Behaviour Research and Therapy 58 (2014) 65e74

Contents lists available at ScienceDirect

Behaviour Research and Therapy

j o u r n a l h o m e p a g e : w w w.e l s e v i e r.c o m / l o c a t e / b r a t

Cognitive-behavioral therapy for hypochondriasis/health anxiety:

A meta-analysis of treatment outcome and moderators
Bunmi O. Olatunji a, *, Brooke Y. Kauffman b, Sari Meltzer a, Michelle L. Davis
Jasper A.J. Smits b, Mark B. Powers b
a
b

Department of Psychology, Vanderbilt University, United States

Department of Psychology, University of Texas-Austin, United States

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 18 March 2014
Received in revised form
12 May 2014
Accepted 16 May 2014
Available online 24 May 2014

The present investigation employed meta-analysis to examine the efcacy of cognitiveebehavioral

therapy (CBT) for hypochondriasis/health anxiety as well as potential moderators that may be associated
with outcome. A literature search revealed 15 comparisons among 13 randomized-controlled trials
(RCTs) with a total sample size of 1081 participants that met inclusion criteria. Results indicated that CBT
outperformed control conditions on primary outcome measures at post-treatment (Hedges's g 0.95)
and at follow-up (Hedges's g 0.34). CBT also outperformed control conditions on measures of
depression at post-treatment (Hedges's g 0.64) and at follow-up (Hedges's g 0.35). Moderator analyses revealed that higher pre-treatment severity of hypochondriasis/health anxiety was associated
with greater effect sizes at follow-up visits and depression symptom severity was signicantly associated
with a lower in effect sizes at post-treatment. Although effect size did not vary as a function of blind
assessment, smaller effect sizes were observed for CBT vs. treatment as usual control conditions than for
CBT vs. waitlist control. A dose response relationship was also observed, such that a greater number of
CBT sessions was associated with larger effect sizes at post-treatment. This review indicates that CBT is
efcacious in the treatment of hypochondriasis/health anxiety and identies potential moderators that
are associated with outcome. The implications of these ndings for further delineating prognostic and
prescriptive indicators of CBT for hypochondriasis/health anxiety are discussed.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Hypochondriasis
Health anxiety
Cognitive-behavioral therapy
Moderation

The central feature of hypochondriasis is a preoccupation with

the inaccurate belief that one has, or is in danger of developing, a
serious medical condition based on misinterpretations of benign
(or minor) bodily sensations (American Psychiatric Association
[APA], 2000). Research has shown that up to 9% of patients in
general medical practice clinics and up to 5% of the general population meets diagnostic criteria for hypochondriasis (Creed &
Barsky, 2004; Gureje, Ustun, & Simon, 1997). The heightened
prevalence in medical practice may reect the fact that hypochondriasis is also characterized by a strong disease conviction
that persists despite appropriate medical evaluation. This preoccupation with medical illness often focuses on specic signs or
symptoms (e.g., sore throat), diseases (e.g., cancer), organs (e.g.,
heart), or vaguely dened somatic phenomena (e.g., my aching
veins). Typically, patients with hypochondriasis attribute
* Corresponding author. Department of Psychology, Vanderbilt University, 301
Wilson Hall, 111 21st Avenue South, Nashville, TN 37203, United States.
E-mail address: olubunmi.o.olatunji@vanderbilt.edu (B.O. Olatunji).

http://dx.doi.org/10.1016/j.brat.2014.05.0 02
0005-7967/ 2014 Elsevier Ltd. All rights reserved.

unwanted bodily sensations to possible disease (e.g., this headache

means I have a brain tumor) and are highly concerned with their
authenticity. Perhaps the most readily observable sign of hypochondriasis is the persistent attempt to seek reassurance about the
feared symptoms or illness. Individuals with this condition may
repeatedly contact doctors, seek additional tests, scour Internet
sites and medical texts, and seek reassurance from signicant
others about bodily sensations which have been appropriately
evaluated and judged to be benign. This preoccupation with
disease can be disruptive to social, occupational, and family
functioning, and is associated with substantial economic costs
(Katon & Walker,
1998).
Although hypochondriasis has historically been viewed as a
somatoform disorder (APA, 2000), the validity of this
categorization has not been without debate. It has been argued
that hypochon- driasis may be best conceptualized as an anxiety
disorder (Olatunji, Deacon, & Abramowitz, 2009). This view
suggests that health anxiety represents a continuous dimension
(ranging from no health anxiety to severe health anxiety) with
hypochondriasis at its

B.O. Olatunji et al. / Behaviour Research and Therapy 58 (2014) 65e74

clinical endpoint (Marcus, Gurley, Marchi, & Bauer, 2007). This

argument is based largely on empirical observations that symptoms of hypochondriasis overlap with certain anxiety disorders:
namely, panic disorder (PD) and obsessive-compulsive disorder
(OCD). Like those with hypochondriasis, patients with PD are
hypervigilant to benign, arousal-related body sensations and often
erroneously attribute them to organic causes such as heart attacks,
strokes, and other serious medical conditions (Abramowitz, 2005;
Barsky, Barnett, & Cleary, 1995). Similarities have also been
observed between hypochondriasis and OCD in terms of preoccupation with health and disease, and the repetitive and pervasive
nature of such preoccupation (Abramowitz, 2005; Fallon, Javitch,
Hollander, & Liebowitz, 1992). Much like PD and OCD, cognitivebehavioral models (Abramowitz, Schwartz, & Whiteside, 2002;
Warwick & Salkovskis, 1990) posit that hypochondriasis is an
extreme form of health anxiety that emerges from the misinterpretation of benign and normally occurring experiences that lead
to anxiety and the use of safety behaviors which paradoxically
maintains the anxiety (Abramowitz & Moore, 2007; Abramowitz,
Olatunji, & Deacon, 2007).
The cognitive-behavioral model of hypochondriasis is depicted
in Fig. 1 (Abramowitz, Deacon, & Valentiner, 2007). This model
suggests that dysfunctional beliefs about bodily symptoms and
illness play a signicant role in the development of hypochondriasis. Such beliefs often increase the likelihood of having catastrophic cognitions when exposed to benign bodily symptoms or
health-related information. Once concerned about the possibility
of acquiring an illness, the person becomes vigilant for any signs of
being ill and is motivated to reduce their worry by gaining
certainty

