ArticlePDF Available

Abstract

Theories of cognitive therapy have long suggested that those with generalized anxiety disorder (GAD) have inaccurate expectations. By challenging them with objective evidence, symptoms are thought to decrease. To test these premises, this study used ecological momentary assessment (EMA) during the Worry Outcome Journal (WOJ) treatment to determine the percentage of GAD worries that did not come true. We then analyzed the association between participants’ untrue worry percentages and GAD symptom change across treatment. Twenty-nine participants with GAD recorded worries when prompted for 10 days, reviewed them online nightly, and tracked their worry outcomes across 30 days. These recordings were then coded by independent raters. Analyses applied bias-correct bootstrapping path analysis on slopes extracted from longitudinal linear mixed models. Primary results revealed that an average of 91.4% (mode = 100%) of worry predictions did not come true. Higher percentages of untrue worries significantly predicted lower GAD symptoms after treatment, as well as a greater slope of symptom reduction from pre- to post-trial. Participants’ average expected likelihoods of worries coming true were much greater than actual observed likelihoods. Thus, worries in those with GAD were mostly inaccurate and greater evidence of this inaccuracy predicted greater improvement in treatment. As theorized, providing evidence for false expectations may significantly contribute to treatment’s effect.
Exposing Worry’s Deceit: Percentage of Untrue Worries in
Generalized Anxiety Disorder Treatment
Lucas S. LaFreniere,
The Pennsylvania State University
Michelle G. Newman
The Pennsylvania State University
Abstract
Theories of cognitive therapy have long proposed that those with generalized anxiety disorder
(GAD) have inaccurate expectations. By challenging them with objective evidence, symptoms are
thought to decrease. To test these premises, this study used ecological momentary assessment
(EMA) during the Worry Outcome Journal (WOJ) treatment to determine the percentage of GAD
worries that did not come true. We then analyzed the association between participants’ untrue
worry percentages and GAD symptom change across treatment. Twenty-nine participants with
GAD recorded worries when prompted for 10 days, reviewed them online nightly, and tracked
their worry outcomes across 30 days. These recordings were then coded by independent raters.
Analyses applied bias-correct bootstrapping path analysis on slopes extracted from longitudinal
linear mixed models. Primary results revealed that 91.4% of worry predictions did not come true.
Higher percentages of untrue worries significantly predicted lower GAD symptoms after
treatment, as well as a greater slope of symptom reduction from pre- to post-trial. Participants’
average expected likelihoods of worries coming true were much greater than actual observed
likelihoods. The most common percentage of untrue worries per person was 100%. Thus, worries
in those with GAD were mostly inaccurate. Greater evidence of this inaccuracy predicted greater
improvement in treatment. As theorized, disconfirming false expectations may significantly
contribute to treatment’s effect.
Keywords
generalized anxiety disorder; worry; cognitive therapy; worry outcome monitoring; self-
monitoring
Numerous efficacy studies have established the power of cognitive behavioral therapies
(CBTs), evincing moderate to large effect sizes (Butler, Chapman, Forman, & Beck, 2006;
Hofmann, Asnaani, Vonk, Sawyer, & Fang, 2012) and marking many as empirically
supported treatments (ESTs; Chambless & Ollendick, 2001). Yet less research has examined
how these therapies create change. Many scholars have argued that we should prioritize the
study of treatment predictors, processes, and targets, elucidating the factors responsible for
There are no conflicts of interest or financial disclosures related to this work.
HHS Public Access
Author manuscript
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Published in final edited form as:
Behav Ther
. 2020 May ; 51(3): 413–423. doi:10.1016/j.beth.2019.07.003.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
therapy outcome (e.g., Arch & Craske, 2008; Kazdin, 2008; Smits, Julian, Rosenfield, &
Powers, 2012). Identifying treatment’s active ingredients can then foster its augmentation
and development. Unfortunately, it is still an empirical question whether many current
therapies truly work via their theorized processes. For example, one of the central premises
of cognitive therapy (CT) is that by challenging unrealistic thoughts with objective evidence,
symptoms can be reduced (Beck, 1979). CT fundamentally assumes that 1) clients’
expectations are inconsistent with reality and 2) greater recognition of that inconsistency
leads to better outcomes. Yet despite being promoted for decades, the disconfirmation of
faulty beliefs has not been examined directly in the treatment of one prevalent syndrome:
Generalized anxiety disorder (GAD).
GAD’s primary criterion—worry—is a cognition that often contains unrealistic content.
Those with GAD are thought to continually expect unreasonable or exaggerated negative
future events, regardless of their improbability. Accordingly, CT for GAD teaches clients to
examine worries as hypotheses and test them against reality, believing that clients’ anxious
predictions will largely be exposed as false (Beck & Emery, 1985). Yet well-controlled
scientific study has not established that GAD worries are, in fact, irrational. Although
excessive worrying and GAD have been associated with higher likelihood estimations for
negative outcomes compared to controls (Berenbaum, Thompson, & Bredemeier, 2007;
Berenbaum, Thompson, & Pomerantz, 2007; Butler & Mathews, 1983; MacLeod, Williams,
& Bekerian, 1991), the only report on worry’s accuracy comes from a secondary description
of an unpublished study. In a theoretical article, Borkovec, Hazlett-Stevens, and Diaz (1999)
briefly mention the results of two questionnaire-based worry outcome monitoring
assessments. In the first, 67.9% of the worry predictions of 29 undergraduate participants
with GAD were rated as
not
having come true over a two-week period. In the second, 17
clients tracked worries and their outcomes during treatment. For this sample, 85% of worries
were rated as having turned out “better than expected.” Unfortunately, the strict accuracy of
their predictions was not reported. When worries did
not
turn out better than expected,
clients still coped better than expected 79% of the time. Although these reports are
promising, they were not peer-reviewed, few methodological details were reported, and no
mention was made of coders or their reliability. Yet not only is the accuracy of GAD worry
predictions unclear, but their association with change during CT has gone unstudied as well.
A true test of the theory underlying CT for GAD would require both establishing that
worries are commonly false and that recognizing this fact predicts change.
One cognitive treatment for GAD is specifically designed to target inaccurate predictions:
The Worry Outcome Journal (WOJ). The WOJ is an ecological momentary intervention in
which clients record their worrisome predictions in daily life, rate the in-the-moment costs
of each worry (such as distress and interference), and then track their actual outcomes as
objectively as possible (LaFreniere & Newman, 2016). It is an augmented form of worry
outcome monitoring, a common technique in CT for GAD (Borkovec & Newman, 1998). By
monitoring worry outcomes and costs with the WOJ, those with GAD are meant to
recognize that their worries are both highly unlikely and costly. Thus, their worry is not
worthwhile (the costs of worry outweigh any benefits). In classic CT fashion, the WOJ is
meant to provide objective, personalized evidence that challenges the unrealistic predictions
associated with worry. In a randomized controlled trial (RCT), the WOJ outperformed an
LaFreniere and Newman Page 2
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
active control in reducing worry over 10 days and at 30-day follow-up (LaFreniere &
Newman, 2016). Yet the primary therapeutic process of the WOJ—revealing worry’s
inaccuracy—has not been directly tested.
Fortunately, the entries made in the WOJ offer useful ecological momentary assessment
(EMA) data that can address this question. Primarily, clients recorded their worrisome
predictions in real time, then reported on their outcomes. Thus, WOJ entries can elucidate
clients’ perceived accuracy of worry after attending to the reality of events—a core process
of CT. Moreover, clients also recorded the distress associated with each worry, its duration
of cognitive interference, and two subjective probabilities that worries would actualize.
These latter ratings included an “emotional,” anxiety-driven probability and a “logical,”
rationally-formed probability. All of this data can inform in-the-moment GAD processes,
including the perceived costs of worries, clients’ recognition of the gap between
emotionally-informed and rationally-informed expectations, and—most importantly—the
degree of worry’s inaccuracy.
Thus, the study at hand aimed to use WOJ data to address several questions. First, we aimed
to more accurately assess how frequently GAD worries did not come true using EMA. We
also sought to report other in-journal data from the WOJ (number of unique worries,
distress, interference, etc.) at a time very early in treatment, when few to no worries had
been tested. Second, we aimed to empirically determine if disconfirmation of poor
expectations truly did predict outcome and change in GAD symptoms during treatment. Is
training a scientific, evidence-based mind actually associated with change in CT?
