Content uploaded by Sarah M C Robertson
Author content
All content in this area was uploaded by Sarah M C Robertson on Dec 01, 2016
Content may be subject to copyright.
This article was downloaded by: [108.85.154.155]
On: 08 May 2015, At: 17:44
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK
Click for updates
Aging & Mental Health
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/camh20
Physiological reactivity during autobiographical
narratives in older adults: the roles of depression
and anxiety
Sarah M.C. Robertsona, Rhonda J. Swickerta, Kathryn Connellya & Ann Galizioa
a Department of Psychology, College of Charleston, Charleston, SC, USA
Published online: 07 Oct 2014.
To cite this article: Sarah M.C. Robertson, Rhonda J. Swickert, Kathryn Connelly & Ann Galizio (2015) Physiological
reactivity during autobiographical narratives in older adults: the roles of depression and anxiety, Aging & Mental Health,
19:8, 689-697, DOI: 10.1080/13607863.2014.962010
To link to this article: http://dx.doi.org/10.1080/13607863.2014.962010
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied
upon and should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Physiological reactivity during autobiographical narratives in older adults:
the roles of depression and anxiety
Sarah M.C. Robertson*, Rhonda J. Swickert, Kathryn Connelly and Ann Galizio
Department of Psychology, College of Charleston, Charleston, SC, USA
(Received 30 May 2014; accepted 22 August 2014)
Objectives: Physiological reactivity (PR) describes the change in physiological functioning (e.g., heart rate, blood
pressure, pulse pressure) that occurs after the induction of a stressful task. This study aims to understand the influence of
mental health symptoms on patterns of PR during autobiographical narratives in an older adult sample.
Method: Eighty older adults completed self-report measures regarding their symptoms of depression and anxiety. Next,
their blood pressure was recorded while they completed two verbal autobiographical narratives.
Results: During the positive narrative, anxiety was positively associated with increased PR while depression was
negatively associated with PR. During the negative narrative, a significant interaction occurred whereby anxiety was
significantly positively associated with PR for those participants low in depression.
Discussion: The above results are explained in the context of the Tripartite Model of Depression and Anxiety, which
predicts different patterns of PR as a function of mental health symptoms. Limitations and future directions are also
discussed.
Keywords: physiological reactivity; anxiety; depression; aging
Introduction
Physiological reactivity (PR) describes the change in physi-
ological functioning (e.g., heart rate, blood pressure, pulse
pressure) that occurs after the induction of a stressful task
(Linden, Earle, Gerin, & Christenfeld, 1997). PR has
received increasing attention from researchers as more has
been discovered about its association with a variety of
physical and mental health problems (Evans et al., 2013).
Recent research indicates that several factors are important
in predicting PR, including both mental health symptoms
and age (Neupert, Miller, & Lachman, 2006;Pattenetal.,
2009; Uchino, Birmingham, & Berg, 2010). This study
extends our understanding of these variables by assessing
the influences of depression and anxiety on PR in a com-
munity-dwelling older adult sample. This study also
increases our understanding of potential self-regulatory
changes in late life by assessing the applicability of the Tri-
partite Model of Depression and Anxiety (Clark & Watson,
1991) to patterns of PR in an older adult sample. Primary
aims of this study are to (1) assess whether changes in PR
occur after the completion of stressful autobiographical
narrative tasks in an older adult sample, (2) assess whether
potential changes in PR occur as a function of mental
health symptoms in an older adult sample, and (3) evaluate
the applicability of the Tripartite Model of Anxiety and
Depression to patterns of PR in an older adult sample.
The Tripartite Model of Anxiety and Depression
Researchers have long noticed that anxiety and depressive
disorders often co-occur. The Tripartite Model represents
an important advancement in clinical psychology because
it describes in theoretical terms how depression and anxi-
ety are both similar to and different from one another
(Clark & Watson, 1991). According to this model, depres-
sion and anxiety are similar in that people with both disor-
ders experience negative affect (NA). NA represents the
extent to which a person experiences unpleasant emotions,
and people with both anxiety and depressive disorders
report experiencing NA (Clark & Watson, 1991). How-
ever, there are distinct clinical features that also separate
the two disorders. According to the Tripartite Model,
depressed people uniquely experience anhedonia, and
have difficulty experiencing pleasure in their lives. Anx-
ious people, on the other hand, uniquely experience physi-
ological hyperarousal, which can include the following
experiences: shaking, heart racing, shortness of breath,
and dizziness (Clark & Watson, 1991). Hyperarousal is a
very common feature of anxiety disorders and diagnostic
criteria for many anxiety disorders include physiological
arousal (American Psychiatric Association, 2013).
Despite the fact that anxiety and depressive disorders
share some symptoms and often co-occur, there is evi-
dence that they have distinct biological patterns as pre-
dicted by the Tripartite Model.
Depression and physiological reactivity
Symptoms of major depressive disorder, as defined in the
Diagnostic and Statistical Manual, Fifth Edition (DSM-5)
include depressed mood, anhedonia, weight loss or
gain, sleeping disturbance, psychomotor agitation or
*Corresponding author. Email: robertsonsm@cofc.edu
Ó2014 Taylor & Francis
Aging & Mental Health, 2015
Vol. 19, No. 8, 689697, http://dx.doi.org/10.1080/13607863.2014.962010
Downloaded by [108.85.154.155] at 17:44 08 May 2015
retardation, fatigue, worthlessness, decreased ability to
concentrate, and suicidality (American Psychiatric Asso-
ciation, 2013). Given this symptom profile, it is not sur-
prising that much data have indicated that depressive
symptoms are often associated with lower levels of PR.
For example, depressed participants demonstrated signifi-
cantly lower systolic blood pressure and heart rate after
the induction of a stressful task when compared to healthy
controls (Salomon, Clift, Karlsd
ottir, & Rottenberg,
2009), a finding that has been replicated by other research-
ers (Carroll, Phillips, Hunt, & Der, 2007). More recent
work has also confirmed the idea that depressive symp-
toms are associated with attenuated patterns of PR.
