This article was downloaded by: [University of Saskatchewan Library]

On: 17 October 2012, At: 15:54

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Anxiety, Stress & Coping: An

International Journal
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/gasc20

Psychological factors predicting stress

symptoms: Metacognition, thought
control, and varieties of worry
a a
Panagiotis Roussis & Adrian Wells
a
Academic Division of Clinical Psychology, University of
Manchester, Manchester, UK

Version of record first published: 18 Aug 2009.

To cite this article: Panagiotis Roussis & Adrian Wells (2008): Psychological factors predicting stress
symptoms: Metacognition, thought control, and varieties of worry, Anxiety, Stress & Coping: An
International Journal, 21:3, 213-225

To link to this article: http://dx.doi.org/10.1080/10615800801889600

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-

conditions

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.
The publisher does not give any warranty express or implied or make any representation
that the contents will be complete or accurate or up to date. The accuracy of any
instructions, formulae, and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.
 Anxiety, Stress, & Coping,
July 2008; 21(3): 213
225

Psychological factors predicting stress symptoms:

Metacognition, thought control, and varieties of worry

PANAGIOTIS ROUSSIS, & ADRIAN WELLS

Academic Division of Clinical Psychology, University of Manchester, Manchester, UK

Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Abstract
This study tested predictions of the metacognitive model of post-traumatic stress disorder (Wells,
2000) concerning relationships between stress symptoms and maladaptive control of thoughts. We
tested the hypothesis that the tendency to use worry as a means of controlling thoughts will be
positively predictive of stress symptoms. In doing so, overlaps with symptomatic worry, stress
symptoms, and stress exposure at time 1 was controlled. One hundred and ten (110) college students
were assessed twice, over a three-month period. Consistent with predictions, use of worry as a thought
control strategy made a significant and independent contribution to the prediction of stress symptoms.
The findings provide further support for Wells’ metacognitive model of stress reactions.

Keywords: metacognition, stress symptoms, worry, thought control strategies

Introduction
Chronic and excessive worry is a common factor across anxiety disorders (Barlow, 1988)
and one major element of post-trauma adaptation. This clinical phenomenon has been
documented in post-trauma populations, often in conjunction with post-traumatic stress
disorder (PTSD). Kessler, Sonnega, Bromet, Hughes, and Nelson (1995) found that
pathological worry in the form of generalized anxiety was present in 16% of individuals
diagnosed with PTSD. This percentage is significantly higher than the 10% prevalence of
generalized anxiety in non-PTSD individuals (Kessler et al., 1995). In another study,
Shore, Tatum, and Vollmer (1986) found that worry was one of the most prevalent
symptoms among survivors of a natural disaster. Even in an analogue population,
individuals with traumatic exposure reported significantly higher levels of symptomatic
worry than did those who denied any previous exposure to stress (Roemer, Molina, Litz, &
Borkovec, 1993).
The relationship between trauma and worry has been investigated in people with
generalized anxiety disorder (GAD). Studies have looked at the prevalence of potentially
traumatizing events among individuals diagnosed with GAD. Higher rates of traumatic
exposure were reported by analogue and clinical GAD individuals compared to non-
anxious people (Roemer, Litz, Pepper, & Borkovec, 1995). These studies indicate that a
history of trauma and worry are likely to co-occur, and perhaps be interrelated. In order to

Correspondence: Adrian Wells, Academic Division of Clinical Psychology Rawnsley Building Manchester Royal
Infirmary, University of Manchester, Manchester M13 9WL, United Kingdom. E-mail: adrian.wells@
manchester.ac.uk

ISSN 1061-5806 print/ISSN 1477-2205 online # 2008 Taylor & Francis

DOI: 10.1080/10615800801889600
 214 P. Roussis and A. Wells

further understand the relationship between trauma exposure and worry it is important to
have a theoretical framework in which to study the function of worry in trauma victims.
Theories of worry propose that avoidance plays a primary role in the maintenance of
anxious apprehension (Borkovec, 1994; Borkovec, Alcaine, & Behar, 2001). Laboratory
studies have demonstrated that worrying about a feared event prior to imagining the event
attenuates autonomic arousal associated with the imagery (Borkovec & Hu, 1990). Thus,
worry is thought to suppress the processing of emotional information, which by itself
maintains anxiety disorder (Borkovec & Lyonfields, 1993). In addition, studies investigating
the perceived functions of worry found that compared to non-worriers worriers tended to
endorse more the belief that worrying allows them to distract themselves from emotional
material (Borkovec & Roemer, 1995). The resulting reduction of anxious arousal is thought
to negatively reinforce worry, thus making it chronic and difficult to control (Roemer &
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Borkovec, 1993).
The strategic role of worry as an attempt to control unwanted experiences has been
implicated in recent theories of psychopathology and PTSD. The S-REF model of clinical
disorder (Wells & Matthews, 1996) and the metacognitive model of PTSD (Wells, 2000)
propose that emotional processing of trauma can be successful only to the extent that
maladaptive self-knowledge and strategies for controlling one’s thought processes are
revised and replaced with new, more adaptive ones. Normally, this process is automatically
initiated by intrusive stress symptoms, such as images and ends with the development of a
coping plan which guides cognition and action in future encounters with threat. Due to its
spontaneous and automatic nature, it is termed ‘‘the Reflexive Adaptation Process’’ (RAP).
Whether the RAP unfolds unimpeded or not depends in part on the strategies used by
individuals to control their post-stress reactions. The use of worry as a thought control
strategy is held in metacognitive theory (Wells, 2000; Wells & Matthews, 1994) to interfere
with successful emotional processing. It focuses the person’s attention on threat-related
stimuli, and depletes cognitive resources necessary to establish flexible control over thinking
and return cognition to its normal threat-free status. Under these circumstances, the person
is unable to respond flexibly to intrusions and arousal. As a result, symptoms persist and
emotional processing fails (Wells, 2000).
The S-REF model of clinical disorder (Wells & Matthews, 1996) and the metacognitive
model of PTSD (Wells, 2000) suggest that the use of worry as a coping strategy is motivated
by the person’s metacognitive beliefs which bias the cognitive system to use perseverative
modes of processing in response to perceptions of threat. In this context, the metacognitive
model distinguishes between two types of worry. One that represents a cognitive symptom
of anxiety and a second type that functions as a thought control strategy. The latter is
assumed to be a particularly dysfunctional type of worry as it impairs the self-regulatory
process of developing a more adaptive way of dealing with intrusive thoughts and arousal.
The first attempt to measure worry as a thought control strategy and, thus, distinguish it
from symptomatic worry was made by Wells and Davies (1994), who developed the
Thought Control Questionnaire (TCQ). In Wells and Davies’ (1994) study, the worry
subscale of the TCQ was found to have a modest correlation (r .49) with scores on the
Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990),
suggesting that the use of worry as a thought control strategy is conceptually and empirically
discriminable from symptomatic worry. Wells and Davies (1994) also found that the use of
worry as a strategy to control thoughts was positively associated with several other indices of
stress vulnerability and psychopathology, such as neuroticism and trait anxiety. TCQ-worry
has been consistently found to have a positive relationship with stress symptoms (Guthrie &
 Anxiety, Stress & Coping 215