Understanding the processes of this common therapeutic approach can both further CT’s
empirical support and allow for its improvement and extension in clinical practice.
To this end, the current study identified worry outcome likelihoods and used bootstrapping
path analysis (Preacher & Hayes, 2008) to test the theories undergirding the WOJ in data
from its sole RCT (LaFreniere & Newman, 2016). First, we hypothesized that the percentage
of worries that did not come true would be relatively high for GAD participants (greater than
75%). Second, we expected that a higher percentage of untrue worries would predict better
treatment outcome—lower GAD symptoms post-trial and greater slopes of change in GAD
symptoms across the trial.
Method
The current study draws on data from LaFreniere and Newman’s (2016) RCT comparison of
the WOJ to an active control. It was approved by an Institutional Review Board. This
research did not receive any specific grant from funding agencies in the public, commercial,
or not-for-profit sectors.
Participants
The condition examined by this study included 29 undergraduate student participants, 26
women and 3 men. Participants met full DSM-IV criteria for GAD (which are identical to
that of DSM-5;American Psychiatric Association, 2013) as assessed by the Generalized
Anxiety Disorder Questionnaire for DSM-IV (GAD-Q-IV; Newman et al., 2002) at both a
LaFreniere and Newman Page 3
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
mass screening and during baseline assessment. Inclusion criteria required being over 18
years of age and speaking English. Participants were required to complete 80% or more
WOJ entries after two days as assessed by a compliance check, which all passed. Where
ethnicity is concerned, the sample was 75.9% White, 10.3% Black, 6.9% Hispanic, and
6.9% Asian. One participant dropped out due to discomfort with recording worries.
Although she reported no harm, she was informed of options for treatment.
Self-Report Measures
The Generalized Anxiety Disorder Questionnaire for DSM-IV
(GAD-Q-IV; Newman et al.,
2002). The GAD-Q-IV was used both to select participants for the study and to measure
general GAD symptoms at outcome. The GAD-Q-IV is a 9-item self-report measure
designed to assess the degree of GAD symptoms as defined by the
Diagnostic and Statistical
Manual of Mental Disorders
(4th ed.; APA, 1994). To its credit, the GAD-Q-IV is the only
self-report measure that evaluates all DSM GAD symptoms (Rodebaugh, Holaway, &
Heimberg, 2008). In addition to items evaluating the presence, amount, and topic matter of
excessive and uncontrollable worry, the GAD-Q-IV addresses six somatic symptoms (e.g.,
tension, restlessness, etc.), as well as distress and interference. The current study profits
from the measure’s ability to be scored both based on diagnostic criteria and dimensionally
(Moore, Anderson, Barnes, Haigh, & Fresco, 2014; Newman et al., 2002). The GAD sample
was selected using the criterion-based scoring system. Treatment effects were assessed with
its continuous score. Research has shown that both functions of the measure are valid and
reliable. Its internal consistency is robust (Cronbach’s alpha = .94; Newman et al., 2002).
Furthermore, it has shown sufficient retest reliability in clinical classification over a two
week period (92% of the sample; Newman et al., 2002). It has also demonstrated convergent
validity with measures of worry and discriminant validity against measures of social anxiety
and panic disorder. Rates of kappa agreement between the GAD-Q-IV and the Anxiety
Disorders Interview Schedule (ADIS) are good as well (κ = .67). The measure also
demonstrates sensitivity to change as a result of treatment (Andersson et al., 2012; Chen,
Liu, Rapee, & Pillay, 2013; Paxling et al., 2011)
Procedure
An online subject pool screening identified university students who met criteria for GAD
based on a criterion scoring of the GAD-Q-IV. Qualifying students were invited to
participate in the study through a mass email. Those wishing to enroll were randomly
assigned to either the WOJ or an active control condition. The current study only addresses
WOJ participants, because only these participants monitored whether or not their worries
came true.
Participants first presented at the research facility and completed the GAD-Q-IV. They then
underwent a 30-minute training session wherein study personnel read from a script as
participants followed along. Participants in both conditions were given a treatment rationale
to replicate actual cognitive-behavioral clinical practice and to control for treatment
expectancy. The rationale for the WOJ read: “Many people who regularly worry believe their
worry is useful to them. As you pay attention to how upsetting, disrupting, and costly your
worries are, and as you see clear evidence in your life that the things you worry about
LaFreniere and Newman Page 4
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
actually do not happen, you will recognize the uselessness of worrying and begin to engage
in it less. Without these anxious thoughts in your life, your anxiety should also lessen.” The
active control was also given a rationale. For the WOJ group, the training then defined worry
for participants, describing it as “a repetitive anxious thought that an event in the future will
turn out badly . . . apprehensive expectation . . . repetitive and hard to control in your mind.”
They were given several examples. Participants were also instructed that “a worry is
not
just
a fleeting concern about the future,” but rather must “forecast a negative outcome.” They
were then provided examples of thoughts that were
not
worries. They were asked to only
record specific, concrete, testable worries that they could be sure either came true or did not
come true within the 30 days of the study; they were given examples of both ideal and
improper entries. WOJ users were also taught how to track outcomes. Next, they were taught
how to complete journal ratings and how to use the online system. Lastly, they were given
the chance to ask questions, as well as contact information for any later inquiries. The entire
session script was included in the initial pages of each WOJ for participants’ reference.
The next morning participants began making prompted entries using a commercial mass
texting service. For ten days they recorded worries or thoughts in their paper journals when
prompted randomly (four times/day within four timeframes: Morning 8:00 AM-12:00 PM,
early afternoon 12:00–3:00 PM, late afternoon 3:00–6:00 PM, and evening 6:00–9:00 PM)
by text message on their personal mobile phone. They were permitted to make additional
entries at any time, but were required to record at least one at the prompted time. Recorded
worries or thoughts must have occurred within the two hours prior to the prompt. For each
recorded entry, they also noted the prompt time and date.
For WOJ-recorded worries, participants noted their specific, testable predictions about the
future. Next, they rated the degree of distress caused by the worry (worry-related distress) on
a scale of 1 (no distress) to 7 (severe distress) and their best estimate answering the question,
“How much time did that worry take up since I first had the worry?” (recent cognitive
interference duration). After completing present-moment cost ratings, they gave two
probability estimates with respect to the worry outcome. First, they rated how likely they
felt
that the feared outcome would occur, according to their “gut feelings or intuition”
(emotional likelihood of worry outcome). Next, they rated how likely “a person would
logically
conclude the worried outcome would occur if they were thinking as realistically as
possible” (logical likelihood of worry outcome). With this rating they were instructed to ask,
“If the most rational person in the world were to give a probability as to how likely this
event would come true, what would it be?” The purpose of these two ratings was to train
participants in realistic prediction formation, draw attention to the discrepancy between their
subjective forecasting and an objective prediction attempt, and to teach distance from worry
by taking the perspective of “the most rational person.” All in-journal ratings were made
first in participants’ paper journals. They were strongly encouraged to record a worry
immediately when prompted, but if they had not experienced a worry since the last prompt
time, they were allowed to write “No worry.”
Each night at 10:00 PM participants copied all content from their day’s paper recordings
into an online survey via PsychData. They also indicated the percentage of thinking time
spent on each specific worry across the entire day. Next, they were asked to review all
LaFreniere and Newman Page 5
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
recorded worries to date, to note if any of these worries had led to a new or unrecorded
outcome, and if so, if their feared outcome came true. They also rated if the outcome was “as
bad as, worse than, or better than expected.” They could note these outcomes at any time.
After two days of recording and online entry, participants received a phone call to check
compliance rates. They were asked the total number of paper and online entries made and
were assessed for any study-related harm. They were required to complete at least 80% of
their entries by this time or they were eliminated from the study. No participant had a
compliance rate below 80%. Consequently, no one was removed from the study due to low
compliance.
The day after the ten-day intervention period they completed the GAD-Q-IV again via
PsychData. Participants then returned their paper records to the lab. Twenty days after the
final day of worry recording they were provided the entire content of their journals
electronically. They were again asked to review each entry and provide the information
described above, including if any worries came true in the past 20 days and whether the
worries turned out as bad as, worse than, or better than expected. They then completed the
outcome measure again. Lastly, participants were compensated with course-specific research
credit.