Schwerdtfeger and Rosenkaimer (2011) found that
depressive symptoms were negatively associated with
blood pressure measurements after the induction of a pub-
lic speaking task, while other work has demonstrated that
depression is associated with decreased reactivity during a
mental arithmetic task (Phillips, 2011). Also, when
required to complete mirror-tracing tasks and speech
tasks, participants with higher scores of depression exhib-
ited less heart rate reactivity (Salomon et al., 2009).
However, other research has also shown that
depression symptoms have been associated with
increased levels of PR. Health psychologists have
noted the increased risk for cardiovascular disease and
high blood pressure among depressed participants (Pat-
ten et al., 2009). Also, participants categorized as hav-
ing high symptoms of depression demonstrated
significantly greater systemic vascular resistance after
the initiation of a stressful task (Matthews, Nelesen, &
Dimsdale, 2005) and higher systolic and diastolic
blood pressure before, during, and after the completion
of a stressful speech task (Light, Kothandapani, &
Allen, 1998). Researchers have been attempting to rec-
oncile the contradictory findings associated with
depression symptoms and PR. One variable that may
explain the differences in these findings relates to
social context, as recent data have indicated that
merely thinking about an important social relationship
can increase the PR that occurs during a psychologi-
cally stressful task (Cyranowski, Hofkens, Swartz, &
Gianaros, 2011).
Anxiety and physiological reactivity
Much research has also been done to identify potential
associations between anxiety symptoms and PR. Hallmark
symptoms of many anxiety and trauma-related disorders
include physiological hyperarousal (e.g., palpitations and
trembling in panic disorder, PR when exposed to trau-
matic cues in post-traumatic stress disorder). Most find-
ings strongly support the notion that higher levels of
anxiety are associated with higher degrees of PR
(McTeague & Lang, 2012). For example, normotensive
students who were high in social anxiety symptoms dem-
onstrated increased levels of heart rate reactivity after the
completion of a speech task when compared to normoten-
sive students who were low in social anxiety symptoms
(Gramer, Schild, & Lurz, 2012). Additional work has
demonstrated that among a sample of children, anxiety
symptoms were more predictive of psychological arousal
after the completion of a stressful task than were depres-
sive symptoms (Dieleman, van der Ende, Verhulst, &
Huizink, 2010). Recent work has also indicated that
higher levels of PR can be identified in participants with
anxiety symptoms as well as their offspring (Jovanovic
et al., 2011). Children of abused mothers demonstrated an
increased startle response when compared to children
whose mothers had not experienced abuse.
While previous studies generally support the notion
that anxiety is associated with the increased levels of PR
and depression is generally associated with decreased lev-
els of PR (especially when one is not required to consider
social context), few studies have examined whether these
two affective patterns may interact. For example, when
considering someone who has high levels of both anxiety
and depression, what pattern of PR might be anticipated?
A recent study by de Rooij, Schene, Phillips, and Rose-
boom (2010) found that higher scores of composite
depression and anxiety (i.e., total scores on the Hospital
Anxiety and Depression Scale) were associated with
decreased diastolic blood pressure assessments in a sam-
ple of 725 participants. It is important to note that the psy-
chological stressor in this study consisted of several
cognitive tasks and one speech task. Given these data, it is
proposed that a significant interaction will occur whereby
anxiety will be a significant positive predictor of PR for
those low in depression, but not a significant predictor of
PR for those high in depression.
Age and physiological reactivity
‘Normal aging affects all physiological processes’ (Boss
& Seegmiller, 1981, p. 434). More specifically, cardiac
output decreases, blood pressure increases, lung capacity
decreases, and rates of ventilation decrease (Boss &
Seegmiller, 1981). What has been less clear in the litera-
ture, however, is how aging is related to patterns of PR.
While previous work has supported the idea that age is an
important factor to consider in the area of PR, it should
also be noted that social and emotional contextual factors
are also important to consider. Work by Kunzmann and
Richter (2009) demonstrated that when participants
between 2070 years of age watched films with age-
appropriate losses (e.g., cognitive impairment in the older
adult), there were no differences in patterns of PR as a
function of age. Other data support the notion that older
adults display similar patterns of reactivity as young
adults when describing past emotional experiences (Lev-
enson, Carstensen, Friesen, & Ekman, 1991). However,
data on this topic have been mixed, and some researchers
have reported that age is associated with differences in
patterns of PR. For example, after viewing emotionally
evocative film clips, older adults demonstrated greater
physiological responses than young adults (Seider, Shiota,
Whalen, & Levenson, 2011). Importantly, a meta-analysis
was conducted to assess the influence of age on PR after
the induction of a stressful emotional task (Uchino et al.,
2010). Results of this study indicated that there are indeed
690 S.M.C. Robertson et al.
Downloaded by [108.85.154.155] at 17:44 08 May 2015
differences in PR as a function of age. More specifically,
after analyzing 31 individual studies, age was associated
with decreased heart rate reactivity and increased systolic
blood pressure reactivity. It is proposed, given the Tripar-
tite Model of Depression and Anxiety, that some of the
inconsistency in findings related to PR and age can be
explained by the presence of depression and anxiety
symptoms. This study uniquely adds to the literature by
assessing the influence of mental health symptoms on pat-
terns of PR during autobiographical narratives in commu-
nity-dwelling older adults.
Autobiographical narratives
Although there is not a clear definition of an autobiographi-
cal narrative in the literature, it is widely accepted that narra-
tives ‘place events in the larger context of one’s life, they
relate events in a meaningful order, and most importantly,
they provide evaluations of events that express their impor-
tance and personal significance’ (Bohanek, Fivush, &
Walker, 2005, p. 51). The autobiographical narrative was
chosen as a psychological stressor for this study because (1)
it is highly ecologically valid and (2) meaningful informa-
tion may be imperceptibly obtained that might be more
readily avoided and less reliably assessed when using more
direct assessment methods. Older adults are known to often
share unstructured stories with other people in their every-
day lives, which enhances the ecological validity of this task
(Alea, Bluck, & Semegon, 2004). Also, utilizing narratives
as a mechanism for assessing verbal emotionality is perhaps
more accurate than asking people to rate how often they use
emotional words. Quantifying emotional word during narra-
tives eliminates some of the demand characteristics associ-
ated with a self-report measure of verbal emotionality.
