Upload
lediep
View
218
Download
0
Embed Size (px)
Citation preview
Worry and Rumination: Do They Prolong Physiological and Affective Recovery From Stress?
Short title:Do Worry and Rumination Prolong Recovery From Stress?
Lora Capobianco, PhD, Julie A. Morris, MSc, and Adrian Wells, PhD University of Manchester
Author NoteDr Lora Capobianco, PhD, School of Psychological Sciences, University of Manchester, [email protected] , Zochonis Building, 2.43, School of Psychological Sciences, University of Manchester, Manchester, UK, M13 9PL, 07 527 989115 Miss Julie A. Morris, Institute of Population Health, University of Manchester; [email protected] , 1st floor, Education & Research Centre, University Hospital of South Manchester, Wythenshawe Hospital, Southmoor Road, Manchester, UK, M23 9LT, 0161 291 5815
Professor Adrian Wells, PhD, School of Psychological Sciences, University of Manchester, & Greater Manchester Mental Health and Social Care NHS Trust, PhD, [email protected] , Rawnsley Building, Manchester Royal Infirmary, Oxford Road, Manchester, UK, M13 9WL, 0161 276 5399
Correspondence concerning this article should be addressed to Prof. Adrian Wells, Rawnsley Building, Manchester Royal Infirmary, Oxford Road, Manchester, UK, M13 9WL. Email: [email protected] Phone Number: +44 (0)161 276 5399
Conflicts of Interest and Source of Funding: The study was conducted for partial completion of the first authors PhD, which was supervised by Professor Adrian Wells and funded by the Presidents Doctoral Scholar Award from the University of Manchester. The authors report no
conflict of interest.
Do Worry and Rumination Prolong Recovery From Stress? 2
Abstract
Background: Anxiety and depression have been linked to repetitive negative self-regulation
styles in the form of worry and rumination respectively. Following predictions of the
metacognitive model [Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical
perspective.Hove: Erlbaum], the current study compared the effect on recovery from stress
of three strategies: worry, rumination, and distraction using a stress-exposure paradigm. It
was predicted that worry and/or rumination would be associated with delayed recovery on
physiological and/or self-report indices when compared to distraction.Method: Fifty-four
participants were randomly assigned to one of the self-regulation conditions and then
completed a modified version of the Trier Social Stress Test [TSST; Kirschbaum, C., Pirke,
K. M., & Hellhammer, D. H. (1993). The “TSST” – A tool for investigating psychobiological
stress responses in a laboratory setting. Neuropsychobiology, 28, 76–81.
doi:10.1159/000119004].. Skin conductance and negative affect (NA) measurements were
obtained at six different time points.
Results: Skin conductance indicated a prolonged recovery in the rumination condition when
compared with the distraction condition. Individuals in the worry condition reported an
immediate-delayed recovery from stress marked by higher NA scores in comparison to the
distraction condition.
Conclusions: These findings may provide important insights into the effects of different
forms of repetitive negative thinking on physiological and psychological recovery from
stress. The different response patterns observed are discussed within the context of the
metacognitive model.
Keywords: Worry; Rumination; Metacognition; Stress Recovery; Psychophysiology
Do Worry and Rumination Prolong Recovery From Stress? 3
Introduction
The emotion regulation strategies that individuals engage in during and after stressful
situations may play an integral part in recovery from a stress exposure. According to the
metacognitive model of psychological disorder (Wells & Matthews, 1994, 1996), repetitive
negative thinking in the form of worry and rumination are paradoxical trans-diagnostic self-
regulatory processes that may prolong the downregulation of psychological distress.
Worry is one of the key cognitive features of Generalized Anxiety Disorder
(GAD), and has been defined as a chain of negative thoughts that are predominantly
verbal and aimed at problem solving (Borkovec, Robinson, Pruzinsky, & DePree,
1983). Worry is a future oriented process that is used to anticipate danger and develop
ways to avoid it. Worry seeks answers to questions such as “What should I do in the
future,” and “How can I avoid danger?” (Wells, 2009). Borkovec, Alcaine, & Behar
(2004) proposed that worry is a cognitive avoidance response to perceived threat,
whereby individuals use worry as a means of problem solving any future danger. Fur-
thermore, worry has been conceptualized as a strategy of avoiding intrusive thoughts
about negative experiences (Borkovec, Alcaine, & Behar, 2004).
A similar process, rumination, has been commonly linked to depression and has
been defined in various ways such as, difficult-to-control repetitive thoughts concern-
ing personal problems. According to Nolen-Hoeksema’s (1991) response styles the-
ory, rumination is defined as, “repetitively focusing on the fact that one is depressed;
on one’s symptoms of depression; and on the causes, meanings, and consequences of
depressive symptoms” ( p. 569). Although rumination has been commonly associated
with depression, there is increasing research that suggests that rumination is also asso-
ciated with anxiety (Nolen-Hoeksema, 2000), which highlights that transdiagnostic
processes such as worry and rumination may maintain psychological distress.
Do Worry and Rumination Prolong Recovery From Stress? 4
The metacognitive model (Wells & Matthews, 1994) proposes that emotional
disorders of anxiety and depression are maintained by the activation of a type of cop-
ing or self-regulation dominated by worry, rumination and paradoxical coping
strategies that impair flexible cognitive control (Wells, 2000; Wells & Matthews,
1994, 1996). This style is called the Cognitive Attentional Syndrome (CAS), for ex-
ample, following stress individuals may try to reduce anxiety by anticipating threats
(worry) in order to avoid them or may analyse (ruminate) over their responses to try
and improve their coping. However, the CAS is associated with negative emotional
outcomes as it maintains a sense of threat and can delay the down-regulation of emo-
tional responses. Therefore, a prediction arising from the model is that repetitive neg-
ative thinking in the form of worry and/or rumination should prolong recovery from
stress-exposure.
Studies have demonstrated that worry and rumination not only have a psychological
impact on stress recovery but also impact physiological recovery. Ruscio et al (2015) found
that rumination was associated with greater NA and increased symptoms of depression and
anxiety following a stressful event. These results are in line with other studies highlighting
that individuals who engage in rumination following a stressful event are more likely to
experience low mood than those who do not use rumination (Driscoll, Lopez & Kistner,
2009; LeMoult, Kircanski, Prasad & Gotlib, 2017; Nolen-Hoeksema & Morrow, 1991;
Nolen-Hoeksema et al., 1993). In addition, engaging in worry following a stressful life event
has been associated with increased severity of stress symptoms three months later (Roussis &
Wells, 2008).