Worry Outcome Coding and Journal Entry Data
Two raters independently coded worry outcome incidence for the WOJ group. Each rater
counted the number of worries recorded, the number of worries that came true, the number
of outcomes that were perceived as turning out better than expected, as expected, and worse
than expected, and the number of worrisome predictions that were untestable within 30 days
(e.g., “I will never get married”; “I won’t become an engineer”; etc.). Specific rule-based
criteria for determining what counted as a worry were established a priori. These rules were:
1) A statement must make a prediction to be counted as a worry; 2) the same worry cannot
be counted twice if it is repeated in subsequent entries; 3) if a prediction has a separate test
of whether it came true or not, then it is considered a separate worry; 4) if a new entry has
the same test of outcome as a past entry, then it is a repeated worry; 5) a statement or entry is
considered “double-barreled” if it has multiple independent predictions within it and
corresponding tests of those predictions. These multiple, testable predictions must be
counted as multiple worries. For example, an entry that reads “I will fail my Monday and
Thursday Spanish quizzes this week” counted as two worries, since it made two predictions
and had two tests of outcome. Raters were extensively trained on what counted as a worry or
certain outcome type and what did not. Prior to final coding, the primary investigator
(serving as a master coder) and the two raters each coded the data of two randomly chosen
participants. Raters’ results were compared to those of the master coder and discrepancies
were discussed and resolved. Afterward, the two raters independently coded all journal data.
Although participants were instructed to only record worries that could be tested within the
30 days of the trial, 16 worries across the entire sample did not have the chance to be tested
in the 30-day timeframe. Thus, 16 was subtracted from the total number of worries counted.
The overall percentage of testable worries that did not come true was calculated by taking
the total number of worries that did not come true divided by the total number of worries
LaFreniere and Newman Page 6
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
counted after subtracting the 16 untestable worries. Interrater reliability was high for each
result (total worries, ICC(2) = 0.920; number of worries that did not come true, ICC(2) =
0.927; worries that turned out better than expected, ICC(2) = 0.980; worries that turned out
as badly as expected, ICC(2) = 0.984; worries that turned out worse than expected, ICC(2) =
0.934). Percentages of untrue worries were also calculated for each individual participant
within their person-specific data with the same method as the overall percentage. These
values served as the “per person” percentages used in inferential analyses. Each reported
finding is the average of the two raters’ calculations.
In our descriptive results, we present both worry outcome data as well as averaged data from
participant ratings in their early WOJ entries. The early rating means, standard deviations,
and ranges presented in Table 1 are based on the average of each participant’s ratings across
the first two days of the trial (the mean of eight ratings). Ratings from the first two days
were averaged in order to capture these worry-related experiences very early in treatment—a
time when the WOJ was likely to have the least influence, with few to no worries having
actually been tested. These two-day averages may approximate normative GAD worry
experiences assessed in participants’ natural environments, but treatment influence is
possible.
2.6 Planned Statistical Analyses
Statistical analyses utilized both
R
and
Mplus 7
software. 6.9% of values were missing
overall. We first conducted paired samples t-tests comparing average worry ratings on day
one to average worry ratings on day two. These tests were meant to verify a lack of
treatment-related change in these ratings, supporting the notion that these ratings may
approximate those of normative worries. For all inferential analyses, we employed bias-
corrected bootstrapping regression-based path analysis (Preacher & Hayes, 2008) with full
information maximum likelihood (FIML) estimation to account for missingness. All
variables in all models were observed. All reported confidence intervals are 95% bootstrap
confidence intervals. Acceptable power was considered to be .80 or greater. Almost all
participants had high percentages of untrue worries near or at 100%. Consequently, we
corrected for a negatively skewed distribution with an arcsine transformation in all analyses
including untrue worry percentages, as is recommended for percentage data (Cohen &
Cohen, 1983). GAD-Q-IV scores and slopes of GAD-Q-IV change were both normally
distributed. To confirm that percentages of untrue worries were not merely a proxy for GAD
severity, we ran a bootstrapping path analysis where pre-trial GAD-Q-IV scores predicted
participants’ percentages of untrue worries. For our primary analyses, we did two separate
bootstrapping path analyses: We first regressed post-trial GAD-Q-IV scores on participants’
percentage of untrue worries during treatment with 10,000 bootstraps. We then regressed
individual slopes of change in GAD-Q-IV scores from pre- to post-trial on percentage of
untrue worries with 10,000 bootstraps. Power analysis with
G*Power
software demonstrated
acceptable power for both analyses. We used multilevel modeling in the
R
statistical
program’s
lme4
and
lmerTest
packages to extract each individual’s slope of change in GAD-
Q-IV scores from pre- to post-trial. To do so, we specified a linear mixed effects model
where GAD-Q-IV scores were predicted by a fixed effect of intercept, a random effect of
intercept, and a random effect of time trend from pre- to post-trial. The fixed effect of time
LaFreniere and Newman Page 7
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
trend was not specified so that its common influence would not be removed from
individuals’ slopes of change. To acquire participants’ slopes of change, we then extracted
the random effects of time trend on GAD-Q-IV scores for each individual. These slopes
were then used as the outcome of our predictor model. Recall that analyses presented in the
original RCT already demonstrated significant change in GAD-Q-IV scores from pre- to
post-trial (LaFreniere & Newman, 2016). We report unstandardized coefficients for path
betas. Cohen’s
d
effect sizes were calculated by the formula
d
=
B
/(√(
n
)*
SE
).
Results
Descriptive Analyses
Averaged across coders, 91.39% of all worries did not come true (91.08% for the first coder
and 91.70% for the second). In other words, 8.61% of worries came true. In regard to the
percentages calculated per person, the average percentage of untrue worries was 89.60% per
person (
SD
= 13.12%; 10.4% true worries). Per person untrue worry percentages ranged
from 53% untrue worries to 100%. The mode of the percentage of untrue worries per person
was 100%: For 7 participants, none of their worries came true. The median was 93.50%
untrue worries. Of all worries that did come true, 30.10% were rated as turning out better
than expected. The average number of unique, distinct worries per person (that had a chance
to be tested) was 34.30 (
SD
= 18.00), ranging from 12.50 to 100.50 (the total number of
unique worries did not influence any study variable of interest1). Other worry outcomes and
WOJ rating descriptive results are presented in Table 1. Average worry distress on day one
did not differ from average worry distress on day two (t(28) = 0.261,
p
= 0.796,
d
= .055).
Similarly, day one ratings did not significantly differ from day two ratings for recent
cognitive interference duration (t(28) = − 0.396,
p
= 0.695,
d
= −0.083), cognitive
interference duration across the entire day (t(28) = 0.1.06,
p
= 0.917,
d
= 0.020), emotional
likelihood of worry outcome (t(28) = 0.265,
p
= 0.793,
d
= 0.047), or logical likelihood of
worry outcome (t(28) = 0.070,
p
= 0.944,
d
= 0.010). Refer to Table 2 for zero order
correlations between study variables. In contrast to the primary study analyses, these
correlations use pairwise deletion rather than accounting for missingness with FIML. They
also do not include bias-corrected bootstrapping. Note that pre-trial GAD symptoms were
not significantly correlated with percentage of untrue worries. This result suggests untrue
worry percentage was not merely a proxy for GAD symptom or worry severity2.
Predictor Analyses
A bias-corrected bootstrap regression demonstrated that higher percentages of worries that
did not come true significantly predicted lower GAD symptoms at post-trial (
B
= −69.729,
bootstrap CI = [−165.511, −4.957],
d
= −0.372). Note that percentage of untrue worries also
1According to a bias-corrected bootstrapping path analysis with FIML estimation, the percentage of untrue worries was not related to
the number of unique worries per person (
B
= .000,
p
= .293, 95% bootstrap CI = [.000, .001]. Cohen's d could not be accurately
calculated due to an unstandardized beta coefficient of 0, yet the
R
-squared value was only 0.050. The simple Pearson correlation was
0.224 and it was not significant (
p
= .293). In addition, number of unique worries per person did not predict response to treatment (i.e.,
pre-post slope of GADQ;
B
= 0.012, CI = [−.001, .043],
d
= .018), nor was it associated with pre-trial GADQ scores (
B
=0.026, CI =
[−.022, .087),
d
= .031).
2Note that pre-trial Penn State Worry Questionnaire scores were not significantly correlated with untrue worry percentage either (
r
=
−.196,
p
= .359).