about their health status. It is important to note that this process in

hypochondriasis is driven largely by mistaken beliefs about illnesses. According to the cognitive-behavioral model, mistaken
beliefs in hypochondriasis are maintained (despite contradictory
information and repeated reassurance of good health from medical
professionals) by maladaptive strategies used to cope with healthrelated anxiety. These strategies include attempts to prevent the
feared illness, avoidance, and attempts to attain certainty about
health status. These safety behaviors prevent individuals with hypochondriasis from acquiring information that would disconrm
their mistaken beliefs about illnesses. This cognitive-behavioral
model of hypochondriasis has been empirically supported and
proven to be clinically useful (Taylor & Asmundson, 2004). Based in
part on these observations, the newly published DSM-5 has
replaced the diagnosis of hypochondriasis with illness anxiety
disorder (APA, 2013).
Based on the ndings of early, predominantly psychodynamic
interventions, hypochondriasis has historically been regarded as
resistant to psychological treatment (Taylor, Asmundson, & Coons,
2005). In fact, treatments were initially considered to be of limited
value for hypochondriasis (Ladee, 1966). Extreme interventions
were considered and some clinicians even employed prefrontal
lobotomies for the treatment of hypochondriasis (Bernstein,
Callahan, & Jaranson, 1975). Fortunately, the emergence of a
cognitive-behavioral model of hypochondriasis has paved the way
for the application of a more appropriate approach to treatment.
The cognitive-behavioral approach is derived largely from the
observation that symptoms of hypochondriasis d at both a topographical and functional level d overlap remarkably with certain

Fig. 1. Cognitive-behavioral model of hypochondriasis outlined by Abramowitz, Deacon, et al. (2007).

anxiety disorders (Olatunji et al., 2009). As with other anxiety

disorders, this approach emphasizes the role of dysfunctional beliefs in maintaining hypochondriasis. This approach also suggests
that avoidance and health-related safety behaviors prevent
dysfunctional beliefs from being disconrmed, and thereby
perpetuate excessive health anxiety and somatic preoccupation
(Warwick & Salkovskis, 2001). This functional analysis of hypochondriasis would then suggest that cognitive behavioral therapy
(CBT) may provide much needed symptom relief. CBT here refers to
the class of interventions that are based on the basic premise that
emotional disorders are maintained by cognitive and behavioral
factors, and that psychological treatment leads to changes in these
factors through cognitive and behavioral techniques (Beck &
Emery, 2005).
Research on the effects of CBT for hypochondriasis has produced
encouraging results. In one study, CBT was found to be superior to
no treatment in reducing health anxiety, the need for reassurance,
and the frequency of checking behavior (Warwick, Clark, Cobb, &
Salkovskis, 1996). A subsequent study found that patients
receiving cognitive therapy showed signicantly more reduction
in symptoms of hypochondriasis at mid-treatment and at posttreatment compared to those receiving a stress management
intervention (Clark et al., 1998). Given that research has shown
that patients with excessive health anxiety have a preference for
psy- chological treatments over drug treatments (Walker,
Vincent, Furere, Cox, & Kjernisted, 1999), a better understanding
of the efcacy of CBT may provide a solid empirical foundation for making it
the gold standard treatment for hypochondriasis. Meta-analysis is
the primary means by which researchers have synthesized the
results from multiple treatment trials examining the efcacy of
various treatments. Although the use of meta-analytic data is not
without limitations, this approach has offered very preliminary
insight into delineating the general effectiveness of CBT in the
treatment of hypochondriasis. An initial meta-analytic review of
the literature concluded that CBT is the most effective treatment for
hypochondriasis (Taylor et al., 2005). However, this investigation
examined only 5 studies utilizing CBT and included uncontrolled
comparisons in the analysis. Thomson and Page's (2007) more
recent Cochrane review, which included only six studies, also
concluded that CBT was the most effective psychological treatment
for hypochondriasis. However, several limitations of the existing
trial data at the time of the publication were noted. For example,
the reliance on wait-list control conditions in many trials makes it
difcult to rule out the possibility that improvements after CBT
were due to nonspecic factors.
Although much remains unknown about the incremental efcacy of CBT for hypochondriasis, even less is known about reliable predictors of treatment outcome. For example, some research
suggests that comorbid psychiatric disorders are poor prognostic
indicators (Barsky, Fama, Bailey, & Ahern, 1998; Noyes et al.,
1994). However, other studies report that comorbid anxiety
and depression is either related to good outcome (House, 1989)
or unrelated to outcome (Greeven et al., 2009). Research has
also shown that younger age is associated with good outcome
(Barsky,
1996; House, 1989) while other studies have failed to observe this
association (Speckens, Spinhoven, van Hemert, Bolk, & Hawton,
1997). Taylor et al. (2005) suggest that the inconsistent ndings
with regards to predictors of treatment outcome in hypochondriasis may be because some variables predict outcome for some
treatments but not others. It may also be the case that the
inconsistent ndings reect the fact that the variables are weak
predictors of outcome. Yet another possibility is that individual
studies lack the power necessary to detect predictors of treatment
outcome. To the extent that this is the case, meta-analysis of many
trials may prove useful in clarifying some of these discrepancies.

Clarifying the substantive predictors of treatment outcome may

have important implications for treatment planning in CBT. For
example, if it is the case that comorbidity predicts worse treatment outcome, this may be justication for the use of adjunctive
treatments to CBT that more directly target the comorbid conditions. This may also have implications for the time course of CBT
for hypochondriasis. Among those with more severe comorbid
conditions, implementation of CBT may be difcult, thereby
interfering with effective treatment. Treatment of comorbid disorders before initiating CBT for hypochondriasis in such cases may
then maximize outcome.
In addition to clarifying predictors of treatment outcome,
examining the extent to which sample-specic characteristics (e.g.,
gender) and study-specic characteristics (e.g., number of treatment sessions) predict outcome in CBT for hypochondriasis may
also prove to be informative. Indeed, it has been observed that the
question towards which all outcome research should ultimately be
directed is the following: What treatment, by whom, is most
effective for this individual with that specic problem, and under
which set of circumstances? (Paul, 1967, p. 111). Although no one
study can address such a complicated question (Beutler, 1991),
examination of moderators in the context of meta-analysis of
treatment outcome studies examining the efcacy of CBT for hypochondriasis may be an important preliminary step in beginning
to address this question. Accordingly, the present investigation
employs a meta-analytic approach to examine the efcacy of CBT
for hypochondriasis. This investigation is especially timely given
the limitations of the prior two meta-analyses CBT for hypochondriasis, one of which consisted of only 5 CBT studies and included
uncontrolled comparisons (Taylor et al., 2005) and the other which
was limited to only 6 studies (Thomson & Page, 2007). In addition,
this investigation builds on prior work by examining moderators of
treatment outcome. Four questions derived from the existing
literature were examined.
1. Do CBT treatments outperform control conditions on hypochondriasis/health anxiety outcome measures at posttreatment and at follow-up?
2. Do CBT treatments outperform control conditions on depression
severity measures at post-treatment and follow-up?
3. Does higher pre-treatment hypochondriasis/health anxiety
severity and higher pre-treatment depression scores predict
lower between-group effect sizes at post-treatment and followup?
4. Does comparator treatment/condition, participant age, percentage of female participants, percentage of comorbidity,
number of sessions, inclusion of treatment integrity checks and/
or blind assessors (i.e., clinical trial quality) moderate treatment
effect sizes?