On a physiological level, both worry and rumination have been associated with
increased blood pressure and cortisol levels, as well as with lower heart rate variability
(Ottaviani et al., 2016). Recently, Papousek, Paechter, Weiss, and Lackner (2017) found that
Do Worry and Rumination Prolong Recovery From Stress? 5
a greater tendency to use rumination following a stressful situation prolonged elevations in
heart rate. These results are consistent with subsequent studies that found that rumination
prolongs cardiovascular recovery as well as salivary cortisol recovery following a stressor
(Key, Campbell, Bacon, & Gerin, 2008; LeMoult & Joormann, 2014, Shull et al., 2016;
Zoccola, Dickerson & Zaldivar, 2008; Zoccola, Quas & Yim, 2010).
Similar findings have been obtained on the physiological concomitant of worry.
Fisher and Newman (2013) evaluated three physiological indices in response to stress
following a worry induction across three groups (healthy controls, high trait worries, and
individuals with GAD). They found that although all groups experienced an increase in heart
rate and decreases in respiratory sinus arrhythmia, differences in salivary alpha amylase were
observed. Higher salivary alpha amylase at baseline predicted decreased change in salivary
amylase during stress in individuals with GAD, while greater baseline levels were associated
with greater change in healthy controls. Previous studies have also found an association
between increased cardiovascular indices and the use of worry in individuals with GAD and
those with high state and trait worry (Brosschot, Van Dijk, & Thayer, 2007; Delgado, Vila, &
Reyes del Paso, 2014; Knepp & Friedman, 2008).
Research evaluating emotion regulation strategies have primarily focused on
physiological outcomes such as cortisol sampling and cardiovascular indices
(parasympathetic response); as a result fewer studies have evaluated the sympathetic nervous
system response. Skin conductance can be used as a physiological measure of stress
reactions as it has been associated with the sympathetic nervous system which regulates a
variety of bodily functions and is associated with the fight or flight response. During
heightened emotional or physiological states the sympathetic nervous system activates the
cardiovascular and adrenal catecholamine systems (Jansen, Nguyen, Karpitsky, &
Mettenleiter, 1995). One measureable indicator of increased catecholamine levels is through
Do Worry and Rumination Prolong Recovery From Stress? 6
sweat gland secretions (Hansen & Sawatzky, 2008), measured using a galvanic skin response
(GSR) device, where greater skin conductance levels have been associated with
psychopathology (Pruneti, Lento, Fante, Carrozzo, & Fontana, 2010; Rosebrock, Hoxha,
Norris, Cacioppo, & Gollan, 2016). Pruneti et al (2010) compared autonomic arousal across
four physiological outcomes (skin conductance, heart rate, peripheral temperature, and
electromyography) in patients diagnosed with GAD, major depression, panic disorder or
obsessive-compulsive disorder across four time points (adaptation, rest, stress, and recovery).
Mean skin conductance levels were compared between groups and across time points, which
indicated that patients with GAD and patients with panic disorders had a greater reaction to
stress in comparison to patients with depression. Interestingly, Rosebrock et al (2016)
compared mean skin conductance and subjective arousal in individuals with anxiety,
depression, comorbid anxiety and depression, and healthy controls while viewing emotional
images. The authors found that all groups demonstrated increased skin conductance reactivity
to threatening images compared to positive, negative, and neutral stimuli. Surprisingly, few
studies on worry and rumination have focused on skin conductance responses. Steinfurth,
Alius, Wendt & Hamm (2017) compared physiological responses (i.e., heart rate and GSR)
in individuals who were asked to worry or ruminate about a keyword from a personally
meaningful topic. SCL were greater during worry and rumination than neutral thinking.
Further studies support these findings evidencing that engaging in worry and rumination is
associated with increased responses in this domain (Hoffman et al., 2005; Ottaviani et al,
2014; Sigmon, Dorhofer, Rohan & Boulard, 2000). Although studies have demonstrated an
association between increased SCL during worry and rumination, no previous studies have
evaluated skin conductance responses of worry and rumination within the context of a stress
paradigm. Therefore the current study aims to overcome this gap in the literature.
In summary, data are consistent with a prediction of the metacognitive model and
Do Worry and Rumination Prolong Recovery From Stress? 7
suggest that worry and rumination impact emotional regulation and prolong recovery from
natural or laboratory based stress exposures. However, so far studies have not compared the
effects of worry and rumination on stress recovery against a control mentation strategy on
subjective as well as physiological (sympathetic) indices of emotion (i.e., skin conductance)
and recovery in the same paradigm. As such, the present study set out to examine these
effects. We hypothesized that there would be a negative effect of worry and/or rumination on
recovery from a stressor when compared with a control condition. We did not know if this
would occur immediately or over a longer time interval but aimed to explore the time-course.
A 30-minute time interval was selected in order to remain consistent with previous lab based
evaluations of stress recovery (i.e. LeMoult, Arditte, D’Avanzato, Joormann, 2013; Shull, et
al., 2016). We hypothesized that worry and/or rumination would lead to delayed recovery
when compared to distraction, and that this effect would be observed in subjective and/or
physiological measures of symptoms. The design of the study enabled us to examine the
pattern of effects of worry and rumination against a control condition, which served to
control for the provision of a post-stress cognitive processing task of fixed duration and was
intended to prevent the occurrence of naturally occurring worry and rumination during that
time.
Methods
Participants
A total of 54 undergraduates (38 women, 16 men) from the University of Manchester
participated in the study. The mean age of the sample was 20.0 years (SD = 1.7, range: 18-26
years). 65% of participants were white British, 15% were Indian, 7% were Pakistani, while
the remaining 13% of participants were of varying ethnicities (e.g., Chinese, African,
Caribbean). Participants were recruited from the online experimental recruitment system,
Do Worry and Rumination Prolong Recovery From Stress? 8
which allows participants to receive course credit in exchange for experimental participation.
Participants were excluded from the experiment if they were not proficient enough in the
English language to follow task instructions and complete the study questionnaires. This
project was approved by the University of Manchester research ethics committee (no. 1, ref
13290).