LaFreniere and Newman Page 8
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
predicted post-trial symptoms when controlling for baseline symptoms (
B
= −53.509,
bootstrap CI = [−129.395, −3.220],
d
= −0.307). A bias-corrected bootstrap regression also
demonstrated that higher percentages of untrue worries predicted greater negative slopes in
GAD symptoms from pre-trial to post-trial (
B
= −19.679, bootstrap CI = [−43.458, −0.854],
d
= −0.336). As hypothesized, participants who experienced higher percentages of their
worries not coming true had lower GAD symptoms after treatment and greater reductions in
symptoms across treatment.
Discussion
The theories supporting cognitive therapy (CT) propose that those with GAD have
inaccurate expectations and beliefs. CT is thought to reduce symptoms by correcting these
expectations with evidence. One CT intervention, the Worry Outcome Journal (WOJ),
proposes that if those with GAD track the actual outcomes of their worrisome predictions
and find their worries improbable, worry and anxiety will decrease. Yet controlled research
has never examined whether GAD expectations truly are unrealistic, or whether correcting
them predicts improvement. Drawing from an RCT of the WOJ (LaFreniere & Newman,
2016), the current study examined EMA data of WOJ users with GAD to determine the
percentage of worries that did not come true. This percentage of untrue worries was then
used to predict GAD symptoms post-trial and the change in GAD symptoms across the trial.
Results revealed that, on average, 91.4% of GAD worries did not come true. As expected,
higher percentages of untrue worries significantly predicted both lower GAD symptoms at
outcome and greater reduction in GAD symptoms across the trial. According to these
findings, the worries of those with GAD do appear to be unlikely. Moreover, greater
disconfirmation of these worries with empirical evidence predicted better treatment
outcomes. We will discuss each of these results in turn.
To begin, this study’s descriptive findings inform several important GAD phenomena.
Notably, it appears quite common that worries did not come true for WOJ users with GAD
(91.4%). In fact, the most common incidence of false worries among participants was 100%
with a median of 93.5%. Clinicians may find these figures valuable in therapy during
psychoeducation and cognitive restructuring of worry’s validity. Note also that it was
participants themselves who determined whether their worries came true or not. Positive
beliefs and the theorized functions of worry in GAD (Behar, DiMarco, Hekler, Mohlman, &
Staples, 2009) suggest people with GAD view worry as a valued means of coping. Since
those with GAD may be motivated to maintain their worry by supporting its accuracy, the
true incidence of worry non-outcomes may actually be higher than shown in participant
journals. For those worries that did come true, 30% were perceived as having turned out
better than expected. These findings partially replicate the informal report of Borkovec et al.
(1999), who found a similar nonoutcome figure of 85% when reviewing homework for 17
clients in treatment. Even so, our 91% figure is higher than their other, non-treatment
undergraduate GAD sample. In this sample, 67.9% of worries were rated as not having come
true in a questionnaire. Yet Borkovec et al.’s studies were not peer-reviewed, they did not
report the specifics of their method, they did not prompt for worry recording as in the WOJ,
and they did not use coders or a coding system. Their report does not explain how
percentages were calculated, how outcomes were reviewed and tracked, whether participants
LaFreniere and Newman Page 9
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
were truly recording testable worrisome predictions by a standardized definition, and how
repeated worries, repeated outcomes, and untestable worries were handled. It is also unclear
whether their participants were reviewing all their past worries regularly to note outcomes
(as in the WOJ) or only did so for each day’s worries or at two weeks’ time. It is likely that
without prompting and regular review, those with GAD would be selectively biased to recall
and attend more to worries that turned out poorly. Regardless, as elaborated by Borkovec et
al., the high non-occurrence of feared outcomes in their client sample and our sample
suggests frequent opportunity for negative reinforcement to promote worry. Interestingly,
that which treats worry when recognized may maintain worry when unattended. Such a
process substantiates the underlying premises of self-monitoring.
In addition to the percentage of untrue worries, other descriptive data from WOJ entries
were of interest. The average number of unique worries per person was rather high: We
found an average of 34.30 distinct worries across 10 days not counting repeated worries,
with the highest person’s count being over 100. We also report various EMA ratings of
worry experiences from very early in treatment (the first two days). The average distress
rating per worry very early in treatment was 4.51 (moderate to high). This finding is
consistent with the Contrast Avoidance Model of worry, which proposes individual worries
are associated with elevated distress (Newman & Llera, 2011). Worries also appeared to take
up significant amounts of thinking time for participants (cognitive interference), with
individual worries consuming 43.12% of the past two hours and 25.88% of the entire day on
average. Lastly, even without practice or instruction, it appears those with GAD are able to
recognize the difference between rationally-formed probabilities of future negative outcomes
and their automatic, emotional predictions. Whereas emotional probabilities for worry were
62.09% on average, participants’ attempts to make logically-formed probabilities were
41.67% on average, even in the first two days of monitoring. Compared to the actual average
likelihood of worry outcome across worries (8.6%), it is clear GAD participants have
distorted expectations both when driven by emotion and when attempting rational
forecasting. We suggest these ratings likely approximate worry experience prior to notable
treatment effects due to 1) a lack of change in ratings between day one and day two, as
demonstrated by our descriptive analyses, and 2) the low chance of worry predictions being
tested (and disconfirmed) within this short time frame.
This study was the first to test whether recognizing greater amounts of these false
expectations actually predicted better outcome for a CT intervention. Higher percentages of
untrue worries were associated with both outcome level and slope of change in GAD
symptoms across WOJ treatment. We recognize that this study does not provide direct
evidence that the accuracy of participants’ worry changed over time, even though greater
worry disconfirmation percentage did predict a change in overall worry severity and
frequency scores. However, we can reasonably infer that participants’ prediction formation
changed from the finding that worry symptoms were reduced: Worry itself is the formation
of “apprehensive expectations.” Our EMA design allowed for the examination of individual
participant expectations (worrisome predictions) and whether they were perceived as
actualized or not in real life (event outcomes). By including disconfirmed worries, analyses
were able to support CT’s theorized belief-correction process for at least one intervention.
Although beliefs considered unrealistic by clinicians have often been assessed by
LaFreniere and Newman Page 10
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
questionnaire in treatment studies (e.g., Epstein & Eidelson, 1981) and the gap between
expectations for therapy and actual therapy experience has been researched (e.g., Westra,
Aviram, Barnes, & Angus, 2010), no study has empirically derived the incidence of false
expectations in daily life for a disordered sample. The assumptions and techniques of
behavioral experiments and cognitive restructuring have been extensively applied. Yet if we
want our treatments to be empirically supported
throughout,
our studies of outcome must be
complemented by tests of such core processes (Kazdin, 2008). Our results contribute to this
end by supporting the disconfirmation strategies often used in CBT. It is recommended that
CT’s disconfirmation processes be examined with further analyses, especially those meeting
full criteria for establishing a treatment mechanism.
Unfortunately, there is limited support for
any
proposed mechanism of GAD treatment.
Although there are a few exceptions testing GAD treatment mediators generally (e.g., Arch,
Wolitzky-Taylor, Eifert, & Craske, 2012; Newman & Fisher, 2013; Newman & Fisher,
2010), none tested the disconfirmation premise of CT. One systematic review by Smits et al.
(2012) did examine threat reappraisal as a mechanism of CBT treatment for anxiety, but they
noted there were no mediation analyses of the construct for GAD. Although they found
reappraisal to be a common treatment mechanism for anxiety across studies, these studies
were almost exclusively focused on social anxiety disorder and panic disorder. Our own
review of the literature did not reveal any such studies in GAD since the time of the Smits et
al. (2012) review. Yet the few studies that
have
examined mechanisms may inform the
processes of the WOJ. Aligned with the Smits et al. (2012) review, recognition of more
worry non-outcomes may lead to beneficial reappraisal. The WOJ likely demonstrates to
participants that their worries are less threatening than previously thought, because those
worries seldom come true. This reappraisal of threat may reduce symptoms the way it has
for several other anxiety disorders in CBT trials. In addition, a study by Arch et al. (2012)
found that reductions in cognitive fusion—or “buying into” maladaptive thoughts—
mediated both CBT and acceptance and commitment therapy (ACT) for GAD.
Disconfirmation of worries may reduce the “believability” or cognitive fusion of worries,
leading to symptom improvement. This would allow for a process that is more generalized
across all worrying, rather than tediously countering singular predictions one-by-one (“I’m
surprised these few worries were so inaccurate. Maybe I should be questioning all my
worries”). Whatever the WOJ’s sub-processes may be, more extensive research on
predictors, mediators, and treatment mechanisms is certainly advisable for all GAD
therapies.