Method
Study selection
Randomized controlled trials (RCTs) of CBT (see inclusion
criteria below) for hypochondriasis/health anxiety were selected
using a comprehensive search strategy. A search was conducted in
PsycINFO and MEDLINE (1966 to March 2014). The searches
included the following terms: cognitive behavioral or cognitive
behavioral therapy, and clinical trial or trial alone and in
combination with hypochondriasis or health anxiety. These
words were searched as key words, title, abstract, and Medical
Subject Headings. In addition, we contacted authors of CBT trials for
emerging publications.

Fig. 2. Study selection and reasons for exclusions.

As depicted in Fig. 2, the initial search strategies produced 2988

potential articles. Examination of the abstracts identied 25
relevant and non-repeating studies. Inclusion criteria for the
meta-analysis were as follows: (a) participants who met full DSMIII-R, DSM-IV, or DSM-IV-TR criteria for hypochondriasis or had
clinical levels of health anxiety; (b) adequate control condition
(psychological pla- cebo, wait-list control, or pill placebo); and (c)
more than one session of CBT. Treatments were classied as CBT if
they included cognitive techniques (e.g. cognitive restructuring,
behavioral experiments, etc.), behavioral techniques (e.g. in-vivo
exposure, imaginal expo- sure, etc.), or a combination of these
strategies. Acceptance-based therapies (e.g.,
ACT), which
emphasize engaging in valued behav- iors despite anxiety, were
included if they also utilized cognitive or behavioral techniques.
Exclusion criteria for the meta-analysis were: (a) single case
studies; (b) treatment conditions aimed at augmentating a
psychological treatment; (c) studies with insuf- cient data,
unless study authors were able to provide such data; and (d)
studies with redundant data. Of the 25 identied studies, 14 were
excluded based on these criteria. Two additional studies were then
identied after reviewing the references of the remaining studies.
Table 1 shows the 15 comparisons within the 13 studies with a
total sample size of 1081 participants that met the nal inclusion
criteria and were included in the meta-analyses.
Software
All analyses were completed with Comprehensive MetaAnalysis (Bornstein & Rothstein, 1999).
Procedure
Data on the following variables were collected when reported
in each of the 13 included studies: treatment conditions (13
studies), control type (13 studies), treatment dose (number of
sessions; 13

studies), number of participants (13 studies), mean age (11

studies), and percentage of female participants (12 studies).
Treatment integrity variables, including inclusion or exclusion of
treatment integrity checks and blind assessors, were collected
categorically and assumed to be absent if not reported in the
manuscript. Dependent variables were labeled as either primary
or secondary outcome
measures as dened by the authors in the original
publication.1
Control conditions were classied into three categories: placebo, treatment as usual, or wait-list. Treatments categorized as
placebo were treatments that did not include the cognitive and
behavioral components of CBT, including: attention control, relaxation, problem-solving, and pill placebo. Wait-list (WL) was dened
as a control condition in which participants did not receive any
treatment for hypochondriasis/health anxiety symptoms for a
specied amount of time. Five of the 13 studies utilized treatment
as usual as the control condition, four of the 13 studies had waitlist as the control condition, three of the 13 studies had a
psychological placebo as the control condition, and one study
employed a pill placebo as the control condition.
Effect size calculation
Between-group effect sizes for each study were computed using
Hedges's g (Rosenthal, 1991). Studies with multiple outcomes were

Primary and secondary outcome measures were dened in accordance with

how the measures were described in the articles with the exception of two articles
(Barsky, Ahern, Bauer, Nolido, & Orav, 2013 and Warwick, 1996). Insufcient data
on the primary outcomes were provided for Barsky et al. and primary outcomes
were not clearly stated in the Warwick (1996) article, therefore we dened the
primary and secondary outcome consistent with the population of articles.
Primary mea- sures were exclusively measures of
hypochondriasis/health
anxiety.
Secondary
measures
generally
included
measures
of
hypochondriasis/health anxiety,
general anxiety,
and somatization, and
depression. Subsequent analyses were also con- ducted to examine the effects of
CBT on depression exclusively.

Table 1
Studies included in the meta-analysis.
Study

Conditions

Sorensen et al. (2011)

Seivewright et al. (2008)
Hedman et al. (2011)
Greeven et al. (2007)

CBT
CBT
CBT
CBT

WL
TAU
Psych PL
Pill PL

56
49
81
75

Barsky and Ahern (2004)

Visser and Bouman (2001)

CBT vs. TAU

Exposure vs. WL

187
42

vs.
vs.
vs.
vs.

Mean age

# of sessions

Primary outcome measure

Secondary outcome measure

Not provided
Not provided
39.05
40.32

16
4.3
12
7.3

HAMA, HAI
HAI
HAI
WI

43.21
37.7

6
12

WI
IAS-HA, IAS-Illness Behavior,
IAS-Body, SAS, SCL-90(Som)
IAS-HA, IAS-Illness Behavior,
IAS-Body, SAS, SCL-90(Som)

BDI, BAI, HAMD

HADS-D
IAS, WI, BAI, ASI, MADRS-S, QOLI
IAS-HA, IAS-Illness Behavior, SCL-90,
MADRS, BAS
HCQ, HAI, SSI, SAS
BDI, MOCI, SCL-Total

Visser and Bouman (2001)

CT vs. WL

40

36.7

12

Bourgault-Fagnou and
Hadjistavropoulos (2013)
Bourgault-Fagnou and
Hadjistavropoulos (2013)
Warwick (1996)
Sumathipala et al. (2008)
Speckens et al. (1995)

CBT vs. WL

39

68.94

WI

SHAI, SSI, STAI-S, STAI-T, GDS

ECBT vs. WL

36

68.53

WI

SHAI, SSI, STAI-S, STAI-T, GDS

37
35
37.1

16
3
16-Jan

HAI
GHQ-30
Intensity-Mean, Intensity-Max

Barsky et al. (2013)

McManus, Surawy, Muse,
Vazques-Montes, and
Williams (2012)
Buwalda, Bouman, and
Van Duijn (2006)