Measures
Positive and Negative Affect Scale
The Positive and Negative Affect Scale (PANAS) is a brief self-report measure of mood
state (Watson, Clark, & Tellegen, 1988). It comprises 20-items consisting of two mood
scales, which measure both positive affect (for example, “excited,” “alert,” “interested.”) and
negative affect (for example, “ashamed,” “nervous,” “hostile.”). Participants respond on a 5-
point Likert-scale which ranges from 1 (very slightly or not at all) to 5 (extremely).
Participants completed the PANAS at six different time points throughout the experiment
(instructions: ‘‘Please rate each emotion to indicate how you are feeling at the moment”).
Only the NA subscale of the PANAS was used in subsequent analyses as the NA subscale of
the PANAS is a measure of subjective distress (an outcome predicted by the metacognitive
model) such that greater negative affect scores indicate greater levels of distress (Watson et
al., 1988).
Skin Conductance Levels
Skin conductance was recorded throughout the entire experiment. To measure skin
conductance a Neulog GSR logger sensor NUL-217 was used. Skin conductance was
recorded in microsiemens at 60 second intervals. Data was then sectioned into relevant time
periods (baseline, stress, post emotion regulation, 10, 20, and 30 minutes into the recovery
period) and mean values were calculated for each of the time intervals. Sensors were placed
Do Worry and Rumination Prolong Recovery From Stress? 9
at the base of participant’s non-dominant index and ring fingers, and participant’s skin was
prepared with an alcohol wipe and moistened with water using a cotton pad as suggested by
the developers. The lab used for testing consisted of a temperature controlled room at 23ºC.
Visual Analogue Scale (VAS).
A VAS was used after the emotion regulation induction to evaluate the extent to which
participants were engaging with the task (worry, rumination or distraction). Participants
completed an 11-point Likert scale item: “Rate how well were you able to stay on task,” with
responses ranging from 0 (did not stay on task) to 100 (stayed on task the entire time).
Materials
Video
A neutral ten-minute nature video was shown to participants as a baseline filler task
and to allow a baseline skin conductance measurement to be taken.
Stimuli
To induce different styles of thinking (worry and rumination) a list of prompts was
used. For the rumination condition, prompts focused participants’ attention on thoughts that
were emotion-, symptom- or self-focused. For example, “Why am I feeling the level of
motivation I feel right now?” Alternatively, participants assigned to the worry condition
received prompts that focused their attention on thoughts concerning the future, danger, and
threat. For example, “What if you were unable to maintain your current lifestyle (standard of
living)?” The rumination and worry stimuli were presented in the form of questions. The
rumination stimuli began with the question, “why?”, because rumination aims to analyze and
question the reason for events. In contrast, the worry stimuli began with the question, “what
if?”, because anxious thoughts, such as worry, represent future-oriented danger-related
Do Worry and Rumination Prolong Recovery From Stress? 10
thinking. The distraction group did not receive any written prompts; instead they were given
a set of simple filler tasks involving word search. The distraction condition was used as an
active control group, as having participants not engage in a task for 8 minutes would not
allow us to minimise the amount of worry or rumination that might spontaneously occur.
Moreover, we wanted to separate the effects of worry/rumination from the effects of simply
engaging in active cognitive processing.
Procedure
Once participants provided informed consent they were randomly assigned to one of
the experimental groups using an online research randomization software (Research
Randomizer Version 4.0). There was no stratification or control for factors used in the
randomization process.
On arrival at the laboratory the GSR sensors were attached to the base of the index and
ring fingers of participant’s non-dominant hand. Skin conductance recordings were obtained
throughout the duration of the experiment, a further description of how the recordings were
analysed is presented in the data analysis section. Participants first watched a ten-minute
nature video, after which, baseline SCL and PANAS ratings were obtained. Following the
baseline measurements participants began the Trier Social Stress Test (TSST) (Kirschbaum,
Pirke, & Hellhammer, 1993). The TSST requires participants to deliver a five- minute speech
on why they are the best candidate for their dream job, and received a standardized amount of
time to prepare for their speech. Participants were told that their speech would be video-taped
and shown to a group of their peers; however, the participants did not know that there was no
tape in the camera and thus their speech would not be evaluated. The use of deception was
explained to participants during the debriefing period after the experiment was completed.
Next, participants completed a five- minute mental arithmetic task. This task requires
participants to count backwards from 1022 in decrements of 13, while the experimenter
Do Worry and Rumination Prolong Recovery From Stress? 11
provided negative feedback such as tapping their fingers impatiently. Both the speech and
math tasks were completed in front of the experimenter who was trained in delivering the
TSST.
Following the stress induction participants completed their second PANAS rating,
following which participants then engaged in their allocated mentation task (worry,
rumination or distraction) which lasted for 8 minutes. As our hypotheses concerned the
impact of emotion regulation strategies on recovery from a stressor the emotion regulation
induction was conducted following the stressor. Participants in the worry and rumination
condition were asked to read and provide written responses to a set of 10 prompts. Prompts
were presented to participants using cue cards, which were presented to all participants in the
same order. Following the experimental manipulation participants provided PANAS ratings,
and completed the visual analogue scale to assess how well they stayed on task.
Finally, participants were asked to sit quietly for 30 minutes, during which three
additional PANAS ratings were obtained at 10- minute intervals. Following the rest period,
participants completed a demographics questionnaire and they were debriefed on the
deception in the experiment.
Data Analysis
Prior to study initiation a power calculation was conducted using G*Power (Faul,
Erdfelder, Lang & Buchner, 2007; Faul, Erdfelder, Buchner & Lang, 2009), which indicated
a sample size of 54 participants based on a medium effect size as indicated by Cohen’s
(1988) guidelines and with 90% power. Skin conductance levels were sectioned into six
intervals (baseline, stress, post emotion regulation, recovery 1, recovery 2, and recovery 3)
and mean values were calculated for each interval. Based on the recommendation by Dixon
and Tukey (1968), outliers were Winsorized, meaning that they were replaced with less
extreme values in order to improve statistical power. Outliers (9% of the data) were replaced
Do Worry and Rumination Prolong Recovery From Stress? 12
with values two standard deviations above the mean.
