPA axis response
to CFI (Dickerson & Kemeny 2004). All samples were
frozen before being assayed for cortisol by Obsidian
Research Ltd (Port Talbot, UK) using an Enzyme-
Linked ImmunoSorbent Assay (ELISA).
Self-Report Measures
Participants completed a demographics form, the
following self-report scales and additional questions.
Forms of Self-Criticism/Self-Reassuring Sca-
le (FSCRS) (Gilbert et al. 2004).
This 22-item scale assesses participants thoughts
and feelings about themselves during a perceived
failure. Two subscales measure forms of self-criticising,
(inadequate self and hated self) and one subscale
measures tendencies to be reassuring to the self
(reassured Self). Participants respond on a Likert scale
0 4. The scale has good reliability with Cronbachs
alphas of .90 for inadequate self, .86 for hated self, and
.86 for reassured self (Gilbert et al. 2004).
Self-Compassion Scale (SCS) (Neff 2003)
This 26-item scale assesses levels of self-
compassion. There are three factors of positive self-
compassion: self-kindness, common humanity and
mindfulness, and three factors that focused on a lack of
self-compassion: self-judgement, isolation and over-
identification. Participants indicate how often they
engage in these ways of self-relating on a Likert scale
1 5. The scale has good reliability (Cronbachs alphas
ranging from .75 to .81).
Adult Attachment Scale (Collins & Read 1990)
This 18-item scale measures three attachment
dimensions. Depend measures abilities to depend on
others, anxious measures degree of worry about
abandonment and close measures ease of getting close
to others. Respondents are asked to rate on a Likert
scale 1 5 how characteristic each statement is of them.
135
Heart Rate Variability and Salivary Cortisol Responses to Compassion-Focused Imagery
Clinical Neuropsychiatry (2008) 5, 3
The Cronbachs alphas were 0.75 for depend, 0.72 for
anxiety and 0.69 for close.
Social Safeness Scale (Gilbert et al. submitted)
This 11-item scale was developed to measure the
extent to which people experience their social world as
safe, warm and soothing, and their ability to enjoy
feelings of closeness with others. Each item is scored
on a Likert scale 0 4. This scale had a Cronbachs
alpha of .82.
Depression, Anxiety and Stress Scale (DASS-
21) (Lovibond & Lovibond 1995)
This 21-item shortened version of the DASS-42
consists of three subscales measuring depression,
anxiety and stress. Participants are asked to rate how
much each statement applied to them over the past
week, on a Likert scale 0 4. The DASS-21 subscales
have Cronbachs alphas of .94 for Depression, .87 for
Anxiety and .91 for Stress (Antony et al. 1998).
Results
Analysis was conducted using SPSS version 14.
The data were screened for normality of distribution
and for outliers.
CFI. Some people showed a reduction in HRV
(measured by SDNN) in response to compassion, while
others showed an increase. We felt these were important
differences and therefore split participants into two
groups. Group 1 contains 11 individuals who showed a
reduction in SDNN value from control to CFI (called
Group 1SDNNdowns), while Group 2 contains 11
individuals who showed an increase in SDNN value
during the CFI (called Group 2SDNNups). This equal
number arose by chance.
Means and standard deviations for each
condition and group are reported in Table 1. In addition,
the mean SDNN for each group during each condition
are depicted in Diagram 1.
The SDNN was analysed using repeated measures
ANOVAs to explore the effects of imagery condition
(relaxation, control and compassion) and group (i.e.
whether participants decreased or increased in SDNN
in response to CFI). The SDNN value calculated for
each imagery condition was used as within-subjects
variable, and group was used as a between-subjects
variable. Post hoc t-tests were performed to explore
where any significant differences lay between different
imagery tasks.
The ANOVA showed both a significant main effect
of imagery condition (F(1.15, 23.00) =4.41, p=.042)
and interaction effects with group (F(1.15, 23.00) =5.18,
p=.028). The difference in mean SDNN value between
Group 1SDNNdowns and Group 2SDNNups is apparent at
relaxation and is accentuated during the CFI. However,
the groups are very similar in SDNN during the control
imagery. The implications of this are that for some po-
sitive imagery tasks (e.g. making a sandwich), there
are no discriminatory effects on HRV. However, for
positive imagery associated with affiliative
interpersonal affects (warmth, kindness and social
connectedness) certain individuals responded with a
decrease in HRV whilst others had an increase in HRV.