While a considerable amount of work has been focused
on narrative expression and associated psychological con-
structs (e.g., identity formation, cognitive capacities, etc.),
there has not been an equivalent focus on understanding
physiological reactions during the expression of a personal
story (Heilmann, Miller, & Nockerts, 2010;Ibarra&Barbu-
lescu, 2010; Keddell, 2009; Robertson & Hopko, 2013;
Watson, 2009). Some important work has begun in this area
that will be reviewed here. Lawler et al. (2003) explored the
relationship between forgiveness and PR during narratives.
Results indicated that participants demonstrating trait for-
giveness experienced lower diastolic blood pressure, while
those showing state forgiveness had lower blood pressure
(both diastolic and systolic) and heart rate. Lawler-Row,
Karremans, Scott, Edlis-Matityahou, and Edwards (2008)
also used narratives to assess whether or not reductions in
anger (i.e., becoming less angry about a negative experi-
ence) had beneficial effects on physiological arousal or if
complete forgiveness had significantly greater physiological
benefits than only decreased anger. Findings from this study
showed that forgiveness, not a decrease in anger, has a bene-
ficial impact on physiological responding. This study builds
upon this work by assessing the roles of mental health varia-
bles in patterns of PR during autobiographical narratives in
a community-dwelling older adult sample.
Hypotheses
When considering the extant literatures on PR, depression,
anxiety, and aging, it seems clear that anxiety is generally
associated with increased PR, while the findings related to
depression and PR have been mixed. It is proposed that
depression will be associated with decreased PR in this
study, given that the majority of data supports this proposi-
tion and the fact that social context was not a specific
requirement of the task in this study. It is also proposed that
anxiety and depression will interact (i.e., anxiety will predict
PR for those low in depression, but not those high in depres-
sion). In this study, participants will complete two autobio-
graphical narratives (one about a positive emotional
experience and one about a negative emotional experience),
and their systolic blood pressure will be assessed during
their verbal description of emotional events. Systolic blood
pressure is defined as the pressure in the arteries when the
heart beats (as opposed to diastolic blood pressure which is
the pressure in the arteries when the heart rests) and is uti-
lized as the dependent variable in this study because it is
more strongly related to psychological stressors in the labo-
ratory and is also relatively more important than diastolic
reactivity in predicting medical outcomes (Everson et al.,
2001;Uchinoetal.,2010). Both positive and negative narra-
tives will be assessed because recent research suggests that
emotional complexity (experiencing NA and positive affect
simultaneously) increases in late life (Carstensen et al.,
2011),andassessingbothnegativelyandpositively
valenced stories allows the opportunity to comprehensively
capture both congruent and non-congruent emotions.
The following hypotheses will be tested within an older
adult sample: (1) It is predicted that participating in autobio-
graphical narrative tasks will result in a significant increase
in systolic blood pressure when compared to a relaxation
phase and a non-word reading phase. (2) It is predicted that
anxiety will be positively associated with systolic blood
pressure reactivity, while depression symptoms will be neg-
atively associated with systolic blood pressure reactivity. (3)
It is predicted that a significant interaction will occur
whereby anxiety will be a significant positive predictor of
PR for those low in depression, but not a significant predic-
tor of PR for those high in depression.
Method
Participants
Older adult participants (ages 5991) were recruited
through an advertisement in the local newspaper (nD80).
Participants were not eligible for the study if they were
currently being treated with medications known to alter
physiological status (e.g., beta-blockers, anti-depressants,
anti-inflammatory medications) or if they were under
55 years of age. Demographic characteristics of the sam-
ple are presented in Table 1.
Measures
Beck Depression Inventory (BDI), second edition: the
BDI is a 21-item scale that assesses affective,
Aging & Mental Health 691
Downloaded by [108.85.154.155] at 17:44 08 May 2015
physiological, and cognitive symptoms of depression
(Beck, Steer, & Brown, 1996). Each question is measured
on a 4-point Likert-type scale, with higher total scores
representing higher levels of depression. In the current
sample, internal consistency was strong (aD.91).
Beck Anxiety Inventory (BAI): the BAI is a 21-item
questionnaire designed to assess cognitive and physiologi-
cal symptoms of anxiety (Beck & Steer, 1993). Each
question is measured on a 4-point Likert-type scale, and
higher total scores on the BAI represent higher levels of
anxiety. In the current sample, internal consistency was
strong (aD.92).
Linguistic Inquiry Word Count (LIWC; Pennebaker,
Booth, & Francis, 2007): this computer program analyzes
the linguistic content of narratives. Researchers were
responsible for transcribing all narratives immediately
after the data were collected. A second member of the
research team reviewed the transcription for accuracy and
data were then entered into the LIWC software program.
This program has good psychometric properties and has
been utilized by several narrative researchers in previous
studies (Fivush, Edwards, & Mennuti-Washburn, 2003;
Pennebaker et al., 2007; Smith, Anderson-Hanley, Lan-
grock, & Compas, 2005). All data, as is consistent with
the LIWC program, are expressed as percentages of total
words utilized. The Total Affect variable is analyzed for
this study, and computes the percentage of all affective
words within a narrative (e.g., happy, cried, abandon).
More specifically, both NA and positive affect words are
captured by this variable. Total Affect was selected (rather
than only NA words or only positive affect words) given
the literature that suggests emotional complexity increases
in late life (Carstensen et al., 2011).
Physiological Assessment: all blood pressure record-
ings were conducted with a General Electric Carescape
V100 Vital Signs Monitor. A blood pressure cuff was
placed on the left arm of all the participants.