CBT vs. WL
CBT vs. TAU
CBT vs. TAU

36
150
79

BDI, MOCI, SCL-Total

CBT vs. Psych PL

89

51.84

HAI

BDI, BAI
BSI,# of visits,# of complaints
IAS-HA, IAS-Illness Behavior, WI,
HADS-A, HADS-D
SSI, MCQ, Beliefs, SCL-90, FSQ

CBT vs. TAU

74

42.6

HAC

BAI, BDI, FFMQ

CBT vs. Psych PL

48

41

GIAS

BDI

Note. CBT Cognitive Behavior Therapy, WL Waitlist, TAU Treatment As Usual, PL Placebo, NP Not Provided, HAMA: Hamilton Anxiety Rating Scale, HAI: Health
Anxiety Inventory, BDI: Beck Depression Inventory, BAI: Beck Anxiety Inventory, HAMD: Hamilton Depression Rating Scale, HADS-D: Hospital Anxiety and Depression ScaleDepression, IAS: Illness Attitude Scale, WI: Whitely inventory, ASI: Anxiety Sensitivity Inventory, MADRS: Montgomery Asberg Depression Rating Scale, QOLI: Quality of Life
Inventory, IAS-HA: Illness Attitude Scale-Health Anxiety, IAS-Illness Behavior: Illness Attitude Scale-Illness Behavior, SCL-SOM: Symptom Check List-Somatization scale, BAS:
Brief Anxiety Scale, HCQ: Hypochondriacal Cognitions Questionnaire, SSI: Somatic Symptoms Inventory, SAS: Somatosensory Amplication Scale, IAS-Body: Illness Attitude
Scale-Body, MOCI: Maudsley Obsessive Compulsive Inventory, SCL-Total: Symptom Check List-Total, SHAI: Short Health Anxiety Inventory, STAI-S: State Trait Anxiety Inventory State-Scale, STAI-T: State Trait Anxiety Inventory-Trait Scale, GDS: Geriatric Depression Scale, GHQ-30: General Health Questionnaire-30, BSI: Bradford Somatic
Inventory, HADS-A: Hospital Anxiety and Depression Scale-Anxiety, MCQ: Metacognitions Questionnaire, SCL-90: Symptom Checklist-90, FSQ: Functional Status Questionnaire, HAC: Health Anxiety Composite, FFMQ: Five Facet Mindfulness Questionnaire, GIAS: Groningen Illness Attitude Scale.

categorized as above (see Table 1) and then combined within each

domain. When the necessary data were available, Hedges's g was
calculated directly using the following formula: g X T - X C
=SP where X T is the mean of the treatment group, X C is the mean of
the comparison group, and SP is the pooled standard deviation. If
these data were not provided, Hedges's g was estimated using
conversion equations for signicance tests (e.g., t, F) (Rosenthal,
1991). All effect sizes were corrected for small sample sizes
according to Hedges and Olkin (1985). Therefore, a smaller
sample size reduces the estimated effect size helping control
for different sample sizes across studies. These controlled effect
sizes may then be inter- preted conservatively with Cohen's
convention of small (0.2), me- dium (0.5), and large (0.8) effects
(Cohen, 1988). Hedges's g also may be computed directly from
Cohen's d with the following for- mula: g d (1 - 3/4 (n1 n2)
- 9). When there were multiple outcomes per domain they were
combined according to Borenstein, Hedges, and Rothstein (2007).
The overall mean effect size for all of the studies combined was
computed
using
P
P the following formula:
g
wj gj = wj where wj is the weight for each study and gj is
the
effect size for each study. Effect sizes were calculated with random
effects models. The random effects analysis estimates the overall
effect size assuming the studies included are only a sample of the
entire population of studies and/or that the studies are
heterogeneous.

points (post and follow-up) on primary outcome measures were

included in this test of heterogeneity. The test was signicant,
Q(14) 58.03, p < .001, suggesting that random effects analyses
were most appropriate for this study. This signicant heterogeneity
also suggests it may be appropriate to employ moderator analyses
to identify potential sources of study variability.

Results

Question 2: does CBT outperform the control conditions on measures

of depression at post-treatment and follow-up?

Question 1: does CBT outperform the control conditions on primary

and secondary outcome measures at post-treatment and follow-up?
Fig. 3 depicts the relative advantage of CBT compared to control
conditions on primary outcomes at post-treatment in an analysis of
9 studies and 11 comparisons (Hedges's g 0.95 [SE 0.15, 95% CI:
0.66e1.22, p < .001]). The analysis involving outcomes at follow-up
included 7 studies and 7 comparisons and indicated that CBT outperforms control conditions on primary outcomes at follow-up,
although the effect size was signicantly smaller than that
observed at post-treatment (Hedges's g 0.34 [SE 0.14, 95% CI:
0.06e0.62, p .016]).
A post-treatment analysis of secondary outcome measures
where CBT outperformed control conditions included 9 studies and
11 comparisons (Hedges's g 0.56 [SE 0.11, 95% CI: 0.35e0.78,
p < .001]). A follow-up analysis of 7 studies and 7 comparisons
revealed that CBT also outperformed control conditions on secondary outcomes at follow-up (Hedges's g 0.20 [SE 0.10, 95% CI:
0.02e0.38, p .028]).1

Heterogeneity
To test the assumption that the effect sizes of this study were
from a homogeneous sample, a heterogeneity analysis was conducted. A total of 13 studies and 15 comparisons across pooled
time

An analysis of measures of depression at post-treatment

included 9 studies and 11 comparisons and showed that CBT

Fig. 3. Effect size estimates (Hedges' g) for the efcacy of CBT compared to control
conditions on primary symptom reduction.

outperformed control conditions (Hedges's g 0.64 [SE 0.12, 95%

CI: 0.41e0.86, p < .001]). The pooled follow-up analysis included a
total of 4 studies and 4 comparisons, where CBT also outperformed
control conditions on depression measures (Hedges's g 0.35
[SE 0.13, 95% CI: 0.01e0.61, p .008]).
Question 3: does higher pre-treatment hypochondriasis/health
anxiety severity and higher pre-treatment depression scores predict
lower effect sizes?
The analysis for pre-treatment severity included 4 studies and 4
comparisons with 218 participants and revealed that higher pretreatment severity was not signicantly associated with higher
effect size at post treatment (b 0.00, p .400). The analysis of
pre-treatment severity at follow up visits included 3 studies and 3
comparisons with 325 participants and revealed that higher pretreatment severity was signicantly associated with higher effect
size at follow-up (b 0.03, p < .050). The analysis for pretreatment
depression included 5 studies and 6 comparisons with 292 participants where higher pre-treatment severity was signicantly
associated with lower effect size at post treatment (b -0.09,
p < .050).2 Insufcient data was provided to analyze pre-treatment
depression severity at follow-up visits.
Question 4: does effect size vary as a function of control type, blind
assessment, number of sessions, percentage of females, mean age,
and comorbidity?
The following analyses were completed using fully random effects categorical moderator analyses. The 4 studies and 6 comparisons employing wait-list control conditions (Hedges's g 1.10
[SE 0.14, 95% CI: 0.83e1.37, p < .001]) showed larger effect sizes
than the 5 treatment as usual control conditions on primary outcomes across pooled time points (post and follow-up) (Hedges's
g 0.46 [SE 0.13, 95% CI: 0.20e0.71, p < .001]). Comparison between other control types and placebo-controlled studies was not