Further research on the WOJ should attend to several other questions as well. We
recommend other process predictors of the WOJ be addressed, as well as possible mediators.
First, the rationale for the WOJ suggests that recognition of in-the-moment costs of worry
promote change alongside recognition of worry likelihood. Accordingly, studies should
examine how attention toward worry costs like distress, cognitive interference, and
relationship strain influence outcomes. Second, the fact that greater untrue worry
percentages predicted greater improvement may be considered a “double-edged sword”: If
certain participants find that more of their worries actually
do
come true, treatment may be
less powerful or even harmful. If such worry confirmations were to occur, other CT
techniques may need to be applied. Possible options are reframing outcomes or directing
LaFreniere and Newman Page 11
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
attention toward client coping ability, nonoutcomes, or the ultimate harmlessness of the
situation’s consequences. Research should then inquire into the effect of augmenting the
WOJ with other CBT techniques, which can be applied in the unlikely event that a worry
comes true. Third, our participants were not given therapists with whom they could process
their WOJ data, outcomes, and maladaptive beliefs. Studying the WOJ in the context of in-
person therapy is equally important to studying it as a stand-alone EMI. Other studies may
also benefit from addressing WOJ use in other populations and disorders, other WOJ
delivery methods, addition of the WOJ EMI to ongoing face-to-face therapies, and
comparisons to other intervention processes, perhaps with moderated mediation analyses.
In designing future extensions of this study, it is important to recognize its limitations. The
most salient is our relatively small sample size of 29, although analyses were sufficiently
powered. Our population had restricted generalizability as well, being largely White, female,
young adult undergraduate students. Participants were also diagnosed by the criterion-
scoring of a self-report questionnaire, rather than a clinical interview. Furthermore, we were
not able to examine change in percentage of untrue worries. Participants did not mark the
time or day that they determined a worry was true or untrue, our coding system did not
partition worries into their time of first occurrence or day of the study, and the occurrence of
repeated worries across time made this impossible. Yet change in percentage or number of
untrue worries may have functioned as a superior predictor of outcome. As mentioned in the
original RCT, we were also unable to examine worry outcomes for worries that needed
longer periods to be tested, and did not track worries past a month’s time. Only tracking
proximal worry events may have influenced certain findings, such as the average number of
unique worries per person. Yet it is important to recognize that even monitoring outcomes of
only proximal worries was able to reduce general worrying and GAD symptoms. In general,
it would be beneficial to test the WOJ for a longer duration and examine the journal data
therein.
Despite its limitations, the results of this study suggest several clinical implications for GAD
treatment. Primarily, therapists should use techniques to draw client attention to evidence
that their worries are unrealistic, unlikely, and unhelpful. If clients can track the actual
results of their expectations, they may learn to form more adaptive, evidence-based beliefs
and predictions. Evidence that worries are unlikely may also increase client expectancy for
CBT success. As clients realize their long-held beliefs may be flawed, expectancy for a
belief-challenging therapy should increase. Therapists should capitalize on this concordance
between the verification of therapy’s predictions and the refutation of anxiety’s predictions.
Increasing client faith in the therapeutic model may then increase treatment’s effect. In
showing support for disconfirmation, our study suggests CT should continue to aim for
correcting maladaptive cognitions with renewed confidence. In cases where a client’s
worries are often shown to be (or perceived to be) true, alternatives to CT may be chosen.
Relaxation training, detachment from one’s thoughts (i.e., cognitive diffusion in ACT), or—
similarly—present moment awareness in mindfulness are all viable options. Since lower
non-outcome percentages did predict less symptom change, worry-consistent events should
be addressed—at least if they seduce clients’ focus. Therapists may target the rare
occurrences that worry did come true, engaging in reframing or identification of errors in
outcome interpretations. When worries are supposedly confirmed, therapists may discuss
LaFreniere and Newman Page 12
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
with clients whether their interpretations of outcome were accurate and whether the
consequences were truly so dire. For example, is an 85% exam score
really
confirming the
worry, “I will do poorly on the exam”? And are its repercussions
really
so bad? The results
of this study support the use of such commonly-used but understudied techniques. In the
spirit of establishing an empirical basis for informed, optimal healing, we hope such
evaluations of both old and new means of help continue long into the future.
References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th
ed.). Washington, D.C.: American Psychiatric Association.
American Psychiatric Association, A. P. A. (2013). Diagnostic and Statistical Manual of Mental
Disorders (DSM-5; 5th ed.). Arlington, Virginia: American Psychiatric Association. doi: 10.1176/
appi.books.9780890425596
Andersson G, Paxling B, Roch-Norlund P, Ostman G, Norgren A, Almlov J, Georen L, Breitholtz E,
Dahlin M, Cuijpers P, Carlbring P, & Silverberg F (2012). Internet-based psychodynamic versus
cognitive behavioral guided self-help for generalized anxiety disorder: A randomized controlled
trial. Psychotherapy and Psychosomatics, 81, 344–355. doi:10.1159/000339371 [PubMed:
22964540]
Arch JJ, Wolitzky-Taylor KB, Eifert GH, & Craske MG (2012). Longitudinal treatment mediation of
traditional cognitive behavioral therapy and acceptance and commitment therapy for anxiety
disorders. Behavior Research and Therapy, 50, 469–478. doi:10.1016/j.brat.2012.04.007
Arch JJ, & Craske MG (2008). Acceptance and commitment therapy and cognitive behavioral therapy
for anxiety disorders: Different treatments, similar mechanisms? Clinical Psychology: Science and
Practice, 15, 263–279. doi: 10.1111/j.1468-2850.2008.00137.x
Beck AT (1979). Cognitive therapy and the emotional disorders. New York: Penguin.
Beck AT, & Emery G (1985). Anxiety disorders and phobias: A cognitive perspective. New York, NY:
Basic Books.
Behar E, DiMarco ID, Hekler EB, Mohlman J, & Staples AM (2009). Current theoretical models of
generalized anxiety disorder (GAD): Conceptual review and treatment implications. Journal of
Anxiety Disorders, 23, 1011–1023. doi:10.1016/j.janxdis.2009.07.006 [PubMed: 19700258]
Berenbaum H, Thompson RJ, & Bredemeier K (2007). Perceived threat: Exploring its association with
worry and its hypothesized antecedents. Behaviour Research and Therapy, 45, 2473–2482.
doi:10.1016/j.brat.2007.03.015 [PubMed: 17512494]
Berenbaum H, Thompson RJ, & Pomerantz EM (2007). The relation between worrying and concerns:
The importance of perceived probability and cost. Behaviour Research and Therapy, 45, 301–311.
doi:10.1016/j.brat.2006.03.009 [PubMed: 16677599]
Borkovec TD, Hazlett-Stevens H, & Diaz ML (1999). The role of positive beliefs about worry in
generalized anxiety disorder and its treatment. Clinical Psychology and Psychotherapy, 6, 126–
138.
Borkovec TD, & Newman MG (1998). Worry and generalized anxiety disorder In Salkovskis P (Ed.),
Adults: Clinical formulation and treatment (Vol. 6, pp. 439–459). Oxford, England: Pergamon
Press. doi: 10.1016/B0080-4270(73)00201-7
Butler AC, Chapman JE, Forman EM, & Beck AT (2006). The empirical status of cognitive-behavioral
therapy: A review of meta-analyses. Clinical Psychology Review, 26, 17–31. doi:10.1016/
j.cpr.2005.07.003 [PubMed: 16199119]
Butler G, & Mathews A (1983). Cognitive processes in anxiety. Advances in Behaviour Research and
Therapy, 5, 51–62. doi:10.1016/0146-6402(83)90015-2
Chambless DL, & Ollendick TH (2001). Empirically supported psychological, interventions:
Controversies and evidence. Annual Review of Psychology, 52, 685–716. doi:10.1146/
Annurev.Psych.52.1.685
LaFreniere and Newman Page 13
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Chen J, Liu X, Rapee RM, & Pillay P (2013). Behavioural activation: A pilot trial of transdiagnostic
treatment for excessive worry. Behaviour Research and Therapy, 51, 533–539. doi:10.1016/
j.brat.2013.05.010 [PubMed: 23827657]
Cohen J, & Cohen P (1983). Applied multiple regression/correlation analysis for the behavioral
sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
Epstein N, & Eidelson RJ (1981). Unrealistic beliefs of clinical couples: Their relationship to
expectations, goals and satisfaction. The American Journal of Family Therapy, 9, 13–22.