Procedures
All the participants initially completed a consent form and
a brief packet of questionnaires, including a demographic
form, the BDI, and the BAI. Next, participants viewed a
relaxing nature video for 3 minutes (Gross & Levenson,
1995). Blood pressure was recorded at two points during
this relaxation phase: once at the beginning and once at
the end. Next, participants completed a reading task that
required them to read non-words out loud. This list of
non-words had been previously developed for the pur-
poses of assessing literacy skills (Good & Kaminski,
2002). Examples of these items include, ‘sed’, ‘tob’, and
‘bil’. The reading task was completed in order to ensure
that changes in PR during the narrative phase were not
due to the effects of speaking alone. Blood pressure was
recorded twice during the reading phase (once at the
beginning and once at the end). Then, participants moved
to the narrative phase of the study, in which they recalled
both a positive and negative story about their lives (the
order of narratives was counterbalanced throughout the
study). Blood pressure was recorded at two points for
both the positive and negative narrative (once at the
beginning and once when a prompt was given). Blood
pressure was recorded after the prompt from the
researcher in order to ensure that all participants were
speaking, thereby reducing experimental noise associated
with variability in speech production. Including the two
readings from the relaxation phase, a total of eight blood
pressure measurements were taken. The two readings
from each phase were averaged. Specific instructions for
the autobiographical narratives were as follows: ‘Today I
am interested in hearing about the different experiences
that people have in life. Please describe a positive emo-
tional experience that you have been through. Try to
imagine the event in your mind and talk continuously
about the thoughts and feelings related to the event you
are describing. Please try to do this in as much detail as
possible. I will record your blood pressure at two different
points during your story: once at the beginning and once
towards the end. Do you have any questions?’ Participants
had up to 10 minutes to complete each narrative. After
participants stopped speaking, they were all prompted
with the question, ‘Is there anything else you can remem-
ber about that event?’
Of the 160 narratives, there were two negative auto-
biographical narratives that lasted for 10 minutes. Once
10 minutes had elapsed, the researcher asked, ‘Ten
minutes have now elapsed. Are you comfortable moving
on now?’ Both participants agreed to move on following
this prompt. Out of the 640 total readings, 5 were not
completed due to errors with the machine. Often, partici-
pants moved their arms while discussing their narratives,
even when they were instructed to stay as still as possible.
Therefore, the blood pressure monitor was unable to attain
a reading for these five assessments. There were three
incomplete readings during the positive narrative and two
Table 1. Demographic descriptive statistics.
Older adults (ND80)
Age (mean, SD) 72.64 (7.54)
Years of education (mean, SD) 15.28 (2.15)
Gender (women) 52 (65.00%)
Ethnicity
Caucasian 77 (96.25%)
African-American 1 (1.25%)
Latino 1 (1.25%)
Asian 1 (1.25%)
Marital status
Single 7 (8.75%)
Married 47 (58.75%)
Separated 2 (2.50%)
Divorced 11 (13.75%)
Widowed 13 (16.25%)
Occupational status
Employed full-time 8 (10.00%)
Employed part-time 9 (11.25%)
Unemployed-seeking employment 2 (2.50%)
Unemployed-not seeking employment 61 (76.25%)
692 S.M.C. Robertson et al.
Downloaded by [108.85.154.155] at 17:44 08 May 2015
incomplete readings during the negative narrative. All of
the incomplete readings were completed by different par-
ticipants. No incomplete readings occurred during the
relaxation or reading phases. Because blood pressure data
during the narratives (i.e., the dependent variable) were
not available for these five participants, this data was not
used when running hierarchical regressions.
Results
Manipulation checks
Prior to the statistical analyses of autobiographical narra-
tives, a manipulation check was performed to determine
whether counterbalancing of narratives was successful.
There were no significant differences in the PR as a func-
tion of narrative task order [average systolic blood pres-
sure during positive narrative F(1,34) D1.98, pD.17;
average systolic blood pressure during negative narrative
F(1,35) D.08, pD.78)]. Also, there was not a signifi-
cant difference between the length of narratives as a func-
tion of narrative valence: [F(1,158) D.00, pD1.00;
average negative narrative length: (MD209.38 seconds,
SD D149.62); average positive narrative length: (MD
194.41 seconds, SD D148.13)].
Blood pressure data across time
A Repeated Measures Analysis of Variance was completed to
determine whether systolic blood pressure differed between
the relaxation phase, the reading phase, the positive narra-
tive, and the negative narrative. There was a significant
difference across the four time points: F(1,75) D4310.92,
pD.00. Post hoc analyses indicated that there was a sig-
nificant difference between relaxation (MD139.52, SE D
2.33) and both the positive narrative [(MD152.72, SE D
2.58), mean difference D13.20, pD.00) and the negative
narrative [(MD155.31, SE D2.69), mean difference D
15.79, pD.00]. There was also a significant difference
between reading (MD137.98, SE D2.10) and both the
positive narrative (mean difference D14.74, pD.00) and
the negative narrative (mean difference D17.33, pD.00).
There was not a significant difference between relaxation
and reading (mean difference D1.54, SE D1.25, pD
.22). In relation to narrative valence, there was not a sig-
nificant difference between the positive and negative nar-
rative (mean difference D2.59, SE D1.43, pD.07).
Please see Table 2 for descriptive statistics related to
blood pressure data and other dependent measures.
The effects of depression and anxiety on systolic
blood pressure
Hierarchical linear regression was utilized to assess the
ability of anxiety and depression to predict PR during
autobiographical narratives after controlling for the influ-
ences of (1) PR during the reading task and (2) the per-
centage of emotional words utilized during the narratives
as measured by the LIWC program. This model was cho-
sen to ensure that PR during the autobiographical narra-
tives was not due to the act of speaking alone. This model
was also chosen to control the potential influence of the
emotional content of the narratives. For example, some
participants might have utilized a higher percentage of
emotional words in their narrative than other participants,
which could influence patterns of PR. Controlling for
these potential confounds allows for a clearer understand-
ing of the relationship between mental health symptoms
and patterns of PR. Preliminary analyses were conducted
to ensure no violation of the assumptions of normality,
linearity, multicollinearity, and homoscedasticity. In order
to reduce collinearity among variables, anxiety and
depression scores were centered prior to forming the prod-
uct term. Systolic blood pressure during the reading phase
and the percentage of emotional word utilized were
entered at Step 1. Anxiety and depression were entered at
Step 2, and the depression X anxiety interaction was
entered at Step 3. Hierarchical linear regressions were
completed for both the negative and positive autobio-
graphical narratives.