Studies that used the Health Anxiety Inventory were used to examine the
relationship between higher pre-treatment hypochondriasis/health anxiety and the
effect size for CBT. Studies that used the Beck Depression Inventory were used to
examine the relationship between higher pre-treatment depression and the effect
size for CBT. These were the most common measures used.

possible as few studies employed these control conditions (1 pill

placebo and 3 psychological placebos).
There was no signicant difference in effect size for posttreatment outcomes for 7 comparisons and 6 studies with blind
assessors (Hedges's g 1.11 [SE 0.19, 95% CI: 0.73e1.49, p ::; .
001]) and 4 comparisons and 3 studies which did not report having
blind assessors (Hedges's g 0.66 [SE 0.18, 95% CI: 0.31e1.01, p <
.001]). This comparison (4 comparisons and 4 studies) was not
signicant for follow-up outcomes either for the studies with
blind assessors (Hedges's g 0.72 [SE 0.19, 95% CI: -0.10 to
0.65, p .155]) and the 3 comparisons and 3 studies without blind
assessors (Hedges's g 0.46 [SE 0.25, 95% CI: -0.03 to 0.95, p
.068]). Treatment integrity was reported in 12 studies, but a
comparison was not possible because only 1 study failed to
report treatment integrity.
The following analyses were completed using unrestricted
maximum likelihood meta regressions. There was a signicant
relation between effect size and number of sessions (11 comparisons and 9 studies; b 0.07, p .038), with more sessions associated with larger effect sizes at post-treatment (Fig. 4). The relation
between effect size and number of sessions at follow-up was not
signicant (7 comparisons and 7 studies; b -0.01, p .858).
There was no signicant relation between effect size and female
percentage at post-treatment (10 comparisons and 8 studies;
b -0.00, p .822) nor at the follow-up analysis (7 comparisons
and 7 studies; b -0.01, p .373). Additionally, there was not a
signicant relation between effect size and mean age at posttreatment (9 comparisons and 7 studies; b 0.00, p .767) or at
follow up (6 comparisons and 6 studies; b -0.02, p .465). An
insufcient amount of studies provided information on comorbidity percentage; therefore, that analysis was not possible.
Publication bias: the le drawer problem
The omission of le drawer studies is a problem because it can
lead to biased meta-analytic estimates if their psychometric properties (i.e., reliability and validity) differ from psychometric properties of published studies (Rosenthal, 1991). Rosenthal's fail-safe N
was an ingenious response to this le drawer problem in the
integration of research (Orwin, 1983; Rosenthal, 1979). This
formula allowed for a direct assessment of the threat posed by
sampling bias in the literature search (Orwin, 1983). Rosenthal
suggested the following equation to compute a fail-safe
2
N:X K K Z - 2:706=2:706 where K is the number of studies in
the meta-analysis and Z is the mean Z obtained from the K studies
(Rosenthal, 1991). Rosenthal also suggested that ndings may be
considered robust if the required number of studies (X) to reduce
the overall effect size to a non-signicant level exceeded 5K 10,
which in this study would be 85 (Rosenthal, 1991). The analyses
conducted on the 15 comparisons amongst the 13 studies revealed
that it would require more than 166 current or future unpublished
studies with an effect size of 0 in order to bring the overall effect
size of the primary analyses within the non-signicant range. This
suggests that the ndings in this meta-analysis are robust as the
fail-safe-N (166) is much larger than the convention for robust
cutoff or 5K 10 (85).
Discussion
Traditionally, hypochondriasis was considered resistant to psychotherapy, possibly because of the shortcomings of psychodynamic and psychoanalytic conceptualizations, which historically
dominated the treatment of this condition. These models were
removed from behavior, cognition, and biology, and the treatments
derived from them produced little enduring benet. Within the last
two decades, however, a model of hypochondriasis as health

Regression of Number of Sessions on Hedges's g

2.00
1.80

Hedges's g

1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
3.13

4.53

5.94

7.34

8.75

10.15

11.55

12.96

14.36

15.77

17.17

Number of Sessions
Fig. 4. The relationship between CBT for hypochondriasis/health anxiety effect size and number of sessions.

anxiety has been advanced that draws from the cognitive (i.e.,
dysfunctional beliefs, body vigilance, anxiety sensitivity, intolerance of uncertainty) and behavioral (i.e., avoidance, safety-seeking)
processes implicated in the development of other anxiety disorders
(Olatunji et al., 2009). This model, depicted in Fig. 1, has informed
the application of CBT to the treatment of hypochondriasis by
helping patients recognize and modify faulty beliefs about illness
and eliminate behavioral responses that prevent the self-correction
of faulty beliefs (Taylor & Asmundson, 2004). The present investigation employed meta-analysis to examine the efcacy of this
approach to the treatment of hypochondriasis/health anxiety. The
ndings showed that CBT outperformed control conditions on
primary symptom outcome measures at post-treatment showing a
large effect size. This nding is consistent with prior meta-analyses
demonstrating that CBT is effective in reducing symptoms of hypochondriasis (Taylor et al., 2005; Thomson & Page, 2007). However, the present study expands this prior research by including a
number of RCTs that have been published since these previous
meta-analyses, and thus adds to the evidence base of CBT for hypochondriasis/health anxiety.
The present investigation also found that CBT outperformed
control conditions on primary hypochondriasis/health anxiety
symptom outcome measures at follow-up showing a small effect
size. Although the sustained efcacy (the ability to produce lasting
symptomatic changes) of CBT-based approaches has been questioned (e.g., Eddy, Dutra, Bradley, & Westen, 2004), these metaanalytic ndings suggest that treatment gains attributed to CBT
(relative to controls) are observed after treatment is completed.
While these ndings are encouraging, additional research is needed
to adequately determine the extent to which CBT produces longer
lasting symptom changes for patients with hypochondriasis/health
anxiety. This will require future studies to include substantially
longer follow-up intervals so more denitive inferences can be
made regarding the durability of CBT for hypochondriasis/health
anxiety. It is important to note that the overall effect size for
CBT was signicantly larger at
post-treatment (g .95)
compared to follow-up (g 0.34). This suggests that the
therapeutic effects of CBT for hypochondriasis/health anxiety
may diminish after the acute treatment phase. Future research
aimed at identifying stra- tegies that may be employed during
and after CBT (i.e., booster sessions) in order to better sustain
treatment gains for patients with hypochondriasis/health anxiety
may prove valuable. Such approaches may be vital in reducing the
health care costs associated with excessive doctor visits and
requests for unnecessary medical tests that is often observed
among those with hypochondriasis/ health anxiety.