Bryant, 2000; Warda & Bryant, 1998) and also be negatively associated with recovery from
PTSD and depression (Reynolds & Wells, 1999). Furthermore, Holeva, Tarrier, and Wells
(2001) established that TCQ-worry was positively associated with stress symptoms in
victims of road accidents, and positively predicted the subsequent development of PTSD at
4
6 months follow-up.
The results of these studies are consistent with the metacognitive model of PTSD, which
proposes that thought control strategies typified by worry contribute to the maintenance of
stress symptoms following trauma. Moreover, the study by Holeva et al. (2001)
demonstrated that the use of worry as a thought control strategy was prospectively related
to the development of PTSD at 4
6 months post-trauma. However, in none of these studies
was the effect of worry as a symptom of anxiety controlled for. In addition, there has been
no systematic effort to investigate the relationship between TCQ-worry and stress
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

symptoms in an analogue sample. It is necessary to test this relationship as the maladaptive

influence of TCQ-worry on post-stress adjustment should be generalized to a non-clinical
population.
Research has found that university students are subjected to high levels of stress (Bishop,
1990; Gall, Evans, & Bellerose, 2000), and perceive academic life as stressful and
demanding (Hammer, Grigsby, & Woods, 1998). Over time, the types of psychological
problems bringing students to counseling centers have changed significantly. A recent
survey of approximately 13,000 students demonstrated that until 1994, adjustment
difficulties and relationship issues were the most common areas of problems among
university students. Since then, there has been an increase in the number of students
manifesting severe symptoms of stress and anxiety (Benton, Robertson, Tseng, Newton, &
Benton, 2003). Specifically, the incidence of major depression among students has doubled
since 1994, while the incidence of suicide attempts has tripled (Benton et al., 2003). In the
United States, the percentage of college undergraduates with psychiatric disabilities has
increased from 2.8% in 1978 to 10% in 2000 (Paul, 2000). Nevertheless, students with
mental health problems are reluctant to disclose or speak about their psychological
difficulties in case they are stigmatized, embarrassed, or treated inequitably (Megivern,
2002). Their reluctance to seek professional help is further augmented by parents who are
afraid of the consequences a mental health diagnosis might have for their children’s career
(Mowbray, Megivern, & Strauss, 2002). These findings suggest that university students are
nowadays struggling to cope with more severe psychological problems than in the past.
They also suggest the possibility that stress-related disorders in this population might be
underestimated due to students’ fears of stigmatization and social exclusion.
Based on a prospective longitudinal design, the present study used a sample of university
students whose stress symptoms were assessed by self-report questionnaires at two time
points: one week after being exposed to a stressor and three months later. The types of
stressors afflicting participants in this study fell within one of the following areas: (i)
academic; (ii) relationship; (iii) health; (iv) financial; and (v) life transition problems. For
instance, moving into a new place to study was reported as a major life stressor by a few
participants. Other studies based on student samples have found that separation from home
environment for the purpose of academic study can be emotionally stressful (Abouserie,
1994; Mori, 2000). Gadzella (1994) demonstrated that students manifest psychological,
physical, cognitive, and behavioral reactions to academic stress. In addition, Misra, Crist,
and Burant (2003) found that higher levels of academic pressure predicted greater stress
reactions. These findings provided further support to earlier studies demonstrating a
significant relationship between depression, physical symptoms, low grade point average,
 216 P. Roussis and A. Wells

and academic stress (Felsten & Wilkox, 1992; Reifman & Dunkel-Schetter, 1990). There is
also evidence that financial difficulties arising from cutbacks in scholarships and student
loans can exacerbate perceptions of stress among students, and precipitate psychological
disorders, such as major depression (Reinherz, Giaconia, Wasserman, Silverman, & Burton,
2003).
Although life-threatening stressors were also reported by participants in this study, these
were less frequent and involved road accidents, physical and sexual abuse, and witnessing
one’s death. In previous studies, college students have reported experiencing assaults and
rapes, as well as witnessing suicides and homicides on and around campuses (Arata &
Burkart, 1996; Utterback & Caldwell, 1989). However, none of the aforementioned studies
investigated the relationships between worry used as a thought control strategy and stress
symptoms. Research on students’ experience of stress has demonstrated that avoidant
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