Negative autobiographical narrative. Step 1: Average
systolic blood pressure during the reading phase and the
percentage of emotional words utilized explained 52.5%
of the variance in average systolic blood pressure during
the negative autobiographical narrative [F
change
(2,74) D
40.89, pD.00]. Step 2: After entry of depression and anx-
iety at Step 2 the total variance explained by the model as
Table 2. Descriptive statistics for study variables.
Older adults (ND80)
NMean (SE) Range
Systolic blood pressure relaxation phase 80 139.52 (2.33) 95.50184.50
Systolic blood pressure reading phase 80 137.98 (2.10) 103.50193.00
Systolic blood pressure positive narrative 77 152.72 (2.58) 114.50209.00
Systolic blood pressure negative narrative 78 155.31 (2.69) 111.50218.00
BDI-II 80 5.90 (4.25) 021
BAI 80 4.29 (4.81) 024
LIWC (total affect-positive narrative expressed as percentage) 80 4.62 (1.97) .9512.50
LIWC (total affect-negative narrative) expressed as percentage) 80 4.51 (1.85) 09.35
Note: There are only 77 participants in the positive narrative and 78 participants in the negative narrative due to errors with the blood pressure machine.
See procedures section for more information.
Aging & Mental Health 693
Downloaded by [108.85.154.155] at 17:44 08 May 2015
a whole was 54.3%, which was not significantly higher
than the previous model [F
change
(2,72) D1.42, pD.25].
Step 3: After the entry of the interaction between depres-
sion and anxiety at Step 3 the total variance explained by
the model as a whole was 57.3%, which was significantly
higher than the previous model [F
change
(1,71) D4.96, p
D.03]. Please see Table 3 for a compiled list of beta
weights and R
2
values.
The significant interaction between anxiety and
depression was decomposed utilizing procedures
described in Bauer and Curran (2005), which describes
the application of the JohnsonNeyman technique for
probing interactions in hierarchical or multilevel models.
The JohnsonNeyman technique allows for the calcula-
tion of a range of values for which there is a significant
interaction, rather than the ‘pick-a-point’ approach, which
calculates significance of simple slopes at a limited num-
ber of values. Results indicated that anxiety was a signifi-
cant predictor of systolic blood pressure when centered
depression scores were less than 3.64. In order to plot this
interaction, simple slopes of the lines predicting systolic
blood pressure from depression were calculated at condi-
tional values of anxiety that were one standard deviation
above and below the mean (Bauer & Curran, 2005). Slope
analyses indicated that for participants who were low in
depression, there was a significant positive effect of anxi-
ety whereby increased anxiety was associated with
increased systolic blood pressure [simple slope D2.03
(SE D.71), tD2.86, pD.01]. However, for participants
who were high in depression, there was no significant
effect of anxiety [simple slope D.91 (SE D.50), tD1.83,
pD.09]. Importantly, these findings were significant even
after controlling for the effects of reading and the percent-
age of emotional words utilized on PR. Please see
Figure 1.
Positive autobiographical narrative. Step 1: Average
systolic blood pressure during the reading phase and the
percentage of emotional words utilized explained 62.2%
of the variance in average systolic blood pressure during
the positive autobiographical narrative [F
change
(2,72) D
59.29, pD.00]. Step 2: After the entry of depression and
anxiety at Step 2 the total variance explained by the model
as a whole was 66.3% [F
change
(2,70) D4.23, pD.02],
which was significantly higher than the previous model.
Step 3: After the entry of the interaction between depres-
sion and anxiety at Step 3 the total variance explained by
the model as a whole was 67.3% [F
change
(1,69) D2.08, p
D.15], which was not significantly higher than the previ-
ous model.
Regression analyses indicated a significant main effect
for anxiety [BD.28, t(73) D2.91, pD.01] whereby
increased anxiety was associated with increased systolic
blood pressure. There was also a significant main effect of
depression [BD¡.19, t(73) D¡2.00, pD.04] whereby
increased depression was associated with decreased sys-
tolic blood pressure. Importantly, these findings were sig-
nificant even after controlling for the effects of reading
and the percentage of emotional words utilized on PR.
Discussion
Results provided moderate support for study hypotheses,
and each hypothesis and the associated data will be dis-
cussed. Results of the study were supportive of Hypothe-
sis 1, in that participants demonstrated significant
increases in systolic blood pressure after the relaxation
phase of the study. Importantly, this increase in PR
occurred for both the positive and negative narratives.
Speaking of an emotional event is associated with
increased PR, regardless of the type of event described.
Table 3. Results of hierarchical regression analyses standardized beta coefficients and R
2
values.
Systolic blood pressure:
negative narrative
Systolic blood pressure:
positive narrative
Independent variable
Step 1
Reading systolic blood pressure (b) .73
.77
Emotional content (b)¡.03 ¡.22
DR
2
.53
.62
Step 2
Reading systolic blood pressure (b) .72
.76
Emotional content (b)¡.03 ¡.21
Depression (b)¡.12 ¡.20
Anxiety (b) .19 .28
DR
2
.02 .04
Step 3
Reading systolic blood pressure (b) .70
.75
Emotional content (b)¡.03 ¡.21
Depression X anxiety interaction (b)¡.25
¡.14
DR
2
.03
.01
Significant at .05 level.
Significant at .01 level.