The post hoc t-test revealed a significant difference
between groups during the CFI (t(20)=-2.20 p=.043).
This is visually represented in Diagram 1.
Cortisol
Cortisol values after relaxation and before CFI
(Cortisol A), and after CFI (Cortisol B) were analysed
using a repeated measures ANOVA, with group as a
between-subjects variable. There was a significant main
effect of imagery on cortisol value (F = 4.54, (df =1,
19) p = .046) but no significant interaction of group.
Post-hoc t-tests revealed a non-significant difference
between the two groups cortisol A values (Group
1SDNNdowns = 5.80ng/ml, Group 2SDNNups = 5.40ng/ml).
Both groups showed a mean cortisol value decrease
after CFI (Group 1SDNNdowns = 5.70ng/ml; Group 2SDNNups
= 4.92ng/ml). However, for Group 2SDNNups this decrease
in cortisol was significant, while for Group 1SDNNdowns
the decrease was non-significant. This greater decrease
in Group 2SDNNups resulted in a significant difference
between the group means for cortisol B (t=2.50, p=
.022) and can be seen clearly in Diagram 2. Cortisol A
correlated negatively with change in cortisol meaning
that the lower a participants initial cortisol value was,
Table 1. Means and standard deviations of SDNN across
conditions for both groups
Condition SDNN
Group1M(SD) Group2M(SD)
Baseline (Relaxation) 55.28 (17.74) 64.06 (23.35)
Compassion imagery 44.61 (17.09) 60.05 (16.44)
Control imagery 53.85 (15.25) 51.51 (18.87)
Participants reported boredom and fatigue
during the second CFI (C2) and therefore we removed
this data as it was unreliable.
Heart rate changes and HRV changes are related
but distinctly different measures. We therefore
performed an ANOVA to determine whether there were
any differences in heart rate between the three
experimental conditions, or between groups. There were
no significant differences, indicating that any changes
in HRV found were not merely a product of changes in
heart rate. Physical demands were kept constant
throughout the study to ensure that any changes in the
measures taken could be attributed to thought induced
alterations of metabolic demand.
Heart Rate Variability
The HRV data showed two opposing responses to
136
Helen Rockliff et al.
Clinical Neuropsychiatry (2008) 5, 3
&RQGLWLRQ
&RQWURO,PDJHU\&RPSDVVLRQ,PDJHU\%DVHOLQH
0HDQ6'11
*US6'11XSV
*US6'11GRZQV
6'11FKDQJH
JURXSLQJYDULDEOH
'LDJUDPVKRZLQJPHDQ6'11IRUHDFKJURXSDWEDVHOLQHGXULQJLPDJHU\FRQGLWLRQV
Diagram 1
&RUW%&RUW$
6DPSOH
0HDQFRUWLVROQJPO
*USXSV
*USGRZQV
*URXSLQJYDULDELOHIRU
6'11FKDQJH
'LDJUDPVKRZLQJPHDQFRUWLVROIRUHDFKJURXSDWEDVHOLQHDIWHU,PDJHU\
Diagram 2
137
Heart Rate Variability and Salivary Cortisol Responses to Compassion-Focused Imagery
Clinical Neuropsychiatry (2008) 5, 3
the bigger their decrease in cortisol during compassion
also was. These results show that for people who had
an increase in HRV during CFI there was a significant
decrease in cortisol. While for those who had a decrease
in HRV during CFI, there was a minimal non-significant
decrease, as would be expected from diurnal rhythm
alone.
Mauchlys test indicated that the assumption of
sphericity was violated for SDNN. Therefore,
Greenhouse-Geisser estimates of sphericity were used
to correct the degrees of freedom. Calculations of effect
size (indicated by partial eta squared) revealed that
effect sizes associated with cortisol measures were
higher (i.e. ω2>.08). than for SDNN (ω2= <.08).