coping, a function of worry, predicted lower levels of adjustment to college life (Aspinwall &
Taylor, 1992).
The present study sought to investigate whether the use of worry as a thought control
strategy predicted the development of stress symptoms in a student population. Secondary
to this, we attempted to disentangle the effects of worry as a control strategy from worry as a
symptom of anxiety. The metacognitive model gives particular prominence to worry as a
coping strategy in the development of abnormal stress reactions. Accordingly, the
hypothesis tested was that the use of worry as a strategy to cope with intrusive thoughts
about the stressor (i.e., TCQ-worry) should predict significantly long-term symptoms of
stress over and above what can be predicted on the basis of symptoms at baseline, degree of
exposure to stress, and worry assessed as a symptom of anxiety.

Method
Design
This study used a longitudinal design, with two assessment points separated by
approximately three months (M 97.3 days, SD 38.1). This design was chosen in order
to evaluate TCQ-worry as an independent predictor of stress symptoms.

Participants
Six hundred and ten (610) students at the University of Manchester expressed their interest
in participating in the study after it had been advertised on the university electronic mail
list. Respondents had to meet a number of criteria in order to be eligible to participate,
including being over 18 years of age, having experienced a recent stressful event (within the
past seven days), and consenting to complete a battery of questionnaires twice, separated by
several months. Of those initially interested, two hundred and seventeen (217) students met
these admission criteria and participated in the study at time 1 (35.6%). The mean age of
participants at initial assessment was 29.6 years (SD 5.7; range 18
52 years) and 96 of
them were male (44%). Of the 217 participants who were assessed at time 1, one hundred
and three (103) refused to be re-assessed three months later (47.5%). Among these, there
were nine people (8.7%) who had clinically significant symptoms of stress as indicated by
their total score of 35 or greater on the Penn Inventory (PI). Overall, the non-completers
did not differ significantly from those who completed both parts of the study in terms of
their gender, age, and scores on any of the other measures.
 Anxiety, Stress & Coping 217

Out of 217 participants at time 1, one hundred and ten (110) were contactable and
agreed to participate at time 2 (50.7%). Among these, 14 (12.7%) had scored at least 35 on
the PI at time 1, thus manifesting clinically significant symptoms of stress at initial
assessment. The mean age of those who completed both parts of the study was 22.2 years
(SD 10.7) and 78 of them were male (70.9%). All had moderate to severe levels of stress
exposure to their experienced stressor, as assessed by the Inventory of College Students’
Recent Life Experiences (ICSRLE; Kohn, Lafraniere, & Gurevich, 1990). The types of
stressors reported by participants were as follows: relationship conflicts (37.8%), academic
difficulties (23.9%), illness requiring hospitalization (9.3%), moving into a new place in
order to study (7.9%), bereavement (7%), financial difficulties (5.6%), loneliness (3.7%),
road traffic accidents (3.3%), physical and sexual abuse (1.8%), witnessing someone’s
death (0.4%), and parental divorce (0.4%).
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Materials
Dependent measures. The PI (Hammarberg, 1992) is a 26-item measure of PTSD that has
been modelled on the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, &
Erbaugh, 1961). Each item has a score ranging from 0 to 3, which gives a possible range of
scores from 0 to 78 on the total scale. A total estimate is provided by summing all of the
items, with a threshold of 35 being consistent with severe PTSD. The PI has good
psychometric properties with a classification rate of 97%, sensitivity of 98%, and specificity
of 94% with combat veterans. Test
retest reliability is high (r 0.96). The present study
used the PI in order to assess stress symptoms that occurred within at least 23 days after
stress exposure, and three months later. The instructions of the PI were slightly modified
when following up participants, in order to capture the intensity of symptoms since the
initial assessment. That is, participants were instructed to pick out one response of each
group of statements that best described the way they had been feeling over the previous
three months including the day of their responding.

Independent Measures.
1. The ICSRLE (Kohn et al., 1990) was used in order to assess the level of students’
exposure to academic life stressors. The inclusion of this questionnaire in the study
also allowed a distinction between stress exposure and stress symptoms. In summary,
the ICSRLE is a 48-item self-report questionnaire that measures students’ exposure to
sources of stresses or hassles, and also allows for an identification of the extent to
which these stressors are experienced over the past month: (a) ‘‘not at all part of my
life’’; (b) ‘‘only slightly part of my life’’; (c) ‘‘distinctly part of my life’’; (d) ‘‘very much
part of my life’’. Scoring for the ICSRLE can range from 0 to 3 for each item (a 0,
b 1, c2, d 3). Higher scores indicate higher levels of exposure to hassles. Two
scores can be derived from this scale. A frequency score is derived by summing up the
total number of hassles reported by each respondent. An intensity score is the sum of
the severity ratings for each hassle reported. In the present study, the intensity score
only was used in the analysis. The psychometric properties of the ICSRLE are very
good. Kohn et al. (1990) found that the alpha reliability of the scale was .88, while its
correlation against other measures of stress exposure, such as the Perceived Stress
Scale (PSC; Cohen, Kamarack, & Mermelstein, 1983) was .59 ( p B.0005). These
findings support the reliability and validity of this measure.
 218 P. Roussis and A. Wells

In this study, we included an open-ended question at the end of the ICSRLE, which
asked respondents to report their greatest stressor over the past month at time 1 was
assessed separately for descriptive purposes.
2. The TCQ (Wells & Davies, 1994) was used to assess the use of worry to control
thoughts. It consists of five subscales and assesses individual differences in the control
of intrusive thoughts by strategies of distraction (e.g. ‘‘I keep myself busy’’), worry
(e.g. ‘‘I dwell on other worries’’), social control (e.g. ‘‘I find out how my friends deal
with these thoughts’’), punishment (e.g. ‘‘I shout at myself for having the thought’’),
and re-appraisal (e.g. ‘‘I challenge the thoughts validity’’). Responses are rated on a 1

4 scale (1never, 2 sometimes, 3 often, 4almost always). The subscales are
scored by summating the relevant item scores. Internal consistency of subscales is as
follows: Distraction (a 0.72), Worry (a 0.71), Social Control (a 0.79), Punish-
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

ment (a 0.64), and Re-appraisal (a 0.67). Test
retest reliability is high (r 0.83).