694 S.M.C. Robertson et al.
Downloaded by [108.85.154.155] at 17:44 08 May 2015
Hypotheses 2 and 3 are important to consider together
given that they related to each other. Hypotheses 2 and 3
predicted that (1) anxiety would be positively associated
with PR, while depression would be negatively associated
with PR and (2) a significant interaction would occur
whereby anxiety will be a significant positive predictor of
PR for those low in depression, but not a significant predic-
tor of PR for those high in depression. Data from this study
provide support for these hypotheses, and the predictions
will be discussed separately for each narrative phase.
Negative narrative
There was a significant interaction whereby anxiety was a
positive predictor of PR for those low in depression, but not
a significant predictor of PR for those high in depression.
An important finding of this study is that participants with
higher depression scores did not demonstrate differences in
PR as a function of anxiety. While anxiety is typically posi-
tively associated with PR, this pattern does not occur among
older adults high in depression. It appears that depressive
symptoms in the older adult seem to be physiologically
overpowering the influence of anxiety symptoms, and leads
one to wonder whether depression has a more profound
impact on PR during negative narratives than anxiety. In
other words, while anxiety is typically positively associated
with PR, its effects are weakened when considering older
adults with high levels of depression.
Previous work has been done on depression and age
that could play an important role in explaining this inter-
action effect. Research has indicated that older adults
report more somatic symptoms of depression than young
adults, while young adults report more affective
symptoms than older adults (Fiske, Wetherell, & Gatz,
2009). If older adults experience more of the somatic
symptoms of depression, it might require a relatively
greater amount of anxiety to induce PR, given that depres-
sion symptoms are often associated with dampened pat-
terns of PR. Data from this study supports this notion, as
higher levels of depression were associated with attenu-
ated patterns of PR. The implications of this finding are
important for older adults and clinicians alike. Struggling
with depression as an older adult will often be associated
with significant changes in an older adult’s physiological
functioning, and somatic symptoms of depression like
lethargy and sleep disturbances can become especially
pronounced. It is important for clinicians to inquire about
the status of somatic symptoms in the depressed older
adult population and potentially address these symptoms
by implementing an empirically supported treatment
aimed at improving somatic symptoms of depression (i.e.,
behavioral activation; Hopko et al., 2011).
It is also possible that depression could be serving to
influence the schema of an older adult in such a way that
priming could play an important role. Beck et al. have
described a depressive schema as an interconnected system
of thoughts that allows for a negative internal representa-
tion of the self (Beck, Rush, Shaw, & Emery, 1979). When
an older adult describes a negative emotional experience, it
is possible that this narrative description activates underly-
ing negative schemata. In other words, talking about a neg-
ative event can serve as a primer for the activation of
negative internal representations. If a depressive schema is
activated, an older adult might not demonstrate equivalent
physiological arousal, given that depression is so often
associated with the dampened patterns of the PR. Further
Figure 1. Simple slopes of depression at low and high levels of anxiety during negative narrative.
Aging & Mental Health 695
Downloaded by [108.85.154.155] at 17:44 08 May 2015
exploration of the potential role of cognitive schemata on
patterns of the PR is indicated.
Positive narrative
Anxiety was significantly positively associated with PR
during the positive narrative, and depression was signifi-
cantly negatively associated with PR during the positive
narrative. There was not a significant interaction between
anxiety and depression during the positive narrative.
While the main effects of anxiety and depression were
predicted, it is interesting that the anticipated interaction
did not occur as it did with the negative autobiographical
narrative. There seems to be a difference in the role of
depression and anxiety symptoms as a function of narra-
tive valence. Anxiety was positively associated with pat-
terns of PR during the positive narrative for all
participants, not just those who were low in depression as
occurred in the negative narrative. Perhaps recounting a
negative life story is relatively more stressful than
recounting a positive life story. The task of recounting a
negative story might be so emotionally taxing that PR
becomes less influenced by anxiety given that the depres-
sive symptoms are potentially more activated in this spe-
cific type of task. This would explain why participants
high in depression did not demonstrate differences in PR
as a function of anxiety during the negative narrative.
Likewise, when given a relatively easier task to complete
(i.e., recounting a positive story), anxiety symptoms
become more important to consider given that the depres-
sion symptoms are not as activated as they would have
been when recounting a negative story. In future work, it
will be important to empirically measure the subjective
distress one experiences during these tasks to further
understand its potential role in influencing patterns of PR.
There are several limitations to address with this
study. First, this study only included older adults. Ideally,
one would like to have access to a variety of age groups to
understand how potential differences occur across the life-
span. Second, physiological measurement is ideally done
in a continuous manner. However, in situations where this
level of equipment is unavailable, non-continuous assess-
ment has also been supported in the literature (Lawler
et al., 2003). Replication of this work utilizing a continu-
ous method of measurement will help to strengthen the
results presented here. Third, it is important to note the
sample obtained for this study is predominantly Caucasian
and replication of this work in minority groups is greatly
needed. Fourth, it is important to note that the LIWC pro-
gram does not rate the intensity of words utilized. Using
the word ‘upset’, ‘sad’, and ‘furious’ would all be catego-
rized in the Total Affect category. Future work that exam-
ines the potential function of the intensity of these words
is needed. Also, it should be noted that the majority of the
sample was composed of women (nD52, 65.00%), and
additional study is needed to understand how gender
might influence patterns of PR in older adults. Finally, the
older adults in this study covered a large range of ages
(from 59 to 91). It is quite possible that variability in pat-
terns of PR could occur within this range, and future work
that examines PR in youngold, old, and oldold catego-
ries will be helpful.
In sum, data from this study provide moderate sup-
port for the Tripartite Model of Anxiety and Depression.
During the negative autobiographical narrative, a signif-
icant interaction occurred whereby anxiety was signifi-
cantly positively associated with PR for participants low
in depression. During the positive autobiographical nar-
rative, there were two significant main effects that
occurred. Anxiety was positively associated with PR,
while depression was negatively associated with PR. It
will be important for future work to address how these
patterns of PR might be similar to or different from pat-
terns of PR in young and middle aged adults. Having
data from several different age groups would allow
researchers and clinicians to more comprehensively
understand self-regulatory processes across the lifespan.