To assess the overall pattern of rate of change between groups a generalized
estimating equation (GEE) analysis, fitting a linear trend in time, was conducted over the
recovery period (post emotion regulation, 10, 20, and 30 minutes into the recovery period).
Analyses focused on the period following the stress induction as the hypotheses aimed to
evaluate the impact of emotion regulation strategies on recovery from stress. If a significant
interaction was found, follow-up tests involving repeated measures analyses of covariance’s
(ANCOVA) were conducted on the recovery period. The post-stress induction scores were
used as a covariate. F values from the ANCOVA are not reported as the parameter estimates
represent the differences between groups. The parameter estimates describe the average
difference between groups at each time point, with positive values indicating that a group is
above the comparison group, distraction, thus indicating higher values, whereas negative
values indicate that a group is below the comparison group, indicating lower values. These
parameter estimates are shown in Table 2. Repeated measures t-tests were conducted to
evaluate the differences within groups on skin conductance and NA. Each recovery time
point was compared to the stress point.
Results
Participant Characteristics
The demographic characteristics of participants are presented in Table 1. The groups
did not differ in age; F(2,50) = 0.36, p = 0.70, gender; χ2(2, N = 54) = 0.18, p = 0.92, or body
weight; F(2,47) = 46.92, p = 0.72. There were no significant differences between groups in
alcohol use; χ2(2, N = 54) = 0.64, p = 0.73, amount of alcohol intake; F(2,40) = 10.68, p =
0.35, if they participated in any exercise; χ2(2, N = 53) = 0.42, p = 0.81, caffeine use; χ2(2, N
= 54) = 0.64, p = 0.73, or nicotine use; χ2(2, N = 52) = 1.25, p = 0.54. There was no
Do Worry and Rumination Prolong Recovery From Stress? 13
significant difference between groups on diagnosed health conditions; χ2(2, N = 54) = 2.12, p
= 0.35. Three participants reported a health condition including anaemia, an overactive
thyroid, and asthma. There was no significant differences between groups in medication use;
χ2(2, N = 53) = 1.87, p = 0.39, although eight participants reported taking medications such as
paracetamol. In order to evaluate if groups differed on baseline levels of NA and skin
conductance a one-way ANOVA was conducted. A significant between group difference
was demonstrated for baseline levels of NA, F(2, 50) = 7.53, p = 0.001, however there was no
difference between groups on baseline levels of skin conductance, F(2, 51) = 0.04, p = 0.96.
In order to control for baseline differences between groups, baseline levels of NA and skin
conductance were used as a covariate in subsequent analyses.
Manipulation Check
Trier Social Stress Test
To ensure that the TSST elicited stress a repeated measures t-test was conducted on the
baseline and post stressor negative affect values as these were administered directly before
and after the stress induction and before the mentation task. The TSST elicited a significant
amount of stress in the worry group, t(16) = 3.77, p = 0.002, Cohen’s d = 1.11, rumination
group, t(17) = 5.47, p < 0.001, Cohen’s d = 1.60, and control (distraction) group, t(17) = 5.06,
p < 0.001, Cohen’s d = 1.70. Stress levels immediately after the TSST were similar across the
groups, F (2, 51) = 0.09, p = 0.91, ηp2 = 0.004, and can be seen in Figure 1 and Table 1.
Emotion Regulation Induction
A visual analogue scale (manipulation check) was administered after the emotion
regulation induction to evaluate if participants remained on task. A one-way ANOVA
demonstrated no significant difference between groups in ability to engage in the task F(2,51)
= 0.35, p = 0.70. Descriptive statistics for the VAS presented by group are displayed in Table
1, which demonstrates that individuals in each condition were able to comply with the
Do Worry and Rumination Prolong Recovery From Stress? 14
manipulation 76.9-81.1% of the time.
Skin Conductance
GEE analyses revealed a borderline significant group by time interaction [Wald χ2=
5.9 p = 0.053] which indicated that there was some evidence of a difference in recovery
between the three groups.
Follow-up tests to assess group differences demonstrated that there was no significant
difference between worry and distraction (control group) in SCL at any of the time points in
the recovery period. Examination of parameter estimates show that differences in the worry
condition ranged from 0.24 microsiemens to 0.11 microsiemens. However, in contrast the ru-
mination condition did not exhibit recovery; increasing parameter values indicate a worsen-
ing of response. The largest difference occurred at the final time point in the recovery period,
where the average difference between the rumination group and the distraction group was
1.54 microsiemens, p = 0.01. A one-way ANOVA was conducted to evaluate the effect size
on the final time point, F(1,34) = 4.66, p = 0.04, η2 = 0.12. The eta-squared (η2) value is indic-
ative of a large effect size as outlined by Cohen’s (1988) guidelines for eta squared. These
results are illustrated in Figure 1, which shows that immediately after the mentation manipu-
lation the worry group showed a reduction in SCL whilst the rumination group showed an in-
crease when compared to the effect post-stress.
Post-hoc tests were conducted to evaluate within group differences in recovery. For
the worry group there was no significant differences between stress and post-emotion regula-
tion, t(17) = 1.32, p = 0.21, however there was a significant difference between stress and re-
covery time point 1, t(17) = 2.71, p = 0.02, stress and recovery time point 2, t(17) = 2.89, p =
0.01, and stress and recovery time point 3, t(17) = 3.02, p = 0.008. The rumination group did
not exhibit any significant differences between stress and any of the subsequent time points.
These results are consistent with the results of the GEE analyses which indicated that rumina-
Do Worry and Rumination Prolong Recovery From Stress? 15
tion appeared to prolong recovery from stress. The distraction group did not exhibit a signi-
ficant difference between stress and post-emotion regulation, t(17) = 0.51, p = 0.61, however
there was a borderline significant difference between stress and recovery time point 1, t(17) =
2.12, p = 0.05, and a significant difference between stress and recovery time point 2, t(17) =
2.45, p = 0.03, and stress and recovery time point 3, t(17) = 2.35, p = 0.03.
In summary, on physiological indices of stress, rumination appeared to prolong recov-
ery, whereas worry appeared to produce a pattern of recovery that was similar to the distrac-
tion group.
Negative Affect
GEE analyses demonstrated a significant group by time interaction [Wald χ2= 11.7 p
= 0.003]. The pattern of recovery for the worry condition following the emotion regulation
induction demonstrated greater NA than the rumination or distraction conditions, suggesting
that worry had an immediate delay on recovery for NA, this outcome can be seen in Figure 1.