The TCQ subscales are significantly correlated with several indices of psychopathol-
ogy, such as the PSWQ (Wells & Davies, 1994), and the Padua Inventory (Wells &
Davies, 1994). Since it was developed, the TCQ has been used in research on patients
with PTSD (Holeva et al., 2001) and obsessional symptoms (Amir, Cashman, & Foa,
1997). For purposes of the present study we specifically used the worry subscale,
although the entire TCQ was administered so as not to influence its psychometric
properties.
3. PSWQ (Meyer et al., 1990). This is a 16-item self-report questionnaire capturing the
generality, excessiveness, and uncontrollability characteristics of pathological worry
(e.g. ‘‘I worry all the time’’). Each item is rated on a 1 (‘‘not at all typical of me’’) to 5
(‘‘very typical of me’’) Likert type scale. Eleven are worded in a way such that higher
numbers indicate more worry (e.g. ‘‘My worries overwhelm me’’). Five items ask the
respondent to endorse the absence of problems with worry (e.g. ‘‘I find it easy to
dismiss worrisome thoughts’’) and these are reverse scored prior to computing the
total score. In student samples, the PSWQ has demonstrated good internal
consistency (a 0.86
0.93) and good test
re-test reliability over intervals as long as
8
10 weeks (r 0.87). The PSWQ has also exhibited moderate to strong correlations
with measures of anxiety and depression (Meyer et al., 1990).

Procedure
The questionnaires were mailed to participants twice with an interim period of three
months (90
104 days). Occasionally, participants were recruited during home visits by the
researcher. In each case, the aim of the study was described as the identification of factors
predicting academic-related stress.

Statistical Analysis
Pearson’s bivariate correlations were computed to test the relationships between the
independent variables and stress symptoms at both time points of assessment. Partial
correlations tested the prospective relationships of the independent variables to stress
symptoms at time 2, while controlling for stress symptoms at time 1. Hierarchical regression
analysis was run to test if TCQ-worry at time 1 significantly predicted stress symptoms at
three months, over and above symptomatic worry (PSWQ), stress exposure (ICSRLE), and
stress symptoms (PI), assessed at time 1.
 Anxiety, Stress & Coping 219

Results
Prevalence of Symptoms of Stress
From a total of 217 participants at time 1, 19 (8.9%) manifested clinically significant stress
symptoms as determined by a score of 35 or greater on the PI. At time 2, severe stress
symptoms persisted in 12 participants (10.9%) out of the 110 who completed the follow-up
assessment, as indicated by a score of 35 or more on the PI. At time 2, there were six new
cases with severe symptoms of stress, and seven participants who had severe symptoms at
time 1, but remitted at time 2. Of 19 participants with clinically significant symptoms of
stress at time 1, six (31.6%) were found to still have severe stress symptoms. The mean
score on the measure of stress exposure (ICSRLE) suggested that the average stress level in
our student sample at time 1 was ‘‘mild’’. Of those participants who completed both parts
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

of the study, 68 had ‘‘moderate to severe’’ levels of stress exposure at time 1 (61.8%).

Descriptive Statistics
The means and standard deviations for all variables are shown in Table 1. At both times of
assessment, the average scores on the PI were comparable to those found in previous studies
that have used this symptom measure in a college student population (Roussis & Wells,
2006). Mean scores on TCQ-worry, stress exposure measure, and the PSWQ were also
similar to those found in other studies (Kohn et al., 1990; Meyer et al., 1990; Wells & Davies,
1994). However, the average level of stress exposure among students in this study was less
than that found in other studies using college student samples (e.g. Kohn et al., 1990).

Variation in Scores Over Time

In order to examine the pattern of change in scores from time 1 to time 2, repeated
measures t-tests were computed for each variable. Table 1 shows t-estimates for each
variable along with their associated significance levels. There was a statistically significant
decrease in PI scores from time 1 (M 23.5, SD 9.1) to time 2 (M 21.5, SD 8.8),
t (110)2.4, pB.01. The eta squared statistic (.05) indicated a moderate effect size
(Cohen, 1988). Scores on the TCQ-worry, PSWQ, and the ICSRLE were not significantly
different across the two time points.

Tests of Cross-sectional Relationships at Four Weeks and Three Months

In order to test the relationships between TCQ-worry, PSWQ, and ICSRLE, and the PI,
Pearson bivariate correlations were computed among these variables at both time points of
Table 1. Descriptive and t-statistics for all variables at time 1 and time 2.

Time 1 Time 2

Variables Mean SD Mean SD t p

Stress symptoms (PI) 23.5 9.1 21.5 8.8 2.4 .01

TCQ-Worry 9.5 2.8 9.3 3.1
.07 .94
PSWQ 47.1 12.9 48.5 12.2
.34 .73
ICSRLE 88.1 19.1 87.7 17.6 .27 .78

Note: PI, Penn Inventory; TCQ, Thought Control Questionnaire; PSWQ, Penn State Worry Questionnaire;
ICSRLE, Inventory of College Students’ Recent Life Experiences; *p B.05, **p B.01.
 220 P. Roussis and A. Wells
Table 2. Cross-sectional bivariate correlations among predictor variables and stress symptoms at time 1 (i.e., four
weeks), and time 2 (i.e., three months) and prospective partial correlations for predictors at time 1 and symptoms
at time 2.