doi:10.1080/01926188108250420
Hofmann SG, Asnaani A, Vonk IJ, Sawyer AT, & Fang A (2012). The efficacy of cognitive behavioral
therapy: A review of meta-analyses. Cognitive Therapy and Research, 36, 427–440. doi:10.1007/
s10608-012-9476-1 [PubMed: 23459093]
Kazdin AE (2008). Evidence-based treatment and practice: New opportunities to bridge clinical
research and practice, enhance the knowledge base, and improve patient care. American
Psychologist, 63, 146–159. doi:10.1037/0003-066X.63.3.146 [PubMed: 18377105]
LaFreniere LS, & Newman MG (2016). A brief ecological momentary intervention for generalized
anxiety disorder: A randomized controlled trial of the Worry Outcome Journal. Depression &
Anxiety, 33, 829–839. doi:10.1002/da.22507 [PubMed: 27062682]
MacLeod AK, Williams JM, & Bekerian DA (1991). Worry is reasonable: The role of explanations in
pessimism about future personal events. Journal of Abnormal Psychology, 100, 478–486.
doi:10.1037/0021-843x.100.4.478 [PubMed: 1757661]
Moore MT, Anderson NL, Barnes JM, Haigh EA, & Fresco DM (2014). Using the GAD-Q-IV to
identify generalized anxiety disorder in psychiatric treatment seeking and primary care medical
samples. Journal of Anxiety Disorders, 28, 25–30. doi:10.1016/j.janxdis.2013.10.009 [PubMed:
24334213]
Newman MG, & Fisher AJ (2013). Mediated moderation in combined cognitive behavioral therapy
versus component treatments for generalized anxiety disorder. Journal of Consulting and Clinical
Psychology, 81, 405–414. doi:10.1037/a0031690 [PubMed: 23398493]
Newman MG, Zuellig AR, Kachin KE, Constantino MJ, Przeworski A, Erickson T, & Cashman-
McGrath L (2002). Preliminary reliability and validity of the Generalized Anxiety Disorder
Questionnaire-IV: A revised self-report diagnostic measure of generalized anxiety disorder.
Behavior Therapy, 33, 215–233. doi:10.1016/S0005-7894(02)80026-0
Newman MG, & Llera SJ (2011). A novel theory of experiential avoidance in generalized anxiety
disorder: A review and synthesis of research supporting a contrast avoidance model of worry.
Clinical Psychology Review, 31, 371–382. doi:10.1016/j.cpr.2011.01.008 [PubMed: 21334285]
Newman MG, & Fisher AJ (2010). Expectancy/credibility change as a mediator of cognitive
behavioral therapy for generalized anxiety disorder: Mechanism of action or proxy for symptom
change? International Journal of Cognitive Therapy, 3, 245–261. doi :10.1521/ijct.2010.3.3.245
[PubMed: 21132075]
Paxling B, Almlöv J, Dahlin M, Carlbring P, Breitholtz E, Eriksson T, & Andersson G (2011). Guided
Internet-delivered cognitive behavior therapy for generalized anxiety disorder: A randomized
controlled trial. Cognitive Behaviour Therapy, 40, 159–173. doi:10.1080/16506073.2011.576699
[PubMed: 21770848]
Preacher KJ, & Hayes AF (2008). Asymptotic and resampling strategies for assessing and comparing
indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.
doi:10.3758/BRM.40.3.879 [PubMed: 18697684]
Rodebaugh TL, Holaway RM, & Heimberg RG (2008). The factor structure and dimensional scoring
of the Generalized Anxiety Disorder Questionnaire for DSM-IV. Assessment, 15, 343–350. doi:
10.1177/1073191107312547 [PubMed: 18202302]
Smits JA, Julian K, Rosenfield D, & Powers MB (2012). Threat reappraisal as a mediator of symptom
change in cognitive-behavioral treatment of anxiety disorders: A systematic review. Journal of
Consulting and Clinical Psychology, 80, 624–635. doi:10.1037/a0028957 [PubMed: 22686124]
Westra HA, Aviram A, Barnes M, & Angus L (2010). Therapy was not what I expected: A preliminary
qualitative analysis of concordance between client expectations and experience of cognitive-
LaFreniere and Newman Page 14
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
behavioural therapy. Psychotherapy Research, 20, 436–446. doi:10.1080/10503301003657395
[PubMed: 20552534]
LaFreniere and Newman Page 15
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
LaFreniere and Newman Page 16
Table 1
Worry outcome incidence and WOJ ratings across entries of the first two days of monitoring
M
(SD) Min. – Max.
Overall percent of worries that did not come true 91.30% --
Percent of true worries
a
with outcome better than expected 30.10% --
Percent of true worries
a
with outcome as bad as expected 44.15% --
Percent of true worries
a
with outcome worse than expected 25.80% --
Average number of unique worries per person 34.30 (18.00) 12.50 - 100.50
Early worry-related distress
b
4.51 (1.49) 3.13 – 6.75
Early recent cognitive interference duration
c
43.12% (26.08) 18.75% - 75.14%
Early daily cognitive interference duration
d
25.88% (22.57) 1.25% - 67.86%
Early emotional likelihood of worry outcome 62.09% (26.08) 25.00% - 85.71%
Early logical likelihood of worry outcome 41.67% (22.97) 7.50% - 82.86%
a
Percentage of worries that were perceived as having come true by participants during monitoring.
b
Average degree of self-reported distress associated with individual worries on a scale from 1 (
no distress
) to 7 (
severe distress
).
c
Percentage of time spent on thought in past two hours.
d
Percentage of time spent on thought across entire day.
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
LaFreniere and Newman Page 17
Table 2
Zero-order correlations between model variables (using pairwise deletion).
1 2 3 4
1. % Untrue Worries
2. GAD-Q Pre-to-Post Slope −.353
3. GAD-Q Pre-Trial −.270 .150
4. GAD-Q Post-Trial −.400
.950
**
.452
*
Note:
GAD-Q = Generalized Anxiety Disorder Questionnaire for DSM-IV-TR;
=
p
< .10.
*
=
p
< .05.
**
=
p
< .001.
Behav Ther
. Author manuscript; available in PMC 2020 May 18.
... This is the case even though worry does not actually help people problem-solve (e.g., Llera & Newman, 2020) and may interfere with learning (LaFreniere & Newman, 2019). Also, pathological worriers continue to worry despite repeatedly experiencing the nonoccurrance of their feared outcomes (LaFreniere & Newman, 2020). The contradiction between positive views of worry and its reality has led to the development of theories to explain what motivates and maintains such worry. ...
... A PEC can be defined as a sharp decrease in negative affect and/or increase in positive affect. In terms of PECs involving decreased negative emotion, worrisome outcomes occurred less than 9% of the time (LaFreniere & Newman, 2020). ...
... However, clarifying the function of positive, as well as negative emotions, in the reinforcement and perpetuation of pathological worry is essential to effectively intervening on each of the emotion-based mechanisms of worry. CAM theorizes that worry may also lead to acute brief increases in positive emotion when things turn out better than expected, which happens most of the time due to the unrealistic nature of most worries (LaFreniere & Newman, 2020). This could also regularly reinforce worrying. ...
Article
Full-text available
Background: The Contrast Avoidance Model (Newman & Llera, 2011) proposes that worry is reinforced by avoiding a negative contrast and increasing the likelihood of a positive contrast. Objective: To determine if reinforcement of worry occurs naturalistically via contrasts in both negative and positive emotion. Method: Using event-contingent momentary assessment we assessed social interactions, pre-interaction state worry and pre-post interaction positive and negative emotion. Participants with generalized anxiety disorder (GAD; N = 83) completed an online questionnaire after social interactions lasting at least 1 min for 8 days. Three-level multilevel models were conducted. Results: Higher worry was concurrently associated with increased negative emotion and decreased positive emotion. Regardless of pre-interaction worry level, negative emotion decreased, and positive emotion increased from before to after interactions, suggesting that most interactions were benign or positive. At lower levels of pre-interaction worry, participants experienced increased negative emotion and decreased positive emotion from before to after interactions. At higher levels of pre-interaction worry, participants experienced decreased negative emotion and increased positive emotion from before to after interactions. Conclusion: Among persons with GAD, worrying before social interactions may be both negatively and positively reinforced; furthermore, not worrying before social interactions may be both negatively and positively punished.