Follow-up tests based on parameter estimates showed that worry appeared to have an
immediate effect of maintaining negative affect levels following the emotion regulation
induction, b = 4.67, p = 0.01. At this point individuals in the worry group were recovering
more slowly than the distraction condition at 4.67 points on the PANAS. This negative effect
does not persist and with time the worry group caught up with the other groups. In order to
evaluate the magnitude of this effect a one-way ANOVA was conducted, which demonstrated
a medium-large effect, F(1,31) = 7.08, p = 0.01, η2 = 0.19. The rumination condition showed
a similar recovery to the distraction group. The results indicate that worry delayed recovery
on NA immediately following stress induction however rumination showed little effect on the
recovery of negative affect.
Do Worry and Rumination Prolong Recovery From Stress? 16
Post-hoc within group t-tests were conducted between stress and all subsequent time
points to further evaluate within group differences in recovery from stress on negative affect.
The worry group demonstrated a significant difference between stress and all subsequent time
points. There was a significant difference between stress and post emotion regulation, t(17) =
2.91, p = 0.01, stress and recovery 1, t(17) = 5.70, p <0.001, stress and recovery 2, t(17) =
6.06, p <0.001, and stress and recovery 3, t(17) = 6.28, p < 0.001. The rumination group
demonstrated a significant difference between stress and post-emotion regulation, t(17) =
5.27, p <0.001, stress and recovery 1, t(17) = 5.70, p <0.001, stress and recovery 2, t(17) =
6.32, p <0.001, and stress and recovery 3, t(17) = 6.34, p <0.001. Similarly, the distraction
group also exhibited significant decreases in self-reported negative affect across all time
points. There was a significant decrease between stress and post-emotion regulation, t(17) =
4.08, p = 0.001, stress and recovery 1, t(17) = 4.64, p <0.001, stress and recovery 2, t(17) =
4.16, p =0.001, and stress and recovery 3, t(17) = 4.20, p = 0.001. These results indicate that
all groups exhibited significant improvements in negative affect at each recovery point
relative to the point of stress exposure.
Discussion
This study compared the effect of mentation strategies (worry and rumination) that are
linked to delayed stress recovery in the metacognitive model (Wells & Matthews, 1996), with
a control condition of distraction. We hypothesized that individuals in the worry and/or
rumination conditions would have a prolonged trajectory in physiological and/or affective
recovery in comparison to individuals in the distraction condition. This is because worry and
rumination are emotion regulation strategies that maintain processing of threat and negative
mood states, which in the metacognitive model prolong subcortical emotion processing. In
manipulating worry and rumination directly, we were able to explore more specific patterns
Do Worry and Rumination Prolong Recovery From Stress? 17
of effects of these strategies on objective and physiological stress responses. Additionally, we
were able to examine the immediate and more protracted effects on recovery.
In line with our hypotheses rumination appeared to cause a delay in recovery from
stress in comparison to distraction. This effect appeared on physiological (SCL) but not self-
report indices (NA). Furthermore, this effect appeared to be greatest 30 minutes into the
recovery period. Thus, it appears that rumination may have a delayed impact, delaying
physiological recovery to a greater extent as time since stress exposure elapses. This effect
was not observed in connection with worrying. Whilst SCL for the worry and distraction
groups did not return to baseline levels, they did trend towards returning to baseline levels,
and the small continued elevation may be due to the study being in a lab-based environment.
In contrast, self-report NA data revealed a somewhat different pattern. The worry
group displayed delayed recovery from stress that was more immediate and occurred soon
after the worry induction. The rumination manipulation did not appear to impact on self-
report NA at any phase.
These results support our initial hypotheses that repetitive negative thinking in the
form of worry or rumination delays the rate of recovery from exposure to stress. The
inclusion of a control condition means that the effect is unlikely to be a non-specific effect of
cognitive processing. Furthermore, the difference in response effect manifested by worry and
rumination groups supports the interpretation that we are assessing effects of these different
forms of mentation rather than non-specific factors or the effect of a cognitive task.
A systematic review and meta-analysis by Ottaviani et al (2016) highlighted that heart
rate effects were greater in studies that focused on worry rather than rumination. These
findings are in contrast to the results of the current study that found that rumination elicited a
greater physiological response than worry. The discrepancy in these findings could be due to
differences in autonomic nervous system responses. As Ottaviani et al (2016) were not able
Do Worry and Rumination Prolong Recovery From Stress? 18
to evaluate moderator effects for all physiological responses (heart rate, blood pressure, and
cortisol) it may be that not all autonomic nervous system indicators show the same effect,
therefore future research on physiological responses is required.
Worry and rumination are the underlying emotion regulation strategies typically
associated with the maintenance of anxiety and depression, respectively. The results add
further evidence on how these processes may differentially impact psychological and
physiological indices of stress response. The data highlights a discrepancy between self-
report measures and physiological measures and argues for the importance of using a multi-
method approach in evaluating the impact of thinking styles. Discrepancies between self-
report measures and objective measures have previously been reported. For example, a study
by Nater et al (2010) examined psychological and physiological responses to stress in
individuals with borderline personality disorder (BPD). They found that on self-report
measures individuals with BPD reported greater and more threatening levels of acute stress in
comparison to healthy controls. However, cortisol responses for BPD patients indicated that
they had decreased cortisol and alpha-amylase levels at baseline and in response to stress in
comparison. Perhaps data of this kind is a marker for the preponderance of worry based
repetitive negative thinking rather than rumination in these individuals. Future research may
benefit the field by determining if there are response signatures that accompany different
varieties of repetitive negative thinking as the present results might suggest.
Although we did not predict specific differences between worry and rumination, the
identification of early effects of worry on self-report affect but later-effects of rumination on
physiological responses deserves some consideration. These differences may reflect
functional dimensions of these different types of thinking. Specifically, worry is normally
used to anticipate danger and prepare for action and it is likely to exert effects within a short
time-frame as the internal anxiety program for avoidance and vigilance is activated.
Do Worry and Rumination Prolong Recovery From Stress? 19
However, rumination is past oriented with little motivation to detect, avoid, or deal with
danger in the present or future. Thus, it is more likely to call reflective memory-based activity
which could paradoxically delay later stages of stress recovery that might be more dependent
on memory-based processes.