Stress symptoms at time 1 (PI-1) Stress symptoms at time 2 (PI-2)

Variable (bivariate) r (bivariate) r (Partial: controlling for PI-1) r

PSWQ .35** .40** .22*

ICSRLE .55** .51** .27**
TCQ-Worry .22** .37** .30**

Note: PI-1, Penn Inventory at time 1; PI-2, Penn Inventory at time 2; PSWQ, Penn State Worry Questionnaire;
ICSRLE, Inventory of College Students’ Recent Life Experiences; TCQ, Thought Control Questionnaire; *p B
.05, **pB.01.
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

assessment. Table 2 shows these cross-sectional relationships along with their associated
significance values. Scores on TCQ-worry, PSWQ, and the ICSRLE were positively and
significantly correlated with the PI both at time 1 and time 2.

Tests of Prospective Relationships

Partial correlations were computed between each of the independent variables at time 1 and
stress symptoms at time 2, while controlling for stress symptoms at time 1. TCQ-worry,
PSWQ, and stress exposure (ICSRLE) were found to predict positively and significantly PI
scores at time 2, while controlling for PI at time 1. These relationships are presented in
Table 2.
While the partial correlations suggest TCQ-worry causes stress symptoms, it could also
be that the use of worry as a thought control strategy is a consequence of stress symptoms.
In order to evaluate this possibility, backward correlations were computed between scores
on the PI at time 1 and TCQ-worry assessed at time 2, while controlling for PI at time 2.
There was a small and insignificant partial correlation between symptoms of stress at time 1
and TCQ-worry assessed at time 2 [r 
.016; n 110, ns], the nature of this relationship
being negative. This result, along with the significant correlation between TCQ-worry at
time 1 and symptoms of stress at time 2, is consistent with TCQ-worry being a causal
predictor of stress symptoms rather than a consequence of the latter. But this should be
confirmed in experimental studies.

Hierarchical regression analysis. Hierarchical multiple regression analysis tested whether

TCQ-worry predicts stress symptoms at time 2 (PI) over and above what can be predicted
on the basis of stress symptoms at time 1 and initial levels of stress exposure and controlling
for symptoms of worry. Having established that stress symptoms at time 1 and stress
exposure predicted significantly stress symptoms at three months, we controlled for these
variables in the first step of the regression analysis. Both of these variables predicted 48.6%
of the variance, R square .486, F(2, 109) 50.4, p .0005.
When PSWQ (symptom worry) was entered into the second step, there was a small but
statistically insignificant increase in the amount of symptom variance explained, R square
change .007, F change (1, 105) 1.4, p .23.
In the final step of the regression, TCQ-worry significantly improved the prediction of
the severity of stress symptoms at three months, R square change .021, F change (1,
104) 4.5, p B.05, F(4, 108) 27.5, p B.0005. The total amount of variance explained in
the final model with all predictors entered was 51.4%.
 Anxiety, Stress & Coping 221

The Tolerance values were acceptable (.68
.77). These results suggest that the multiple
correlation of each predictor variable with the other independent variables is low, and,
therefore, the observed relationships between our predictor variables and stress symptoms
at time 2 are not attributable to the interrelatedness among predictors. Taken together,
these results indicate that the use of worry as a thought control strategy (TCQ-worry)
predicts long-term PTSD over and above what can be predicted on the basis of initial levels
of stress symptoms, stress exposure (ICSRLE), and worry as a symptom of anxiety
(PSWQ). Table 3 presents summary statistics for each of the predictors in the final step of
the regression analysis. Only TCQ-worry and symptoms measured at time 1 independently
predict symptoms at time 2. Stress exposure at time 1 makes a contribution that marginally
misses the conventional significance level.
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Discussion
The present study demonstrated that university students are exposed to a range of levels of
stress as assessed by the ICSRLE. With more than half who completed both assessments
reporting moderate to severe levels of stress exposure at time 1. Students’ responding to an
open-ended question revealed that the most prevalent stressor in this sample involved
relationship issues with a prevalence rate of 37.8%. The second most prevalent stressor was
academic difficulties (23.9%), followed by life-threatening illness (9.3%), and moving into
a new place to study (7.9%). Other stressors reported by students in this study concerned
bereavement (7%), financial difficulties affecting academic study (5.6%), and loneliness
(3.7%). To a lesser extent, there were also reports of potentially traumatic stressors,
including road accidents (3.3%), experiences of physical and sexual abuse (1.8%), and
witnessing someone’s death (0.4%). The mean level of stress reported by students in this
study is smaller than that found in previous studies that have employed the ICSRLE (Kohn
et al., 1990). However, the types of stressors reported by students in our sample are similar
to those found elsewhere, although their incidence rates tend to vary across studies
(Heckert, 1999; Roussis & Wells, 2006; Sgan-Cohen & Lowental, 1988).
The prevalence of clinically significant levels of stress symptoms in our sample, as
established by the PI, was 8.9% at time 1 and 10.9% at time 2. Of those who completed
both parts of the study, 12.7% manifested clinically significant symptoms of stress. It is
difficult to compare the levels of stress symptoms in our sample with those reported in other
samples, given that no other prospective study has used the PI in order to assess stress
symptoms in a student population. However, the mean score on stress symptoms at time 1
in this study is consistent with a previous sample of university students assessed by the PI
cross-sectionally (Roussis & Wells, 2006).
Our findings support the role of TCQ-worry as a metacognitive predictor of stress
symptoms in a non-clinical population. The present study demonstrated that the tendency
to worry as a thought control strategy (TCQ-worry) measured soon after a stressful life
Table 3. Regression summary statistics for final step of equation (dependent variable: Penn Inventory at time 2).