References
Alea, N., Bluck, S., & Semegon, A.B. (2004). Young and older
adults’ expression of emotional experience: Do autobio-
graphical narratives tell a different story? Journal of Adult
Development, 11, 235250. doi:10.1023/B:JADE.
0000044527.52470.5d
American Psychiatric Association. (2013). Diagnostic and sta-
tistical manual of mental disorders: DSM-V. Washington,
DC: Author.
Bauer, D.J., & Curran, P.J. (2005). Probing interactions in fixed
and multilevel regression: Inferential and graphical techni-
ques. Multivariate Behavioral Research, 40(3), 373400.
doi:10.1207/s15327906mbr4003_5
Beck, A.T. Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cogni-
tive theory of depression. New York, NY: Guilford Press.
Beck, A.T., & Steer, R.A. (1993). Beck anxiety inventory: Man-
ual. San Antonio, TX: Psychological Corporation.
Beck, A.T., Steer, R.A., & Brown, G.K. (1996). Manual for the
Beck depression inventory-II. San Antonio, TX: Psychologi-
cal Corporation.
Bohanek, J.G., Fivush, R., & Walker, E. (2005). Memories of
positive and negative emotional events. Applied Cognitive
Psychology, 19,5166. doi:10.1002/acp.1064
Boss, G.R., & Seegmiller, J.E. (1981). Age-related physiological
changes and their clinical significance. Journal of Western
Medicine, 135, 434440.
Carroll, D., Phillips, A.C., Hunt, K., & Der, G. (2007). Symptoms
of depression and cardiovascular reactions to acute psycholog-
ical stress: Evidence from a population study. Biological Psy-
chology, 75,6874. doi:10.1016/j.biopsycho.2006.12.002
Carstensen, L.L., Turan, B., Scheibe, S., Ram, N., Ersner-Hersh-
field, H., Samanez-Larkin, G.R., ... Nesselroade, J.R.
(2011). Emotional experience improves with age: Evidence
based on over 10 years of experience sampling. Psychology
and Aging, 26,2133. doi:10.1037/a0021285
Clark, L.A., & Watson, D. (1991). Tripartite model of anxiety
and depression: Psychometric evidence and taxonomic
implications. Journal of Abnormal Psychology, 100,
316336. doi:10.1037//0021-843X.100.3.316
Cyranowski, J.M., Hofkens, T.L., Swartz, H.A., & Gianaros, P.J.
(2011). Thinking about a close relationship differentially
impacts cardiovascular stress responses among depressed
and nondepressed women. Health Psychology, 30,
276284. doi:10.1037/a0023005
de Rooij, S.R., Schene, A.H., Phillips, D.I., & Roseboom,
T.J. (2010). Depression and anxiety: Associations with bio-
logical and perceived stress reactivity to a psychological
stress protocol in a middle-aged population.
696 S.M.C. Robertson et al.
Downloaded by [108.85.154.155] at 17:44 08 May 2015
Psychoneuroendocrinology, 35, 866877. doi:10.1016/j.
psyneuen.2009.11.011
Dieleman, G.C., van der Ende, J., Verhulst, F.C., & Huizink, A.
C. (2010). Perceived and physiological arousal during a
stress task: Can they differentiate between anxiety and
depression? Psychoneuroendocrinology, 35, 12231234.
doi:10.1016/j.psyneuen.2010.02.012
Evans, B.E., Greaves-Lord, K., Euser, A.S., Tulen, J.M., Franken,
I.A., & Huizink, A.C. (2013). Determinants of physiological
and perceived physiological stress reactivity in children
and adolescents. PLoS ONE, 8,113. doi:10.1371/journal.
pone.0061724
Everson, S.A., Lynch, J.W., Kaplan, G.A., Lakka, T.A.,
Sivenius, J., & Salonen, J.T. (2001). Stress-induced
blood pressure reactivity and incident stroke in middle-aged
men. Stroke, 32,12631270. doi:10.1161/01.STR.32.6.1263
Fiske, A., Wetherell, J.L., & Gatz, M. (2009). Depression in
older adults. Annual Review of Clinical Psychology, 5,
363389. doi:10.1146/annurev.clinpsy.032408.153621
Fivush, R., Edwards, V.J., & Mennuti-Washburn, J. (2003). Nar-
ratives of 9/11: Relations among personal involvement, nar-
rative content and memory of the emotional impact over
time. Applied Cognitive Psychology, 17, 10991111.
doi:10.1002/acp.988
Good, R.H., & Kaminski, R.A. (Eds.). (2002). Dynamic indica-
tors of basic early literacy skills (6th ed.). Eugene, OR: Insti-
tute for the Development of Educational Achievement.
Gramer, M., Schild, E., & Lurz, E. (2012). Objective and per-
ceived physiological arousal in trait social anxiety and post-
event processing of a prepared speaking task. Personality
and Individual Differences, 53, 980984. doi:10.1016/j.
paid.2012.07.013
Gross, J., & Levenson, R.W. (1995). Emotion elicitation using
films. Cognition and Emotion, 9,87108. doi:10.1080/
02699939508408966
Heilmann, J., Miller, J.F., & Nockerts, A. (2010). Sensitivity of
narrative organization measures using narrative retells pro-
duced by young school-age children. Language Testing, 27
(4), 603626. doi:10.1177/0265532209355669
Hopko, D.R., Armento, M.A., Robertson, S.C., Ryba, M.M.,
Carvalho, J.P., Colman, L.K., & ... Lejuez, C.W. (2011).