... Despite this, as specified by the free energy principle, the system is driven toward minimizing free energy. The biological system therefore expects that its own modeling and action policies are neither sufficient nor insufficient-and yet still necessary-and generates the belief that the inherent uncertainty will persist across time (see Dickstein et al., 2010;Grupe and Nitschke, 2013;Fonzo and Etkin, 2016;Kannis-Dymand et al., 2020;LaFreniere and Newman, 2020). External uncertainty therefore generates internal uncertainty about external certainty, along with the necessary action policies that would resolve it. ...
Article
Full-text available
Generalized anxiety disorder is among the world's most prevalent psychiatric disorders and often manifests as persistent and difficult to control apprehension. Despite its prevalence, there is no integrative, formal model of how anxiety and anxiety disorders arise. Here, we offer a perspective derived from the free energy principle; one that shares similarities with established constructs such as learned helplessness. Our account is simple: anxiety can be formalized as learned uncertainty. A biological system, having had persistent uncertainty in its past, will expect uncertainty in its future, irrespective of whether uncertainty truly persists. Despite our account's intuitive simplicity-which can be illustrated with the mere flip of a coin-it is grounded within the free energy principle and hence situates the formation of anxiety within a broader explanatory framework of biological self-organization and self-evidencing. We conclude that, through conceptualizing anxiety within a framework of working generative models, our perspective might afford novel approaches in the clinical treatment of anxiety and its key symptoms.
... Horror also provides opportunities for clients to propose coping strategies for dealing with threat, if it occurs. Although the vast majority of worries for people with GAD do not come true (LaFreniere & Newman, 2020), there are circumstances in which anxiety accurately warns people of potential threats and times in which what previously had been considered "less probable" suddenly becomes salient. In these cases, simulation with the threatening situation could enhance resilience and coping with the situation. ...
Preprint
Full-text available
From scary stories to horror films and haunted houses, the horror genre is wildly popular. Although horror aims to elicit fear and anxiety in its audience, many people with anxiety are horror fans and some report using horror to cope with their anxiety. In this article, we provide a theoretical rationale for why people with anxiety might find relief in horror films. First, we discuss aspects of horror that could make it particularly alluring to people with anxiety and how some features of horror films have the potential to reduce anxiety in the moment. Next, we examine how engagement with horror could build skills for resilience in more generalized situations. We build on processes from evidence- based therapies (i.e., cognitive behavioral therapy and exposure therapy) to explain how horror media has the potential to be used as a therapeutic tool. Finally, we discuss steps for future research on horror as a therapeutic tool for anxiety-related disorders.
... The most common worry topics seen in persons with GAD are not different from what we all worry about (work, health, family), but are more frequent and intense compared with the worry observed in non-clinical populations Olatunji, Wolitzky-Taylor, Sawchuk, & Ciesielski, 2010). A study by LaFreniere and Newman (2020) investigating the nature of worry in GAD showed that the average number of unique worries during a ten-days period was 34.3 for GAD patients. Furthermore, on average, 89.60% of the worries did not come true during a 30-day period. ...
... Studies have revealed that the majority of worries of people with GAD do not come true. The percentage of untrue worries in two different studies was calculated as 85-91% (Borkovec et al., 1999) and 91.4% (LaFreniere & Newman, 2019a). Thus, the possibility of experiencing positive emotional contrast in these individuals is very high. ...
Article
Full-text available
The Contrast avoidance model (CAM) suggests that individuals with generalized anxiety disorder (GAD) avoid negative emotional contrasts (shifts) by creating and sustaining negative emotions through worry, and the main fear in these individuals is negative emotional contrasts. The purpose of the present study is to review studies on the CAM and its underlying assumptions, studies supporting and criticizing the model, and emotional factors related to the development of contrast avoidance (CA) tendencies with a focus on the transdiagnostic nature of the CAM and respective principles. Moreover, we reviewed a two-pronged treatment method based on the CAM. Finally, we proposed several research suggestions and emphasized on the importance of studying different aspects of the CAM in order to support its applicability to emotional disorders other than GAD.
... This is reflected in a habitual tendency to predict unreasonable or exaggerated future events (regardless of their actual probability), often by asking 'what if' something bad was to occur and underestimating his or her ability to cope with adversity. The problem is that worrying about low probability events (91% of worry predictions among 'worriers' never happens [30]) is negatively reinforced by their very nonoccurrence. To illustrate, the IBS patient who worries that s/he will have an accident at a restaurant is more likely to worry intensely in the future when she does not have an accident because the relief that she experiences when her worry does not happen actually reinforces worry and anxiety due to a conditioning principle called negative reinforcement. ...
Article
Introduction Irritable bowel syndrome is a common, painful and often disabling GI disorder for which there is no satisfactory medical or dietary treatment. The past 10 years has seen the development and validation of a number of psychological treatments of which CBT is arguably the most effective based on two recently conducted multiple site trials from two investigative teams in the UK and USA. Areas covered The purpose of this review is to describe the principles, processes, procedures, and empirical basis supporting CBT and distinguish it from other psychological treatments available to clinical GE whose patients suffer from refractory IBS. Expert commentary The efficacy of CBT in treating refractory IBS patients is well established but there is limited understanding of why it works and for whom it is most beneficial. Further, its availability is generally limited to tertiary care settings which may undermine its value proposition if improved self-management is not accompanied by other health care efficiencies. Systematic efforts to increase both the efficiency of CBT and the way it is delivered (e.g., digital therapeutics, primary care) is critical to optimizing CBT’s potential and reducing the public health burden IBS imposes.
Article
More and more research has focused on the role of the cerebellum in emotions and social cognition. Structural cerebellar and cerebello-cerebral connectivity abnormalities have been identified in several prevalent neuropsychiatric conditions, which have in some cases even been linked to the severity of the emotional disorder.Non-invasive brain stimulation (NIBS) techniques are currently used to modulate neuronal excitability and tune the connectivity within and between neuronal networks. Targeting the cerebellum with NIBS in order to improve emotions and social behavior in neuropsychiatric conditions seems to be a very interesting and innovative approach. Several studies have already explored the effect of cerebellar vermis stimulation in patients with schizophrenia with promising results. Other neuropsychiatric disorders such as bipolar disorder (BD), obsessive-compulsive disorder (OCD), major depressive disorder, or generalized anxiety disorder (GAD) have received less attention with respect to cerebellar stimulation, although the cerebellum has been implicated in these disorders. We will address NIBS and neuropsychiatric disorders in this chapter. Future research should focus on combining cerebellar NIBS with neuroimaging to unravel the specific role of the cerebellum in emotional disorders. Such studies will be very valuable in establishing causal relationships between the structural and functional abnormalities that can be observed in these disorders, and in the search for neurophysiological biomarkers for emotions. However, it is still unclear which stimulation parameters are optimal. Moreover, an important factor to consider when applying cerebellar NIBS in order to improve emotional or other functioning is cerebellar reserve. Although the cerebellum has a wide variety of plasticity mechanisms and its structural organization intrinsically incorporates a lot of redundancy, this redundancy can be depleted. A certain amount of cerebellar reserve should be preserved to successfully apply NIBS.Systematic studies are therefore needed to clarify the optimal stimulation parameters, and methods should be developed to quantify cerebellar reserve in order to estimate the possible added value of NIBS in the rehabilitation of emotions.
Article
Positive beliefs about worry are an important factor that has been shown to be associated with the reduction of worry severity with Cognitive Behavioral Therapy. The present study evaluated the psychometric properties of a measure of positive beliefs about worry, the Why Worry Questionnaire II (WW-II; Hebert et al., 2014) with a clinical sample. The present study also compared mean scores on the WW-II in the present clinical sample with scores found in non-clinical samples. Finally, the study compared mean scores on the WW-II between groups with primary diagnoses of Generalized Anxiety Disorder (GAD), Major Depressive Disorder (MDD), and Anxiety Disorder Not Otherwise Specified (NOS). The confirmatory factor analysis found the five-factor model an adequate to good fit to the data, and the WW-II demonstrated excellent internal consistency within this clinical sample. Additionally, scores on the WW-II in the present sample were significantly higher than scores found by Hebert et al. (2014) in their non-clinical sample. Finally, no significant mean differences were found between primary diagnoses of GAD, Anxiety Disorder NOS, or MDD. Important theoretical and clinical implications and suggestions for future research are discussed.