Although the present ndings suggest that the therapeutic effects of CBT for hypochondriasis/health anxiety may diminish after
the acute treatment phase, they also suggest that the acute therapeutic effects of CBT for patients with hypochondriasis/health
anxiety may generalize to other symptoms. More specically, CBT
outperformed control conditions on secondary outcome measures
at post-treatment. Although secondary outcome measures did
include assessment of symptoms of hypochondriasis/health anxiety, they also included assessment of psychiatric symptoms broadly
dened, as well as specic symptoms of general anxiety, somatization, and depression. Prior research has shown that hypochondriasis/health anxiety does demonstrate signicant associations
with general psychopathology (Gropalis, Bleichhardt, Witthoft, &
Hiller, 2012; Weck, Neng, Richtberg, & Stangier, 2012). However,
the present study suggests that CBT that is specically targeted
towards the reduction of hypochondriasis/health anxiety may also
signicantly reduce other symptoms of distress.
The present study also examined if CBT treatments outperformed control conditions on reducing depressive symptoms
specically at post-treatment. The experience of depression is quite
common among those with hypochondriasis/health anxiety
(Noyes et al., 1994), so much so that some have speculated that
hypo- chondriasis is a masked form of depression (Lesse,
1980). The present ndings revealed that CBT for patients with
hypochon- driasis/health anxiety also resulted in signicant
reductions in symptoms of depression. These ndings are
consistent with prior work showing that CBT that specically
targets anxiety-related disorders also
outperform control
conditions in reducing symp- toms of depression (Hofmann &
Smits, 2008; Olatunji, Davis, Powers, & Smits, 2013). Research
has shown that symptoms of hypochondriasis/health anxiety may
emerge before those of depression (Noyes et al., 1994; Simon
Gureje, & Fullerton, 2001), suggesting that depressive symptoms
may represent a response to the distress and functional
impairment associated with hypo- chondriasis/health
anxiety
(Abramowitz, 2004). Given this functional relationship, directly
targeting symptoms of hypochon- driasis/health anxiety may be
expected to also lead to reductions in symptoms of depression.
The present investigation also examined the extent to which
initial pre-treatment hypochondriasis/health anxiety severity and
pre-treatment depression scores predicted CBT effect sizes. Prior
longitudinal research suggests that severity of hypochondriasis
may be of prognostic signicance (Buwalda & Bouman, 2008;
Hartman et al., 2009). Although it is intuitive to predict that
higher pre-treatment hypochondriasis/health anxiety severity will
predict lower effect sizes, the opposite was true in the present

study. That is, higher pre-treatment hypochondriasis/health anxiety was found to be associated with an increase in CBT effect size at
follow-up visits. This nding does replicate those of a recent study
that found that health anxiety at baseline was positively associated
with symptom improvement after internet-based CBT for severe
health anxiety (Hedman, Andersson, et al., 2013; Hedman,
Lindefors, et al., 2013). The present ndings also compliment
those of Nakao, Shinozaki, Ahern, and Barsky (2011) who found
that more baseline anxiety predicted larger improvements
associated with a CBT intervention for hypochondriasis. Contrary
to
conven- tional wisdom, the nding that psychological
treatment might be more efcacious for high-severity than for
low-severity patients has
also
been observed in
the
treatment of other disorders (Driessen, Cuijpers, Hollon, &
Dekker, 2010; Smits et al., 2013; Smits, Minhajudin, Thase, &
Jarrett, 2012).
The question remains as to why CBT is more effective (relative
to controls) for patients with more severe hypochondriasis/health
anxiety. Is this merely regression toward the mean? According to
Driessen et al. (2010), evidence of moderation is most likely to be
found when efcacious treatments are compared with stringent
controls and the sample contains both more and less severe
patients. In such cases, nonspecic treatments may be sufcient for
low-severity patients, but high-severity patients will require a
treatment that has specic effects beyond the simple provision of
treatment as usual in order to fully benet. Should CBT be proven
to be more efcacious relative to control conditions for patients
with more severe symptoms of hypochondriasis/health anxiety
than for those with less severe symptoms, this does not mean
that treat- ment guidelines should be revised to recommend CBT
as a mono- treatment for patients with severe symptoms. Indeed,
the moder- ated nding of symptom severity should be
interpreted with caution given that the effect was small (b
0.03) and the clinical signicance of such an effect is not
sufciently apparent. The question of the incremental efcacy of
CBT for this group of patients requires more head-to-head
comparisons between CBT and other bona de psychological
treatments before more denitive in- ferences can be made.
Unlike pre-treatment hypochondriasis/health anxiety severity,
pre-treatment depression symptom severity was signicantly
associated with a decrease in the advantage of CBT over control
conditions at post-treatment. This nding is consistent with recent
research showing that depressive symptoms at baseline were
negatively related to improvement after internet-based CBT for
severe health anxiety (Hedman, Andersson, et al., 2013; Hedman,
Lindefors, et al., 2013). As previously noted, depressive symptoms
often co-occur with hypochondriasis/health anxiety (Noyes et al.,
1994). The presence of severe depression may then impede
response to CBT for hypochondriasis/health anxiety through multiple mechanisms. For example, depressive symptoms may
decrease motivation and compliance with difcult exposure assignments that are often employed as homework in CBT, thereby
preventing meaningful reductions in symptoms of hypochondriasis/health anxiety. It is important to note that the nding showing
that pre-treatment depression symptom severity may predict
worse outcomes is not unique to hypochondriasis/health anxiety.
Indeed, prior research has shown that pretreatment levels of
depression also predict worse outcomes for patients with OCD
(Abramowitz, Franklin, Street, Kozak, & Foa, 2000; Stewart, Yen,
Egan, & Jenike, 2006), a disorder that is related to hypochondriasis/health anxiety (Abramowitz, 2005). Based on these ndings,
there have been efforts to develop and implement a treatment
program specically for depressed OCD patients (Abramowitz,
2004). Although the present ndings did indicate that CBT for
hy- pochondriasis/health anxiety also
reduces depressive
symptoms, the ndings also suggest that patients with
hypochondriasis/health