The effect of worry and rumination found in the study adds further evidence to the
proposal that perseverative thinking styles following stress exposure impacts on symptoms
(Nolen-Hoeksema, 1991; Wells & Papageorgiou, 1995). Although we did not assess the
potential underlying mechanism by which these processes may transmit such effects, the
results provide further support for the metacognitive model that gives a central role to worry
and rumination (Wells & Matthews, 1994; 1996) in the development of chronic stress
symptoms.
The clinical implications of the present findings must remain tentative, but the data
suggests that interventions that directly reduce the propensity to self-regulate with strategies
of worry and rumination may improve immediate and/or delayed reactions to stress. The data
adds to the literature on traumatic stress reactions that has shown that the tendency to use
worry to deal with negative thoughts soon after negative life events increases the severity of
symptoms later on (Holeva, Tarrier, & Wells, 2001; Warda & Bryant, 1998; Roussis &
Wells, 2008). Furthermore, metacognitive therapy aimed at reducing worry and rumination
appears effective in treating prolonged traumatic stress responses (Wells & Colbear, 2012;
Wells, Walton, Lovell, & Proctor, 2015) suggesting that a greater focus on regulating worry
and rumination could provide clinical benefits and potentiate stress recovery.
Future studies should examine the recovery period in greater detail. We noticed
patterns of response that appear time-dependent, thus future studies could extend the range of
monitoring to explore the longer-term effect and determine when symptom parameters return
Do Worry and Rumination Prolong Recovery From Stress? 20
to baseline. Additionally, future studies should examine effects using a range of indices of
stress such as cortisol levels and cardiovascular recovery.
The present study is limited by its sample which consisted of undergraduate
participants, thus decreasing the generalizability of the findings. Furthermore, other research
has demonstrated age differences in emotion regulation (Orgeta, 2009). Thus, it is uncertain
whether rumination, worry, or distraction would function similarly in an older adult or
younger population. In addition, we did not measure trait worry or rumination in order to
avoid potentially priming participants by completing measures of worry and rumination prior
to group assignment, but this means it is unclear if participants differed on these dimensions
at the outset. In addition, we did not control for amount of time since caffeine intake prior to
the experiment. The duration of our recovery period was short so we could not examine the
long term effect of worry and rumination. Finally, there was a difference in the pronouns
used for the worry and rumination prompts in the current study. The pronoun “I” was used in
the rumination prompts and “you” was used in worry. We did not want to change the prompts
as these have been used in previous studies of rumination and worry induction, but we
acknowledge that this inequality may account for some of the differences observed. Our
study did not set out to test for statistically significant differences between worry and
rumination and was not powered to do so, but this is clearly an area for future studies.
In conclusion, we observed that worry and rumination each had a negative but
specific effect on recovery from stress compared to a control condition; however long term
effects require further examination.
References
Borkovec, T.D., Alcaine, O., & Behar, E.S. (2004). Avoidance theory of worry and
generalized anxiety disorder. In Heimberg, R., Mennin, D., Turk, C.,
Do Worry and Rumination Prolong Recovery From Stress? 21
(Eds). Generalized anxiety disorder: Advances in research and practice, (pp. 77-
108). New York: Guilford.
Borkovec, T. D., Robinson, E., Pruzinsky, T., & DePree, J. A. (1983). Preliminary
exploration of worry: some characteristics and processes. Behaviour
Research and Therapy, 21, 9–16. doi: 10.1016/0005-7967(83)90121-3
Brosschot, J. F., Van Dijk, E., & Thayer, J. F. (2007). Daily worry is related to low heart rate
variability during waking and the subsequent nocturnal sleep period. International
Journal of Psychophysiology, 63, 39–47.
doi: 10.1016/j.ijpsycho.2006.07.016
Cohen J. E. (1988). Statistical Power Analysis for the Behavioral Sciences. New York, NY:
Routledge Academic.
Delgado, L. C., Vila, J., & Reyes del Paso, G. A. (2014). Proneness to worry is negatively
associated with blood pressure and baroreflex sensitivity: Further evidence of the
blood pressure emotional dampening hypothesis. Biological Psychology, 96, 20–27.
doi: 10 .1016/j.biopsycho.2013.11.005
Dixon, W. J., & Tukey, J. W. (1968). Approximate Behavior of the Distribution of
Winsorized t (Trimming/Winsorization 2). Technometrics, 10, 83–98.
doi: 10.2307/1266226
Driscoll, K.A., Lopez, C.M., Kistner, J.A. (2009). A diathesis-stress test of response styles in
children. Journal of Social and Clinical Psychology, 28,1050–1070. doi:
10.1521/jscp.2009.28.8.1050.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical
power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39, 175-191. doi:10.3758/BF03193146
Do Worry and Rumination Prolong Recovery From Stress? 22
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using
G*Power 3.1: Tests for correlation and regression analyses. Behavior Research
Methods, 41, 1149-1160. doi: 10.3758/BRM.41.4.1149
Fisher, A. J., & Newman, M. G. (2013). Heart rate and autonomic response to stress after
experimental induction of worry versus relaxation in healthy, high-worry, and
generalized anxiety disorder individuals. Biological Psychology, 93, 65–74. doi:
10.1016/j.biopsycho.2013.01.012
Hansen, F., & Sawatzky, J. V. (2008). Stress in Patients With Lung Cancer: A Human
Response to Illness. Oncology Nursing Forum, 35, 217-223. doi:
10.1188/08.ONF.217-223
Hofmann, S.G.,Moscovitch, D.A., Litz, B.T., Kim, H.J., Davis, L.L., & Pizzagalli, D.A.
(2005). The worried mind: autonomic and prefrontal activation during worrying.