Self-Report Scales
Data were screened for normality of distributions,
skewness values ranged from 0.09 to 1.43 and Kurtosis
values from -0.23 to 1.42. The hated self subscale
derived from the forms of self-criticism and
reassurance scale was skewed and kurtotic. This is
not surprising to find in a non-clinical population due
to a floor effect, and we chose to remove this subscale
from further analyses. T-tests were conducted to
examine differences in self-report scores between
groups 1SDNNdowns and 2 SDNNups. A significant difference
was found between social safeness scores (t=0.03),
with Group 1SDNNdowns having a lower mean score
(38.00) than Group 2SDNNups (44.64). We also noted that
the group differences on the other self-report variables
are all in the predicted directions, however statistical
significance was not reached, possibly a limitation of
the small numbers.
Correlation Analysis
Pearsons correlation coefficients for self-report
scales, cortisol measures, baseline HRV and HRV
change data are given in Table 2.
Change in SDNN was positively correlated with
ability to depend on others (adult attachment scale),
and with the social safeness and pleasure scale (r =
.52, .014 and r = .57, p = .006 respectively). In other
words, HRV response to CFI is linked to peoples
current experiences of themselves in social
relationships. This is further indicated with the
contrasting result that anxious attachment was
negatively correlated with change in SDNN (r = -.48,
p = .025). We also found self-criticism that focuses on
feelings of self-inadequacy was negatively associated
with reduced HRV when trying to engage in CFI (r=-.54).
In addition, although not significant, the ability to be
self-reassuring (r=.36) and self-compassionate (r=.33)
were positively correlated with SDNN change.
Discussion
This study aimed to explore the impact of CFI on
two different physiological measures (HRV and
cortisol) we also explored individual differences
informed by clinical observations that some people find
CFI difficult or threatening (Gilbert 2007). Our data
indicates two key processes. First, CFI does impact on
HRV. Some people show a clear increase in HRV,
whereas others show a more threat-like response with
a reduction in HRV. We also found that those individuals
who showed elevated HRV to compassionate imagery
experienced a drop in cortisol, indicative of a soothing
effect on the HPA axis. In contrast, those who showed
a reduction in HRV experienced a non-significant
cortisol change. There is research suggesting that some
individuals with secure attachments, typically engage
in soothing images and memories of feeling cared for
and these can aid emotional regulation (Baldwin 2005,
Mikulincer & Shaver 2007). Buddhism also specifically
focuses on training people in CFI (Leighton 2003).
It appears that different people have different
physiological responses to CFI. So the question is to
try to identify the sources of these differences. Why do
some people benefit from CFI and experience it as
soothing while others appear threatened? We explored
this with self-report measures. The self-report data
suggests that peoples experiences of CFI are related
to current experiences of social safeness; Group 2SDNNups
had higher mean scores of social safeness. Although
other self-report measures did not show significant
differences between the two groups, all differences
showed trends in the expected directions (e.g. Group
1SDNNdowns had higher mean scores of self-criticism, self-
coldness, anxious attachment and psychopathologies;
whilst Group 2SDNNups had higher mean scores of self-
compassion, self-reassurance and ability to depend on
6HOIUHSRUWPHDVXUH ,QDGHTXDWH
VHOI
5HDVVXUHG
VHOI
6HOI
&RPSDVVLRQ
6HOI
&ROGQHVV
'HSHQG
$WWDFKPHQW
$Q[LRXV
$WWDFKPHQW
&ORVH
$WWDFKPHQW
6RFLDO
6DIHQHVV
'HSUHVVLRQ
'$66
$Q[LHW\
'$66
6WUHVV
'$66
6'11 U
&KDQJH S
S
S
Table 2. Correlations between change in HRV measures and self-report scales
Note for readers interested in the results of other HRV metrics: RMSSD change also showed a correlation with the Depend
attachment subscale (r = .44, p =.043), Toichis Cardiac Vagal Index change also showed positive correlations with the Depend
attachment subscale (r = .45, p =.034), Toichis Cardiac Sympathetic Index also showed a significant correlation with the social
safeness and pleasure scale (r = .45, p =.035)
138
Helen Rockliff et al.