anxiety that present with severe depression may benet from

supplementary interventions that specically target depressive
symptoms.
Examination of the effect size as a function of treatment as usual
and waitlist controls revealed that CBT showed smaller effect sizes
when compared to treatment as usual control conditions than
when compared to waitlist controls. One interpretation of these
ndings is that non-specic factors may inuence treatment
outcome in CBT for hypochondriasis/health anxiety to some degree.
In fact, examination of the effect size for CBT compared to waitlist
controls and that of CBT compared to treatment as usual control
conditions in the present investigation would suggest that the
differences are far from negligible in magnitude. Examination of
the extent to which other sample-specic and study-specic
characteristics predicted outcome also revealed some important
ndings. For example, the CBT effect size for hypochondriasis/
health anxiety did not vary as a function of blind assessment at post
treatment or follow-up.
More CBT sessions were also found to be associated with larger
effect sizes at post-treatment. Although a relationship between
number of CBT treatment sessions and treatment efcacy has not
been consistently observed in other disorders (Abramowitz, 1996;
Rosa-Alca'zar, Sa'nchez-Meca, Go' mez-Conesa, & MarnMartnez,
2008), this nding suggests that there may be added benet to
more CBT sessions for patients with hypochondriasis/health anxiety. This nding does appear to be in line with recent work showing
that treatment adherence (i.e. the number of completed treatment
modules) in internet-based CBT for severe health anxiety signicantly predicted outcome (Hedman, Andersson, et al., 2013;
Hedman, Lindefors, et al., 2013). In addition to the number of sessions, future research is needed to examine if the length of CBT
treatment sessions is associated with outcomes as there may be
some advantages to a more intensive approach to the treatment of
hypochondriasis/health anxiety.
Hypochondriasis is a chronic, disabling, prevalent and costly
disorder that has been generally viewed as refractory to treatment
(Barsky & Ahern, 2004). Although the specic and problem focused
nature of CBT has been the basis of concerns that it may not
generalize to real-life patients who frequently present with cooccurring conditions (Westen, Novotny, & Thompson, 2004), the
present meta-analytic investigation suggests that CBT is effective in
reducing both disorder-specic and nonspecic (e.g., depression)
symptoms for those with hypochondriasis/health anxiety. The
present investigation also identied some moderators of outcome
that may have prognostic and prescriptive implications. A prognostic variable is one that predicts outcome irrespective of the
treatment, whereas a prescriptive variable (often referred to as a
moderator) predicts a different pattern of outcomes between two
or more treatment modalities (Hollon & Beck, 1986). Important
prescriptive indicators were observed in the present study that
may guide future research. For example, the prescriptive
indicators showed that CBT is more effective than control
conditions for
pa- tients with more severe symptoms of
hypochondriasis/health anxiety but less effective for patients with
more severe symptoms of depression. However, most RCTs that
have been evaluated for the treatment of hypochondriasis/health
anxiety have been CBT focused. One
study did
nd that
patients with hypochondriasis who received CBT did signicantly
better on all measures relative to the wait-list control group, and
on a specic measure of health anxiety compared with shortterm psychodynamic psychotherapy (Sorensen, Birket-Smith,
Watta, Buemann, & Salkovskis, 2011). Furthermore, the short-term
psychodynamic psychotherapy group did not signicantly differ
from the waiting-list group on any outcome measures. More
examination of the efcacy of other psychological treatments
for hypochondriasis/health anxiety will

be valuable in identifying more substantive prescriptive variables

that may inform individually tailored treatment.
Although the present meta-analysis represents a signicant
improvement on prior work that was limited by a small number of
studies, some of which were uncontrolled (Taylor et al., 2005;
Thomson & Page, 2007), the use of only RCTs is perhaps another
limitation of the study as it resulted in the availability of a
relatively small number of studies for
analysis. Thus,
examination of some moderators (i.e., effects of blind assessors)
in
the
meta-analysis were potentially underpowered.
Accordingly, some signicant moderators had relatively small
beta weights and these ndings should be interpreted with some
caution. Another issue is that an insufcient amount of studies
provided information on variables (i.e., comorbidity percentage)
that would allow for a more comprehensive examination of
potential moderators of outcome. This observation highlights the
importance of more thorough assessment and reporting of data in
RCTs. Although the use of RCTs does allow for greater condence
in
the
present
ndings, more RCTs of
CBT for
hypochondriasis/health will also allow for the identication of
more reliable moderators and mediators of treat- ment outcome.
A recent study examining the mediating role
of putative
mechanisms in internet-delivered CBT for severe health anxiety
found that reduced perceived risk of disease, less attention to
bodily symptoms, and reduced intolerance of uncertainty
signicantly mediated improvement in health anxiety (Hedman,
Andersson, et al., 2013; Hedman, Lindefors, et al., 2013). The
mediating effects of these processes are consistent with the
cognitive-behavioral model represented in Fig. 1. Future research
delineating additional mediating mechanisms is crucial because
such mechanisms provide important insight into the processes
that should be targeted during treatment.
It is important to note that the very nature of RCTs may make
it difcult to detect meaningful moderators and mediators,
especially if stricter eligibility criteria and overly rigid delivery of
CBT pro- tocols eliminate the very processes that inuence how
patients respond to treatment. The use of more effectiveness trials
in com- munity settings may complement RCTs in efforts to
identify mod- erators and mediators of treatment outcome.
Indeed, a conuence of efcacy and effectiveness trials may
ultimately reveal what treatment, by whom, is most effective for
patents with hypochon- driasis/health anxiety, and under which
set of circumstances. Although the newly published DSM-5 has
replaced the diagnosis of hypochondriasis with illness anxiety
disorder (APA, 2013), many of the underlying processes remain
the same. Patients with illness anxiety disorder experience
heightened bodily sensations, are intensely anxious about the
possibility of an undiagnosed illness, are not easily reassured, and
engage in excessive safety behaviors to minimize anxiety about
health concerns. Given the overlap in symptoms, CBT should
prove to be equally as effective for illness anxiety disorder as it is
for hypochondriasis.
References3
Abramowitz, J. S. (1996). Variants of exposure and response prevention in the
treatment of obsessiveecompulsive disorder: a meta-analysis. Behavior Therapy,
27, 583e60 0.
Abramowitz, J. S. (20 04). Treatment of obsessive compulsive disorder in patients
who have comorbid depression. Journal of Clinical Psychology, 60, 133e1141.
Abramowitz, J. S. (20 05). Hypochondriasis: conceptualization, treatment, and
relationship with obsessiveecompulsive disorder. Annals of Clinical Psychiatry,
17, 211e217.
Abramowitz, J. S., Deacon, B. J., & Valentiner, D. P. (20 07). The Short Health Anxiety
Inventory: psychometric properties and construct validity in a non-clinical
sample. Cognitive Therapy and Research, 31, 871e883.