Brief behavioral activation and problem-solving therapy for
depressed breast cancer patients: Randomized trial. Journal
of Consulting and Clinical Psychology, 79(6), 834849.
doi:10.1037/a0025450
Ibarra, H., & Barbulescu, R. (2010). Identity as narrative: Preva-
lence, effectiveness, and consequences of narrative identity
work in macro work role transitions. The Academy of Man-
agement Review, 35(1), 135154. doi:10.5465/
AMR.2010.45577925
Jovanovic,T.,Smith,A.,Kamkwalala,A.,Poole,J.,Samples,T.,
Norrholm, S.D., & ... Bradley, B. (2011). Physiological
markers of anxiety are increased in children of abused mothers.
Journal of Child Psychology and Psychiatry and Allied Disci-
plines, 52,844852. doi:10.1111/j.1469-7610.2011.02410.x
Keddell, E. (2009). Narrative as identity: Postmodernism, multi-
ple ethnicities, and narrative practice approaches in social
work. Journal of Ethnic & Cultural Diversity in Social
Work: Innovation in Theory, Research & Practice, 18(3),
221241. doi:10.1080/15313200903070973
Kunzmann, U., & Richter, D. (2009). Emotional reactivity
across the adult life span: The cognitive pragmatics make a
difference. Psychology and Aging, 24, 879889.
doi:10.1037/a0017347
Lawler, K.A., Younger, J.W., Piferi, R.L., Billington, E., Jobe,
R., Edmondson, K., & Jones, W.H. (2003). A change of
heart: Cardiovascular correlates of forgiveness in response
to interpersonal conflict. Journal of Behavioral Medicine,
26, 373393. doi:10.1023/A:1025771716686
Lawler-Row, K.A., Karremans, J.C., Scott, C., Edlis-Matitya-
hou, M., & Edwards, L. (2008). Forgiveness, physiological
reactivity and health: The role of anger. International Jour-
nal of Psychophysiology, 68,5158. doi:10.1016/j.
ijpsycho.2008.01.001
Levenson, R.W., Carstensen, L.L., Friesen, W.V., & Ekman, P.
(1991). Emotion, physiology, and expression in old age.
Psychology and Aging, 6,2835. doi:10.1037/0882-
7974.6.1.28
Light, K.C., Kothandapani, R.V., & Allen, M.T. (1998).
Enhanced cardiovascular and catecholamine responses in
women with depressive symptoms. International Journal of
Psychophysiology, 28, 157166. doi:10.1016/S0167-8760
(97)00093-7
Linden, W., Earle, T.L., Gerin, W., & Christenfeld, N. (1997),
Physiological stress reactivity and recover: Conceptual sib-
lings separated at birth? Journal of Psychosomatic Research,
42, 117135. doi:10.1016/S0022-3999(96)00240-1
Matthews, S.C., Nelesen, R.A., & Dimsdale, J.E. (2005).
Depressive symptoms are associated with increased sys-
temic vascular resistance to stress. Psychosomatic Medicine,
67, 509513. doi:10.1097/01.psy.0000160467.78373.d8
McTeague, L.M., & Lang, P.J. (2012). The anxiety spectrum and
the reflex physiology of defense: From circumscribed fear to
broad distress. Depression and Anxiety, 29, 264281.
doi:10.1002/da.21891
Neupert, S.D., Miller, L.M.S., & Lachman, M.E. (2006). Physio-
logical reactivity to cognitive stressors: Variations by age
and socioeconomic status. International Journal of Aging
and Human Development, 26, 221235. doi:10.2190/17DU-
21AA-5HUK-7UFG
Patten, S.B., Williams, J.V., Lavorato, D.H., Campbell, N.R., Eli-
asziw, M., & Campbell, T.S. (2009). Major depression as a
risk factor for high blood pressure: Epidemiologic evidence
from a national longitudinal study. Psychosomatic Medicine,
71, 273279. doi:10.1097/PSY.0b013e3181988e5f
Pennebaker, J.W., Booth, R.J., & Francis, M.E. (2007). Linguis-
tic inquiry and word count: LIWC 2007. Austin, TX: LIWC.
Phillips, A.C. (2011). Blunted cardiovascular reactivity
relates to depression, obesity, and self-reported health. Bio-
logical Psychology, 86(2), 106113. doi:10.1016/j.
biopsycho.2010.03.016
Robertson, S.M.C., & Hopko, D.R. (2013). Emotional expres-
sion during autobiographical narratives as a function of
aging: Support for the Socioemotional Selectivy Theory.
Journal of Adult Development, 20,7686. doi:10.1007/
s10804-013-9158-6
Salomon, K., Clift, A., Karlsd
ottir, M., & Rottenberg, J. (2009).
Major depressive disorder is associated with attenuated car-
diovascular reactivity and impaired recovery among those
free of cardiovascular disease. Health Psychology, 28,
157165. doi:10.1037/a0013001
Schwerdtfeger, A., & Rosenkaimer, A.K. (2011). Depressive
symptoms and attenuated physiology reactivity to laboratory
stressors. Biological Psychology, 87, 430438. doi:10.1016/
j.biopsycho.2011.05.009
Seider, B.H., Shiota, M.N., Whalen, P., & Levenson, R.W.
(2011). Greater sadness reactivity in late life. Social Cogni-
tive and Affective Neuroscience, 6, 186194. doi:10.1093/
scan/nsq069
Smith, S., Anderson-Hanley, C., Langrock, A., & Compas, B.
(2005). The effects of journaling for women with newly
diagnosed breast cancer. Psycho-Oncology, 14(12), 1075
1082. doi:10.1002/pon.912
Uchino, B.N., Birmingham, W., & Berg, C.A. (2010). Are
older adults less or more physiologically reactive? A meta-
analysis of age-related differences in cardiovascular reactiv-
ity to laboratory tasks. Journal of Gerontology: Psychologi-
cal Sciences, 65B(2), 154162. doi:10.1093/geronb/gbp127
Watson, C. (2009). ‘Teachers are meant to be orthodox’: Narra-
tive and counter narrative in the discursive construction of
‘identity’ in teaching. International Journal of Qualitative
Studies in Education, 22(4), 469483.
Aging & Mental Health 697
Downloaded by [108.85.154.155] at 17:44 08 May 2015