Clinical Neuropsychiatry (2008) 5, 3
others and experience close relationships). This would
fit with Gilberts (1989) concepts of a safeness system
and Porges (2003, 2007) view that higher HRV is linked
to interpersonal approach and socially adaptive
behaviours.
Further evidence for this link was revealed in the
pattern of correlations. A significant positive correlation
was found between change in SDNN during CFI and
social safeness and ability to depend on others. In
contrast, anxious attachment was negatively associated
with SDNN change; in other words, anxious attachers
may find CFI more threatening. The data raises
important implications for psychotherapy (Gilbert &
Irons 2005; Gilbert & Procter 2006). The differences
in HRV response seen between participants reflect
clinical observations; that for some people (particularly
self-critics and those scoring low in social safeness),
focusing on compassion can at first be unfamiliar,
threatening and feel unsafe (Gilbert 2007; Gilbert &
Irons 2005; Gilbert et al. 2006). Self-critics often report
feeling reluctant to let go of their self-criticism for
fear of their standards slipping; that they might
become selfish or arrogant, or that it constitutes a
change to self-identity. They can also fear compassion
because they feel they do not deserve compassion or
because it is unfamiliar, triggers sadness, or it is
frightening to let others (even imagined ones) get close
(Gilbert & Procter 2006).
It is unclear which aspects of compassion are
particularly threatening because our CFI involved
experiencing acceptance, loving kindness, warmth, and
compassion. Clinically, people often experience sadness
and grief when their attachment systems are activated
(Bowlby 1969, 1980). Compassionate imagery involves
activating these systems and drawing on emotional
memories of attachment (Gilbert 2007). The finding
that self-criticism was linked to an SDNN decrease
when engaging in CFI provides further support that self-
critics may find CFI difficult at first, this could also be
related to a lack of compassionate memories on which
to draw. This difficulty, and/or negative response to
feeling compassion for the self, can be a barrier to the
development of self-compassion, and could also be a
cause of the HRV decrease seen in Group 1SDNNdowns.
This suggestion fits with data by Segerstrom and
Solberg Nes (2007) who found HRV decreases to be
associated with self-regulatory processes, such as
inhibition of emotion.
A number of authors have noted that self-criticism
is a major vulnerability factor for low mood and
psychopathologies (Whelton & Greenburg 2005). Both
criticism from others (Dickerson & Kemeny 2004) and
self-criticism is linked to HPA axis arousal and cortisol
release (Mason et al. 2001). Gilbert (2007) suggests
that chronic self-criticism continually stimulates the
threat system thus having a detrimental impact on po-
sitive affect.
We concentrated on reporting the SDNN as
recommended by the global HRV metric Task Force,
(1996). However, there are other methods for exploring
the periodic processes involved in HRV, which may
shed more light on the relationships indicated by this
study. Second, we did not measure respiration or muscle
tension, and consequently the HRV changes could be
attributable to changes in breathing rate/depth, or
muscle tension. However, physical demands were kept
constant throughout the study making this unlikely.
Yet to be explored is whether training and working
through the fears and blocks to self-compassion, will
impact on HRV and other neurophysiological processes
linked to social soothing, and if this methodology could
be adapted for evaluating psychotherapies. Given the
increasing interest in compassion as a therapeutic aid,
further research into the neurophysiological mediators
and effects of compassion, may indicate ways of
developing psychotherapeutic techniques (Gilbert &
Irons 2005).
Acknowledgments
The authors would like to thank Linton
Instrumentation (Diss, Suffolk) for their facilitative
attitude and help with using Biopac software. This work
was undertaken by the Mental Health Research Unit who
received a proportion of its funding from the Department
of Health. The views expressed in this publication are those
of the authors and not necessarily those of the Department
of Health.
References
Appelhans BM & Luecken LJ (2006). Heart rate variability as
an index of regulated emotional responding. Review of
General Psychology 10, 229-240.