Chapter
Full-text available
Generalized anxiety disorder (GAD) is a widespread and debilitating disorder. In this review, we present updated research on worry, along with the emotional, information processing, interpersonal, and biological factors underlying this disorder, through the lens of a comprehensive model of GAD: the Contrast Avoidance Model. This model sheds light on the developmental and maintaining factors of this disorder, and provides new insights for treatment. We also review research on the efficacy of extant treatments of GAD, along with presenting new directions for future research.
Article
Full-text available
Introduction: Generalized anxiety disorder (GAD) is prevalent among college students. Smartphone-based interventions may be a low-cost method of treatment. Method: College students with self-reported GAD were randomized to receive smartphone-based guided self-help (n = 50), or no treatment (n = 50). Post-treatment and six-month follow-up outcomes included the Depression Anxiety Stress Scales-Short Form Stress Subscale (DASS Stress), the Penn State Worry Questionnaire (PSWQ-11), and the State-Trait Anxiety Inventory-Trait (STAI-T), as well as diagnostic status assessed by the GAD-Questionnaire, 4th edition. Results: From pre- to posttreatment, participants who received guided self-help (vs. no treatment) experienced significantly greater reductions on the DASS Stress (d = -0.408) and a greater probability of remission from GAD (d = -0.445). There was no significant between-group difference in change on the PSWQ-11 (d = -0.208) or STAI-T (d = -0.114). From post to six-month follow-up there was no significant loss of gains on DASS stress scores (d = -0.141) and of those who had remitted, 83.3% remained remitted. Yet rates of remitted participants no longer differed significantly between conditions at follow-up (d = -0.229). Conclusion: Smartphone-based interventions may be efficacious in treating some aspects of GAD. Methods for improving symptom reduction and long-term outcome are discussed.
Article
Full-text available
This review summarizes the current meta-analysis literature on treatment outcomes of CBT for a wide range of psychiatric disorders. A search of the literature resulted in a total of 16 methodologically rigorous meta-analyses. Our review focuses on effect sizes that contrast outcomes for CBT with outcomes for various control groups for each disorder, which provides an overview of the effectiveness of cognitive therapy as quantified by meta-analysis. Large effect sizes were found for CBT for unipolar depression, generalized anxiety disorder, panic disorder with or without agoraphobia, social phobia, posttraumatic stress disorder, and childhood depressive and anxiety disorders. Effect sizes for CBT of marital distress, anger, childhood somatic disorders, and chronic pain were in the moderate range. CBT was somewhat superior to antidepressants in the treatment of adult depression. CBT was equally effective as behavior therapy in the treatment of adult depression and obsessive-compulsive disorder. Large uncontrolled effect sizes were found for bulimia nervosa and schizophrenia. The 16 meta-analyses we reviewed support the efficacy of CBT for many disorders. While limitations of the meta-analytic approach need to be considered in interpreting the results of this review, our findings are consistent with other review methodologies that also provide support for the efficacy CBT.
Article
Full-text available
This study examined (a) duration of generalized anxiety disorder (GAD) as a moderator of cognitive behavioral therapy (CBT) versus its components (cognitive therapy and self-control desensi- tization) and (b) increases in dynamic flexibility of anxious symptoms during the course of psychother- apy as a mediator of this moderation. Degree of dynamic flexibility in daily symptoms was quantified as the inverse of spectral power due to daily to intradaily oscillations in four-times-daily diary data (Fisher, Newman, & Molenaar, 2011). Method: This was a secondary analysis of the data of Borkovec, Newman, Pincus, and Lytle (2002). Seventy-six participants with a principle diagnosis of GAD were assigned randomly to combined CBT (n = 24), cognitive therapy (n = 25), or self-control desensitization (n = 27). Results: Duration of GAD moderated outcome such that those with longer duration showed greater reliable change from component treatments than they showed from CBT, whereas those with shorter duration fared better in response to CBT. Decreasing predictability in daily and intradaily oscillations of anxiety symptoms during therapy reflected less rigidity and more flexible responding. Increases in flexibility over the course of therapy fully mediated the moderating effect of GAD duration on condition, indicating a mediated moderation process. Conclusions: Individuals with longer duration of GAD may respond better to more focused treatments, whereas those with shorter duration of GAD may respond better to a treatment that offers more coping strategies. Importantly, the mechanism by which this moderation occurs appears to be the establishment of flexible responding during treatment.
Article
Full-text available
Objective: This study examined (a) duration of generalized anxiety disorder (GAD) as a moderator of cognitive behavioral therapy (CBT) versus its components (cognitive therapy and self-control desensitization) and (b) increases in dynamic flexibility of anxious symptoms during the course of psychotherapy as a mediator of this moderation. Degree of dynamic flexibility in daily symptoms was quantified as the inverse of spectral power due to daily to intradaily oscillations in four-times-daily diary data (Fisher, Newman, & Molenaar, 2011). Method: This was a secondary analysis of the data of Borkovec, Newman, Pincus, and Lytle (2002). Seventy-six participants with a principle diagnosis of GAD were assigned randomly to combined CBT (n = 24), cognitive therapy (n = 25), or self-control desensitization (n = 27). Results: Duration of GAD moderated outcome such that those with longer duration showed greater reliable change from component treatments than they showed from CBT, whereas those with shorter duration fared better in response to CBT. Decreasing predictability in daily and intradaily oscillations of anxiety symptoms during therapy reflected less rigidity and more flexible responding. Increases in flexibility over the course of therapy fully mediated the moderating effect of GAD duration on condition, indicating a mediated moderation process. Conclusions: Individuals with longer duration of GAD may respond better to more focused treatments, whereas those with shorter duration of GAD may respond better to a treatment that offers more coping strategies. Importantly, the mechanism by which this moderation occurs appears to be the establishment of flexible responding during treatment.
Article
Background: The efficacy of many cognitive behavioral component interventions has not been examined, with worry outcome monitoring among them. Methods: To address this issue, 51 participants with clinical levels of generalized anxiety disorder were randomly assigned to a treatment or control condition for 10 days. The treatment condition consisted of a brief ecological momentary intervention termed the Worry Outcome Journal (WOJ). WOJ participants recorded worries and tracked their outcomes, rating worry distress, interference, and expected outcome probabilities. Thought log (TL) control participants completed a record of their everyday thoughts and rated associated distress. All participants made four entries on paper each day when randomly prompted by text message. They then entered their paper contents online each night. After 30 days they reviewed their contents electronically and completed follow-up measures. Results: Primary results revealed significant reductions in worry for WOJ users compared to TL users at postintervention. A marginally significant difference was found at 20-day follow-up and treatment gains were maintained. Secondary analyses showed no harmful increases in worry beliefs for WOJ users, as well as preliminary evidence for decreases in beliefs about the uncontrollability of thoughts in both groups. Conclusion: The WOJ may be a viable therapist-independent treatment for reducing worry, even after only 10 days of use.
Article
Forty-seven marital therapy couples completed inventories measuring unrealistic beliefs about self and unrealistic beliefs about marital relationships. In addition, they completed questionnaire measures of their expectations and goals for therapy and their levels of marital satisfaction. As hypothesized, the clients' unrealistic beliefs, particularly those regarding relationships, were negatively associated with their estimated chance for improvement in therapy, desire to improve rather than terminate the relationship, preference for marital versus individually oriented treatment, and overall marital satisfaction. These results are consistent with the theoretical rationale for cognitive therapy with clinical couples and suggest specific targets for intervention in this process. Directions for future research are discussed.
Article
Transdiagnostic interventions present pragmatic benefits in treatment dissemination and training of mental health professionals when faced with emotional disorders such as anxiety and depression. Excessive worry is a common feature across emotional disorders and represents an ideal candidate target for transdiagnostic intervention. The current pilot trial examined the efficacy of a behavioural activation treatment for worry (BAW) in a community population. 49 individuals experiencing excessive worry were randomised to waitlist or BAW receiving an 8 week group based intervention. Results demonstrated that BAW was successful in reducing excessive worry, depressive symptoms, cognitive avoidance, Intolerance of Uncertainty and improving problem solving orientation. Twice as many individuals showed clinically significant reductions in excessive worry after treatment compared to the waitlist control. Despite limitations to sample size and power, this study presents promising support for BAW as a practical transdiagnostic treatment for worry.