Emotion, 5, 464–475. doi: 10.1037/1528-3542.5.4.464
Holeva, V., Tarrier, N., & Wells, A. (2001). Prevalence and predictors of acute stress
disorder and PTSD following road traffic accidents: Thought control strategies
and social support. Behavior Therapy, 32, 65-83. doi: 10.1016/S0005-
7894(01)80044-7
Jansen, A., Nguyen, X., Karpitsky, V., & Mettenleiter, M. (1995). Central command neurons
of the sympathetic nervous system: Basis of the fight-or-flight response. Sci Mag,
5236, 644–646. doi: 10.1126/science.270.5236.644
Key, B.L., Campbell, T.S., Bacon, S.L., Gerin, W. (2008). The influence of trait and state
rumination on cardiovascular recovery from a negative emotional stressor. Journal of
Behaviour Medicine, 31(3), 237-248. doi: 10.1007/s10865-008-9152-9
Do Worry and Rumination Prolong Recovery From Stress? 23
Kirschbaum, C., Pirke, K. M., & Hellhammer, D. H. (1993). The “Trier Social Stress
Test”—A tool for investigating psychobiological stress responses in a
laboratory setting. Neuropsychobiology, 28, 76–81.
doi: 10.1159/000119004
Knepp, M. M., & Friedman, B. H. (2008). Cardiovascular activity during laboratory tasks in
women with high and low worry. Biological Psychology, 79, 287–293.
doi: 10.1016/j.biopsycho.2008.07 .002
LeMoult, J., Arditte, K.A., D’Avanzato, C., & Joormann, J. (2013). State Rumination:
Associations with Emotional Stress Reactivity and Attention Biases. Journal of
Experimental Psychopathology, 4(5), 471-484. doi: 10.5127/jep.029112
LeMoult, J., & Joormann, J. (2014). Depressive Rumination Alters Cortisol Decline in Major
Depressive Disorder. Biological Psychology, 100, 50-55.
doi: 10.1016/j.biopsycho.2014.05.001
LeMoult, J., Kircanski, K., Prasad, G., & Gotlib, I H. (2017). Negative self-referential
processing predicts the recurrence of major depressive episodes. Clinical
Psychological Science, 5, 174-181. doi: 10.1177/2167702616654898
Nater, U.M., Bohus, M., Abbruzzese, E., Ditzen, B., Gaab, J., et al. (2010). Increased
Psychological and Attenuated Cortisol and Alpha-Amylase Response to Acute
Psychosocial Stress in Female Patients with Borderline Personality Disorder.
Psychoneuroendocrinology, 35, 1565-1572. doi: 10.1016/j.psyneuen.2010.06.002
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the
duration of depressive episodes. Journal of abnormal psychology, 100,
569–82. doi: 10.1037/0021-843X.100.4.569
Do Worry and Rumination Prolong Recovery From Stress? 24
Nolen-Hoeksema, S., (2000). The role of rumination in depressive disorders and
mixed anxiety/depressive symptoms. Journal of Abnormal Psychology, 109,
504–511. doi: 10.1037/0021-843X.109.3.50
Nolen-Hoeksema, S. & Morrow, J. (1991). A prospective study of depression and
posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta
earthquake. Journal of Personality and Social Psychology, 61, 115–121.
doi: 10.1016/j.psyneuen.2010.06.002
Nolen-Hoeksema, S., Morrow, J., & Fredrickson, B. L. (1993). Response styles and the
duration of episodes of depressed mood. Journal of Abnormal Psychology, 102, 20–
28. doi: 10.1037/0021-843X.102.1.20
Orgeta, V. (2009). Specificity of age differences in emotion regulation. Aging &
Mental Health, 13(6), 818–826. doi: 10.1080/13607860902989661.
Ottaviani C., Borlimi R., Brighetti G., Caselli G., Favaretto E., Giardini I., et al. (2014).
Worry as an adaptive avoidance strategy in healthy controls but not in pathological
worriers. International Journal of Psychophysiology, 93, 349–355.
doi: 10.1016/j.ijpsycho.2014.05.010.
Ottaviani, C., Thayer, J.F., Verkuil, B., Lonigro, A., Medea, B., Couyoumdjian, A., &
Brosschot, J.F. (2016). Physiological Concomitant of Perseverative Cognition: A
Systematic Review and Meta-Analysis. Psychological Bulletin, 142, 231-259.
doi: 10.1037/bul0000036
Papousek, I., Paechter, M., Weiss, E.M., & Lackner, H.K. (2017). The tendency to ruminate
and the dynamics of heart rate recovery after an ordinary, mildly stressful
performance situation. Personality and Individual Differences, 104, 150-154.
doi: 10.1016/j.paid.2016.08.003
Do Worry and Rumination Prolong Recovery From Stress? 25
Pruneti, C. A., Lento, R. M., Fante, C., Carrozzo, E., & Fontana, F. (2010). Autonomic
arousal and differential diagnosis in clinical psychology and psychopathology.
Giornale Italiano di Psicopatologia, 16, 43–52.
Rosebrock, L. E., Hoxha, D., Norris, C., Cacioppo, J. T., & Gollan, J. K. (2016). Skin
Conductance and Subjective Arousal in Anxiety, Depression, and Comorbidity:
Implications for Affective Reactivity. Journal of Psychophysiology.
doi: 10.1027/0269-8803/a000176
Roussis, P., & Wells, A. (2008). Psychological factors predicting stress symptoms:
metacognition, thought control, and varieties of worry. Anxiety, Stress, and
Coping, 21(3), 213–225. doi: 10.1080/10615800801889600.
Ruscio, A. M., Gentes, E. L., Jones, J. D., Hallion, L. S., Coleman, E. S., & Swendsen, J.
(2015). Rumination Predicts Heightened Responding to Stressful Life Events in
Major Depressive Disorder and Generalized Anxiety Disorder. Journal of Abnormal
Psychology, 124, 17–26. doi: 10.1037/abn0000025
Shull, A., Mayer, S.E., McGinnis, E., Geiss, E., Vargas, I., & Lopez-Duran, N.L. (2016).Trait
and state rumination interact to prolong cortisol activation to psychosocial stress in
females. Psychoneuroendocrinology, 74, 324-332.
doi: 10.1016/j.psyneuen.2016.09.004.