Predictors R2 Beta t p

PI time 1 .51 .52 6.6 .0005

ICSRLE .16 1.8 .06
PSWQ .04 0.5 .59
TCQ-Worry .16 2.1 .03
 222 P. Roussis and A. Wells

event predicted the severity of stress symptoms three months after the event. This
relationship was independent of initial levels of stress symptoms, worry as an anxiety
symptom (PSWQ) and stress exposure.
Backward correlations indicated that TCQ-worry at time 2 could not be predicted by
stress symptoms at time 1 when the effects of time 2 stress symptoms and either stress
exposure or symptomatic worry, assessed at time 2, were partialled out. While the strength
of the relationships between TCQ-worry at time 1 and stress symptoms at time 2 was small,
this combination of results is consistent with the idea that TCQ-worry causes symptoms
rather than symptoms causing change in TCQ-worry.
This pattern of results extends previous findings in victims of road traffic accidents
(Holeva et al., 2001) and is in line with the hypothesis that the long-term influence of TCQ-
worry on stress symptoms is independent of worry as a symptom of anxiety (Wells, 2000).
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

The findings of the present study extend the results of previous studies which have
demonstrated that TCQ-worry is positively associated with stress symptoms following
traumatic life experiences (Guthrie & Bryant, 2000; Warda & Bryant, 1998).
The evidence pertaining to the independent effects of TCQ-worry compared to
symptomatic worry in predicting stress symptoms is new and potentially important. Worry
as a symptom has been found to be common in all of the anxiety disorders (Barlow, 1988).
In addition, Shearer and Gordon (2006) maintain that following trauma pathological worry
is tied up with heightened levels of arousal both of which maintain an increased perception
of the likelihood of danger. However, in the present study, worry symptom levels did not
predict stress symptoms but TCQ-worry did. It is possible that the effects of worry on
arousal may reflect the individuals coping strategy.
The present findings are consistent with the metacognitive model of PTSD (Wells,
2000), which suggests that worry as a strategy can have deleterious effects on emotional
processing of stress. For instance, in the initial aftermath of trauma TCQ-worry is likely to
anchor attention on threat and increases the sense of danger so that the individual cannot
return thinking to its normal threat-free status. Prolonged use of worry as a thought control
strategy is likely to use up executive processing resources that are necessary for more
adaptive emotional control and coping. These data support the targeting of worry and
thought control strategies that represents a central feature of metacognitive therapy (Wells
& Sembi, 2004a; Wells & Sembi, 2004b).
Studies using exposure to analogue stress have demonstrated that worrying about a
stressor has an incubative effect on the frequency of stress-related images over the next
three days compared to a control group (Butler, Wells, & Dewick, 1995; Wells &
Papageorgiou, 1995). These data suggest that worry of other types might contribute to
intrusive thoughts and it is important to refine studies to test effects of the type and timing
of worry on different stress symptoms.
The present study has both strengths and weaknesses. Among the strengths was the use
of a prospective longitudinal design that allowed for testing the relationship between TCQ-
worry and change in symptoms of stress. Among the weaknesses was the high attrition rate,
which limits our certainty about the predictive effects of TCQ-worry. However, completers
were found not to differ significantly from non-completers in terms of their stress
symptoms, age, gender, or any other predictor variable.
The use of a sample that consisted of college students is also a limitation. Clearly, our
findings cannot generalize to populations afflicted by serious traumas. The types of stressors
reported by students in this study were relatively mild and although no measures were taken
of threat to personal integrity, it is unlikely that they threatened students’ lives, or that they
 Anxiety, Stress & Coping 223

produced feelings of intense fear, horror, or helplessness (American Psychiatric Association

[APA], 2000; DSM-IV-TR). Research on the neurobiology of PTSD indicates that
symptomatic individuals show a pattern of biological responses that is significantly different
from that found in depressed patients and other individuals exposed to normative life stress.
For instance, in some of the studies that have examined the function of the hypothalamic

pituitary
adrenal (HPA) axis, PTSD individuals have been found to have abnormally low
levels of cortisol (Mason, Giller, Kosten, Ostroff, & Podd, 1986; Resnick, Yehuda, &
Acierno, 1997; Yehuda et al., 1995).
Another weakness is that we did not control comorbid psychological condition. Major
depression has been found to increase vulnerability to negative effects of traumatic events
(Fullerton et al., 2000). It is often comorbid with PTSD (Kessler et al., 1995), and has also
been found to afflict college students (Benton et al., 2003; Vrana & Lauterbach, 1994).
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Personality factors, such as neuroticism, might also be responsible for the observed
relationships between TCQ-worry and stress symptoms. Measures of neuroticism are
significantly correlated with both anxiety (Morgan, Matthews, & Winton, 1995) and
strategic worry (Wells & Davies, 1994). In addition, past research has demonstrated that
neuroticism is positively associated with exposure to interpersonal stress (Bolger &
Schilling, 1991) and affective reactivity to daily stressors (Bolger & Zuckerman, 1995).
Future studies should replicate current findings controlling for levels of depression and
neuroticism. Finally, the relationships explored in this study were between self-report
measures and the findings would be strengthened by similar results from more objective
measures or from manipulation of predictors or stress exposure.
In conclusion, the findings of the present study support distinctions between types of
worry and suggest the use of worry as a thought control strategy is relevant to the
development of persistent stress reactions. This finding is consistent with the metacognitive
model of PTSD (Wells, 2000), which suggests that perseverative-verbal processing of
threat-related information is a dysfunctional strategy that interferes with post-stress
adjustment.