Allen JJB, Chambers AS & Towers DN (2007). The many metrics
of cardiac chronotropy: A pragmatic primer and brief
comparison of metrics. Biological Psychology 74, 243-262.
Antony MM, Bieling PJ, Cox BJ, Enns MW & Swinson RE
(1998). Psychometric properties of the 42-item and 21-
item versions of the depression anxiety stress scales in
clinical groups and a community sample. Psychological
Assessment 10, 176-181.
Arbuthnott KD, Arbuthnott DW & Rossiter L (2001). Guided
imagery and memory: Implications for psychotherapies.
Journal of Counselling Psychology 48, 123-132.
Baldwin MW (Ed) (2005). Interpersonal Cognition. Guilford,
New York.
Berntson GG, Lozano DL & Chen YJ (2005). Filter properties
of root mean square successive difference (RMSSD) for
heart rate. Psychophysiology 42, 246-252.
Bowlby J (1969). Attachment and Loss: Attachment, vol. 3. Basic
Books, New York.
Bowlby J (1980). Loss: Sadness and Depression. Attachment
and Loss, Vol. 3. Hogarth Press, London.
Carter CS (1998). Neuroendocrine perspectives on social
attachment and love. Psychoneuroendorinlogy 23, 779-
818.
Collins NL & Read SJ (1990). Adult attachment, working models
and relationship quality in dating couples. Journal of
Personality and Social Psychology 58, 644-663.
Depue RA & Morrone-Strupinsky JV (2005). A neurobehavioral
model of affiliative bonding: Implications for conceptua-
lizing a human trait of affiliation. Behavioural and Brain
Sciences 28, 313-350.
Dickerson SS & Kemeny ME (2004). Acute stressors and cortisol
response: A theoretical integration and synthesis of
laboratory research. Psychological Bulletin 130, 335-39.
Frederick C & McNeal S (1999). Inner Strengths: Contemporary
Psychotherapy and Hypnosis for Ego Strengthening.
Lawrence Erlbaum Associates, Mahwah, NJ.
139
Heart Rate Variability and Salivary Cortisol Responses to Compassion-Focused Imagery
Clinical Neuropsychiatry (2008) 5, 3
Gilbert P (1989). Human Nature and Suffering. Lawrence
Erlbaum Associates, London.
Gilbert P (1993). Defence and safety: Their function in social
behaviour and psychopathology. British Journal of Clinical
Psychology 32, 131-153.
Gilbert P (2000). Social mentalities: Internal social conflicts
and the role of inner wamth and compassion in cognitive
therapy In Gilbert P, Baily KG (Eds). Genes on the Couch:
Explorations in Evolutionary Psychotherapy. Brunner-
Routledge, Hove, pp.118-50.
Gilbert P (2007). Psychotherapy and Counseling for Depression.
Sage, London.
Gilbert P & Irons C (2004). A pilot exploration of the use of
compassionate images in a group of self-critical people.
Memory 12, 507-516.
Gilbert P & Irons C (2005). Focused therapies and compassionate
mind training for shame and self-attacking In Gilbert P
(Ed) Compassion: Conceptualisations, Research and Use
in Psychotherapy. Routledge, London, pp. 263-325.
Gilbert P & Procter S (2006). Compassionate mind training for
people with high shame and self criticism: Pilot study of a
group therapy approach. Clinical Psychology and
Psychotherapy 13, 353-379.
Gilbert P, Clarke M, Hempel S, Miles JNV & Irons C (2004).
Criticizing and reassuring oneself: An exploration of forms
style and reasons in female students. British Journal of
Clinical Psychology 43, 31-50.
Gilbert P, Baldwin M, Irons C, Baccus J & Palmer M (2006).
Self-criticism and self-warmth: An imagery study exploring
their relation to depression. Journal of Cognitive
Psychotherapy: An International Quarterly 20, 183-200.
Gilbert P, McEwan K, Mitra R, Mills A, Bellew R, Irons C, Gale
C & Legg L (submitted). Safeness and two types of positi-
ve affect. Cognition and Emotion.