Sigmon, S. T., Dorhofer, D. M., Rohan, K. J., & Boulard, N. E. (2000).The impact of anxiety
sensitivity, bodily expectations, and cultural beliefs on menstrual symptom reporting:
Do Worry and Rumination Prolong Recovery From Stress? 26
A test of the menstrual reactivity hypothesis. Journal of Anxiety Disorders, 14,
615–633. doi: 10.1016/S0887-6185(00)00054-2
Steinfurth, E. C. K., Alius, M. G., Wendt, J. and Hamm, A. O. (2017). Physiological and
neural correlates of worry and rumination: Support for the contrast avoidance model
of worry. Psychophysiology, 54, 161–171. doi:10.1111/psyp.12767
Warda, G., & Bryant, R. A. (1998). Thought control strategies in acute stress
disorder. Behaviour Research and Therapy, 36, 1171–1175.
doi: 10.1016/S0005-7967(98)00082-5
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measure of positive and negative affect: The PANAS scales. Journal of Personality
and Social Psychology, 55(1), 128-141.
doi: 10.1037/0022-3514.54.6.1063
Wells, A. (2000). Emotional disorders and metacognition: innovative cognitive therapy.
Chichester: Wiley.
Wells, A. (2009). Metacognitive therapy for anxiety and depression. New York: Guilford
Press.
Wells, A., & Colbear, J. S. (2012). Treating Posttraumatic Stress Disorder With
Metacognitive Therapy: A Preliminary Controlled Trial. Journal of Clinical
Psychology, 68, 373–381. doi: 10.1002/jclp.20871
Wells, A., Matthews, G. (1994). Attention and emotion: A clinical perspective. Hove:
Erlbaum.
Do Worry and Rumination Prolong Recovery From Stress? 27
Wells, A., Matthews, G. (1996). Modelling Cognition in Emotional Disorder: The SREF
MODEL. Behaviour Research Therapy, 34, 881–888.
doi: 10.1016/S0005-7967(96)00050-2
Wells A, Papageorgiou C. (1995). Worry and the incubation of intrusive images
following stress. Behaviour Research and Therapy, 33, 579-583.
doi: 10.1016/0005-7967(94)00087-Z
Wells, A., Walton, D., Lovell,K. & Proctor, D. (2015). Metacognitive therapy versus
prolonged exposure in adults with chronic post-traumatic stress disorder: A
parallel randomized controlled trial. Cognitive Therapy and Research, 39,
70-80. doi: 10.1007/s10608-014-9636-6
Zoccola P.M., Dickerson S.S., & Zaldivar F.P. (2008). Rumination and cortisol
responses to laboratory stressors. Psychosomatic Medicine, 70, 661–667.
doi: 10.1097/PSY.0b013e31817bbc77.
Zoccola, P. M., Quas, J. a, & Yim, I. S. (2010). Salivary cortisol responses to a
psychosocial laboratory stressor and later verbal recall of the stressor: The
role of trait and state rumination. Stress, 13, 435–43.
doi: 10.3109/10253891003713765.
Do Worry and Rumination Prolong Recovery From Stress? 28
Tables
Table 1. Participant Characteristics
Variable Worry(N=18)
Rumination(N=18)
Distraction(N=18)
Age, M(SD) 19.72 (0.89) 20.11 (1.79) 20.18 (2.27)
Sex (Female : Male) 13:5 12:6 13:5
Weight (kg), M(SD) 62.32 (13.76) 59.40 (13.42) 62.47 (7.15)
Alcohol Use (Yes :No) 14:4 13:5 16:2
Alcohol Intake (units) M(SD)
6.14(3.39) 5.85(2.51) 4.56(3.37)
Exercise (Yes :No) 13:5 13:4 12:6
Health Conditions (Yes :No) 0:18 2:16 1:17
Medications (Yes :No) 3:15 1:16 4:14
Caffeine Use (Yes :No) 8:10 6:12 6:12
Manipulation Check, M(SD)
77.35 (17.69) 81.11 (12.43) 76.94 (22.76)
NA Baseline, M(SD) 14.94 (4.13) 12.06 (2.13) 11.22 (1.59)
NA Post-Stress, M(SD) 20.59 (5.98) 19.50 (6.40) 20.33 (7.92)
NA Post-Emotion Regulation Induction, M(SD)
16.94 (5.74) 13.67 (3.79) 13.83 (4.08)
NA Recovery 1, M(SD) 13.59 (3.78) 12.83 (2.94) 12.83 (4.45)
NA Recovery 2, M(SD) 12.10 (2.76) 11.94 (1.63) 13.12 (3.76)
NA Recovery 3, M(SD) 10.76 (1.35) 11.50 (1.46) 12.89 (2.91)
SCL Baseline, M(SD) 3.65 (2.80) 3.64 (3.09) 3.40 (2.94)
SCL Post-Stress, M(SD) 7.06 (3.13) 6.25 (2.50) 5.40 (3.44)
SCL Post-Emotion Regulation Induction, M(SD)
6.80 (3.31) 6.78 (2.58) 5.33 (3.54)
SCL Recovery 1, M(SD) 6.29 (3.37) 6.61 (2.82) 4.95 (3.44)
Do Worry and Rumination Prolong Recovery From Stress? 29
SCL Recovery 2, M(SD) 6.14 (3.50) 6.73 (2.82) 4.73 (3.50)
SCL Recovery 3, M(SD) 5.96 (3.32) 6.91 (2.89) 4.63 (3.44)
Note. M = Mean, SD = Standard Deviation, NA = Negative Affect, SCL = Skin Conductance Levels
Table 2. Parameter Estimates: Skin Conductance & Negative Affect Time Point Group b pSkin Conductance Post Emotion Regulation Induction
Worry -0.15 0.67Rumination 0.62 0.08
Recovery 1Worry -0.24 0.61Rumination 0.85 0.07
Recovery 2Worry -0.17 0.75Rumination 1.21 0.03
Recovery 3Worry -0.11 0.86Rumination 1.54 0.01
Negative AffectPost Emotion Regulation Induction Worry 4.67 0.01
Rumination 0.14 0.94Recovery 1
Worry 0.96 0.52Rumination 0.98 0.51
Recovery 2Worry -0.19 0.31Rumination -1.51 0.20
Recovery 3Worry -2.21 0.03Rumination -1.56 0.12
Note: The parameter estimates are differences with the comparison condition (distraction group)
Do Worry and Rumination Prolong Recovery From Stress? 30
Figures
Figure 1. Mean Skin Conductance Levels and Negative Affect Over Time
INSERT FIGURE 1 HERE
Note: Vertical line = Time point of emotion regulation induction, * = significant between
group difference, a = significant within group difference between indicated time point and
stress