References
Abouserie, R. (1994). Sources and levels of stress in relation to locus of control and self-esteem in college students.
Educational Psychology: An International Journal of Experimental Educational Psychology, 14(3), 323
330.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed). Washington,
DC: Author.
Amir, N., Cashman, L., & Foa, E. B. (1997). Strategies of thought control in obsessive-compulsive disorder.
Behavior Research and Therapy, 35(8), 775
777.
Arata, C. M., & Burkart, B. R. (1996). Post-traumatic stress disorder among college student victims of
acquaintance assault. In E. S. Byers, & L. F. O’Sullivan (Eds.), Sexual coercion in dating relationships (pp. 79
92).
New York: Haworth Press.
Aspinwall, L., & Taylor, S. (1992). Modelling cognitive adaptation: A longitudinal investigation of the impact of
individual differences and coping on college adjustment and performance. Journal of Personality and Social
Psychology, 63(6), 989
1003.
Barlow, D. (1988). Anxiety and its disorders: The nature and treatment of anxiety and panic. New York: Guildford
Press.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression.
Archives of General Psychiatry, 4, 561
571.
Benton, S. A., Robertson, J. M., Tseng, W., Newton, F. B., & Benton, S. L. (2003). Changes in counselling center
client problems across 13 years. Professional Psychology: Research and Practice, 34, 66
72.
Bishop, J. B. (1990). The university counselling center: An agenda for the 1990s. Journal of Counselling and
Development, 68, 408
413.
 224 P. Roussis and A. Wells
Bolger, N., & Schilling, E. A. (1991). Personality and the problems of everyday life: The role of neuroticism in
exposure and reactivity to daily stressors. Journal of Personality, 59, 355
386.
Bolger, N., & Zuckerman, A. (1995). A framework for studying personality in the stress process. Journal of
Personality and Social Psychology, 69, 890
902.
Borkovec, T. D. (1994). The nature, functions, and origins of worry. In G. C. L. Davey, & F. Tallis (Eds.),
Worrying: Perspectives on theory, assessment, and treatment. Wiley series in clinical psychology (pp. 5
33). Chichester,
UK: Wiley.
Borkovec, T. D., Alcaine, O., & Behar, E. (2001). Avoidance theory of worry and generalized anxiety disorder. In
R. G. Heimberg, C. L. Turk, & D. S. Mennin (Eds.), Generalized anxiety disorder: Advances in research and
practice (pp. 77
108). New York: Guildford.
Borkovec, T. D., & Hu, S. (1990). The effect of worry on cardiovascular response to phobic imagery. Behaviour
Research and Therapy, 28, 69
73.
Borkovec, T. D., & Lyonfields, J. D. (1993). Worry: Thought suppression of emotional processing. In H. W.
Krohne (Ed.), Attention and avoidance: Strategies in coping with aversiveness (pp. 101
118). Seattle, WA: Hogrefe
& Huber.
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

Borkovec, T. D., & Roemer, L. (1995). Perceived functions of worry among generalized anxiety disorder subjects:
Distraction from more emotional topics? Journal of Behavior Therapy and Experimental Psychiatry, 26, 25
30.
Butler, G., Wells, A., & Dewick, H. (1995). Differential effects of worry and imagery after exposure to a stressful
stimulus: A pilot study. Behavioural and Cognitive Psychotherapy, 23, 45
56.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.
Cohen, S., Kamarack, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and
Social Behavior, 24, 385
396.
Felsten, G., & Wilkox, K. (1992). Influences of stress and situation-specific mastery beliefs and satisfaction with
social support on well-being and academic performance. Psychological Reports, 70, 292
303.
Fullerton, C. S., Ursano, R. J., Epstein, R. S., Crowley, B., Vance, K. L., Kao, T. C., et al. (2000). Peritraumatic
dissociation following motor vehicle accidents: Relationship to prior trauma and prior major depression. Journal
of Nervous Mental Disease, 188, 267
272.
Gadzella, B. M. (1994). Student-Life Stress Inventory: Identification of and reactions to stressors. Psychological
Reports, 74, 395
402.
Gall, T. L., Evans, D. R., & Bellerose, S. (2000). Transition to first year university: Patterns of change in
adjustment across life domains and time. Journal of Social and Clinical Psychology, 19, 544
567.
Guthrie, R., & Bryant, R. (2000). Attempting suppression of traumatic memories over extended periods in acute
stress disorder. Behaviour Research and Therapy, 38, 899
907.
Hammarberg, M. (1992). Penn Inventory for posttraumatic stress disorder: Psychometric properties. Psychological
Assessment, 4, 67
76.
Hammer, B. L., Grigsby, D. T., & Woods, S. (1998). The conflict demand of work, family, and school among
students at an Urban University. The Journal of Psychology, 132(2), 220
226.
Heckert, T. M. (1999). Sources of stress among college students. College Student Journal, 33(2), 312
317.
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
84.
Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B. (1995). Posttraumatic stress disorder in the
National Comorbidity Survey. Archives of General Psychiatry, 52, 1048
1060.
Kohn, M. P., Lafraniere, K., & Gurevich, M. (1990). The Inventory of College Students’ Recent Life Experiences:
A decontaminated hassles scale for a special population. Journal of Behavioral Medicine, 13(6), 619
630.
Mason, J. W., Giller, E. L., Kosten, T. R., Ostroff, R. B., & Podd, L. (1986). Urinary free cortisol levels in
posttraumatic stress disorder patients. Journal of Nervous and Mental Disease, 174, 145
149.
Megivern, D. (2002). Disability services and college students with psychiatric disabilities. Journal of Social Work in
Disability and Rehabilitation, 1(3), 25
42.
Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. J. (1990). Development and validation of the Penn
State Worry Questionnaire. Behaviour Research and Therapy, 28(6), 487
495.
Misra, R., Crist, M., & Burant, C. (2003). Relationships among life stress, social support, academic stressors, and
reactions to stressors of international students in the United States. International Journal of Stress Management,
10(2), 137
157.
Morgan, I., Matthews, G., & Winton, M. (1995). Coping and personality as predictors of posttraumatic intrusions,
numbing, avoidance, and general distress: A study of victims of the Perth flood. Behaviour and Cognitive
Psychotherapy, 23, 251
264.
Mori, S. C. (2000). Addressing the mental health concerns of international students. Journal of Counselling and
Development, 78, 137
144.
 Anxiety, Stress & Coping 225
Mowbray, C. T., Megivern, D., & Strauss, S. (2002). College students’ narratives of high school experiences: Coping with
serious emotional disturbance. New York: John Wiley & Sons.
Paul, S. (2000). Students with disabilities in higher education: A review of the literature. College Student Journal, 34,
200
210.
Reifman, A., & Dunkel-Schetter, C. (1990). Stress, structural social support, and well-being in university students.
College Health, 38, 271
277.
Reinherz, H. Z., Giaconia, R. M., Wasserman, M. S., Silverman, A. B., & Burton, L. (2003). Coming of age in the
1990s: Influences of contemporary stressors on major depression in young adults. In P. Cohen, C. Slomkowski,
& L. N. Robins (Eds.), Historical and geographical influences on psychopathology (pp. 141
162). Mahwah, NJ:
Erlbaum.
Resnick, H. S., Yehuda, R., & Acierno, R. (1997). Acute post-rape plasma cortisol, alcohol use, and PTSD
symptom profile among recent rape victims. Annuals of New York Academy of Science, 821, 433
436.
Reynolds, M., & Wells, A. (1999). The Thought Control Questionnaire
Psychometric properties in a clinical
sample, and relationships with PTSD and depression. Psychological Medicine, 29, 1089
1099.
Roemer, L., & Borkovec, T. D. (1993). Worry: Unwanted cognitive activity that controls unwanted somatic
Downloaded by [University of Saskatchewan Library] at 15:54 17 October 2012