Hall E, Hall C, Stradling P & Young D (2006). Guided Imagery:
Creative Interventions in Counselling and Psychotherapy.
Sage, London.
Holmes E, Arntz A & Smucker M (2007). Imagery rescripting
in cognitive behaviour therapy: Images, treatment
techniques and outcomes. Journal of Behavior Therapy
and Experimental Psychiatry 38, 297-305.
Kirsch P, Esslinger C, Chen Q et al. (2005). Oxytocin modulates
neural circuitry for social cognition and fear in humans.
The Journal of Neuroscience 25, 11489-11493.
Lovibond SH & Lovibond PF (1995). Manual for the Depression
Anxiety Stress Scales, 2nd Ed. Sidney Psychology
Foundation.
Leighton TD (2003). Faces of Compassion: Classic Bodhisattva
Archetypes and their Modern Expression. Boston, Wisdom
Publications.
Mason JW, Wang S, Yehuda R, Riney S, Charney DS &
Southwick SM (2001). Psychogenic lowering of urinary
cortisol levels linked to increased emotional numbing and
a shame-depressive syndrome in combat-related
posttraumatic stress disorder. Psychosomatic Medicine 63,
387-401.
Mikulincer M & Shaver PR (2007). Attachment in Adulthood:
Structure, Dynamics, and Change. Guilford, New York.
Neff K (2003). Self-compassion: An alternative conceptua-
lization of a healthy attitude toward oneself. Self and
Identity 2, 85-102.
Osumi T, Shimazaki H, Imai A, Sugiura Y, Ohira H (2006).
Psychopathic traits and cardiovascular responses to
emotional stimuli. Personality and Individual Differences
42, 1391-1402.
Porges S (2003). The Polyvagal theory: phylogenetic
contributions to social behaviour. Physiology & Behavior
79, 503-513.
Porges SW (2007). The polyvagal perspective. Biological
Psychology 74, 116-143.
Rector NA, Bagby RM, Segal ZV, Joffe RT & Levitt A (2000).
Self-criticism and dependency in depressed patients treated
with cognitive therapy or pharmacotherapy. Cognitive
Therapy & Research 24, 571-584.
Segerstrom SC & Solberg Nes L (2007). Heart rate variability
reflects self- regulatory strength, effort, and fatigue.
Psychological Science 18, 275-281.
Singer JL (2006). Imagery in Psychotherapy. American
Psychological Press, Washington DC.
Takahashi T, Murata T, Hamada T, Omori M, Kosaka H, Kikuchi
M, Yoshida H & Wada H (2005). Changes in EEG and
autonomic nervous activity during meditation and their
association with personality traits. International Journal
of Psychophysiology 55, 199-207.
Task Force of the European Society of Cardiology and the North
American Society of Pacing and Electrophysiology (1996).
Heart rate variability: Standards of measurement,
physiological interpretation, and clinical use. Circulation
93, 1043-1065.
Thayer J & Lane RD (2007). The role of vagal function in the
risk for cardiovascular disease and mortality. Biological
Psychology 74, 224-242.
Wang S (2005). A conceptual framework for integrating research
related to the physiology of compassion and the wisdom
of Buddhist teachings. In Gilbert P (Ed) Compassion:
Conceptualisations, Research and Use in Psychotherapy.
London, Routledge, pp. 75-120.
Wheatley J, Brewin CR, Patel T, Hackmann A, Wells A, Fischer
P & Myers S (2007). Ill believe it when I see it: Imagery
re-scripting of intrusive sensory memories. Journal of
Behaviour Therapy and Experimental Psychiatry 39, 371-
385.
Whelton WJ & Greenberg LS (2005). Emotion in self-criticism.
Personality and Individual Differences 38, 1583-1595.
Zuroff DC, Santor D & Mongrain M (2005). Dependency, self-
criticism, and maladjustment. In Auerbach JS, Levy KN,
Schaffer CE (Eds) Relatedness, self-Definition and Mental
Representation: Essays in Honour of Sidney J. Blatt.
London, Routledge, pp.75-90.