experience. In D. M. Wegner, & J. W. Pennebaker (Eds.), Handbook of mental control (pp. 220
238). Englewood
Cliffs, NJ: Prentice-Hall.
Roemer, L. Litz, B. T. Pepper, C. M. & Borkovec, T. D. (1995). The relationship between worry and PTSD following
exposure to potentially traumatizing events. Paper presented at the annual convention of the Association for
Advancement of Behavior Therapy, Washington, DC.
Roemer, L., Molina, S., Litz, B. T., & Borkovec, T. D. (1993). A preliminary investigation of the role of potentially
traumatizing events in the development of generalized anxiety disorder. Depression and Anxiety, 4, 134
138.
Roussis, P., & Wells, A. (2006). Post-traumatic stress symptoms: Tests of relationships with thought control
strategies and beliefs as predicted by the metacognitive model. Personality and Individual Differences, 40, 111
122.
Sgan-Cohen, H. D., & Lowental, U. (1988). Sources of stress among Israeli dental students. The Journal of the
American College Health Association, 36, 317
321.
Shearer, S., & Gordon, L. (2006). The patient with excessive worry. American Family Physician, 73(6), 1049
1056.
Shore, J. H., Tatum, E. L., & Vollmer, W. M. (1986). Psychiatric reactions to disaster: The Mount St Helens
experience. American Journal of Psychiatry, 143, 590
595.
Utterback, J., & Caldwell, J. (1989). Proactive and reactive approaches to posttraumatic stress disorder in the
aftermath of campus violence: Forming a traumatic stress reaction team. Journal of Traumatic Stress, 2(2), 171

184.
Vrana, S., & Lauterbach, D. (1994). Prevalence of traumatic events and post-traumatic psychological symptoms in
a non-clinical sample of college students. Journal of Traumatic Stress, 7(2), 289
302.
Warda, G., & Bryant, R. (1998). Thought control strategies in acute stress disorder. Behaviour Research and
Therapy, 36, 1171
1175.
Wells, A. (2000). Emotional disorders and metacognition: Innovative cognitive therapy. Chichester, UK: Wiley.
Wells, A., & Davies, M. (1994). The Thought Control Questionnaire: A measure of individual differences in the
control of unwarranted thoughts. Behaviour Research and Therapy, 32, 871
878.
Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical perspective. Hove, UK: Erlbaum.
Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-REF model. Behaviour
Research and Therapy, 34, 881
888.
Wells, A., & Papageorgiou, C. (1995). Worry and the incubation of intrusive images following stress. Behaviour
Research and Therapy, 33, 579
583.
Wells, A., & Sembi, S. (2004a). Metacognitive therapy for PTSD: A core treatment manual. Cognitive and
Behavioral Practice, 11, 365
377.
Wells, A., & Sembi, S. (2004b). Metacognitive therapy for PTSD: A preliminary investigation of a new brief
treatment. Journal of Behavior Therapy and Experimental Psychiatry, 35, 307
318.
Yehuda, R., Kahana, B., Binder-Brynes, K., Southwick, S., Mason, J. W., & Giller, E. L. (1995). Low urinary
cortisol excretion in Holocaust survivors with posttraumatic stress disorders. American Journal of Psychiatry, 152,
982
986.