*Indicates the studies used in the meta-analysis.

Abramowitz, J. S., Franklin, M. E., Street, G., Kozak, M., & Foa, E. B. (20 0 0). Effects of
comorbid depression on response to treatment for obsessiveecompulsive disorder. Behavior Therapy, 31, 517e528.
Abramowitz, J. S., & Moore, E. L. (20 07). An experimental analysis of hypochondriasis. Behaviour Resea rch and Therapy, 45, 413e424.
Abramowitz, J. S., Olatunji, B. O., & Deacon, B. J. (20 07). Health anxiety, hypochondriasis, and the anxiety disorders. Behavior Therapy, 38, 86e94.
Abramowitz, J. S., Schwartz, S. A., & Whiteside, S. P. (20 02). A contemporary conceptual model of hypochondriasis. Mayo Clinic Proceedings, 77, 1323e1330.
American Psychiatric Association. (20 0 0). Diagnostic and statistical manual of mental
disorders (4th ed. text revision). Washington, D.C: Author.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental
disorders (5th ed., rev.). Washingt on, DC: Author.
Barsky, A. (1996). Hypochondriasis: medical management and psychiatric treatment. Psychosomatics, 37, 48e56.
* Barsky, A. J., & Ahern, D. K. (20 04). Cognitive behavior therapy for hypochondriasis. A randomized trial. Journal of the American Medical Association, 291,
1464e1470.
* Barsky, A. J., Ahern, D., Bauer, A., Nolido, N., & Orav, J. (2013). A randomized trial of
treatments for high-utilizing somatizing patients. Journal of General Internal
Medicine, 28, 1396e1404.
Barsky, A., Barnett, M. C., & Clear y, P. D. (1995). Hypochondriasis and panic disorder:
boundary and overlap. Archives of Gene ral Psychiatry, 51, 918e925.
Barsky, A., Fama, J., Bailey, E., & Ahern, D. K. (1998). A prospective 4- to 5-year
study of DSM-III-R hypochondriasis. Archives of Gene ral Psychiatry, 55,
737e74 4.
Beck, A. T., & Emer y, G. (20 05). Anxiety disorders and phobias: A cogniti ve perspective
(15th annual ed.). New York: Basic Books.
Bernstein, I. C., Callahan, W. A., & Jaranson, J. (1975). Lobotomy in private practice:
long-term
follow-up. Archives of
General Psychiatry, 32,
1041e1047.
Beutler, L. E. (1991). Have all won and must all have prizes? Revisiting Luborsky
et al'.s ver dict. Journal of Consulting and Clinical Psychology, 59, 226e232.
Borenstein, M., Hedges, L. V., & Rothstein, H. (20 07). Meta-analysis d Fixed effects
versus random effects. Biostat Inc. (Tech. Rep).
Bornstein, M., & Rothstein, H. (1999). Comp rehensive meta-analysis: A computer
program for research synthesis. New Jerse y: Biosta.
* Bourgault-Fagnou, M. D., & Hadjistavropoulos, H. D. (2013). A randomized trial of
two forms of cognitive behaviour therapy for an older adult population with
subclinical health anxiety. Cognitive Behaviour Therapy, 42, 31e4 4.
Buwalda, F. M., & Bouman, T. K. (20 08). Predicting the effect of psychoeducational
group treatment for hypochondriasis. Clinical Psychology and Psychotherapy, 15,
396e403.
* Buwalda, F. M., Bouman, T. K., & Van Duijn, M. A. J. (20 06). Psychoeducation for
hypochondriasis: a comparison of a cognitiveebehavioural approach and a
problemsolving approach. Behaviour Resea rch and Therapy, 45, 887e899.
Clark, D. M., Salkovskis, P. M., Hackmann, A., Wells, A., Fennell, M., Ludgate, J., et al.
(1998). Two psychological treatments for hypochondriasis. A randomised
controlled trial. British Journal of Psychiatry, 173, 218e225.
Cohen, J. (1988). Statistical power analysis for the behavio ral sciences (2nd ed.).
Hillsdale, NJ: Erlbaum.
Creed, F., & Barsky, A. (20 04). A systematic review of the epidemiology of somatisation disorder and hypochondriasis. Journal of Psychosomatic Research, 56,
391e408.
Driessen, E., Cuijpers, P., Hollon, S. D., & Dekker, J. J. (2010). Does pretreatment
severity moderate the efcacy of psychological treatment of adult outpatient
depression? A meta-analysis. Journal of Consulting and Clinical Psychology, 78,
668e680.
Eddy, K., Dutra, L., Bradley, R., & Westen, D. (20 04). A multidimensional metaanalysis of pharmacotherapy for obsessiveecompulsive disorder. Clinical Psychology Review, 24, 1011e1030.
Fallon, B., Javitch, J., Hollander, E., & Liebowitz, M. (1992). Hypochondriasis and
obsessive compulsive disorder: overlaps in diagnosis and treatment. Journal of
Clinical Psychiatry, 52, 457e460.
Greeven, A., van Balkom, A. J., van der Leeden, R., Merkelbach, J. W., van den
Heuvel, O. A., & Spinhoven, P. (20 09). Cognitive behavioral therapy versus
paroxetine in the treatment of hypochondriasis: an 18-month naturalistic
follow-up. Journal of Behavior Therapy and Experimental Psychiatry, 40, 487e496.
* Greeven, A., van Balkom, A. J., Visser, S., Merkelbach, J. W., van Rood, Y. R., van
Dyck, R., et al. (20 07). Cognitive behavior therapy and paroxetine in the treatment of hypochondriasis: a randomized controlled trial. American Journal of
Psychiatry, 164, 91e99.
Gropalis, M., Bleichhardt, G., Witthoft, M., & Hiller, W. (2012). Hypochondriasis,
somatoform disorders, and the anxiety disorders: sociodemographic variables,
general psychopathology, and naturalistic treatment effects. Journal of Nervous
and Mental Disease, 200, 406e412.
Gureje, O., Ustun, T. B., & Simon, G. E. (1997). The syndrome of hypochondriasis: a
cross-national study in primary care. Psychological Medicine, 27, 10 01e1010.
Hartman, T. C., Borghuis, M., Lucassen, P., van de Laa r, F., Speckens, A., & van Weel,
C.
(20 09). Medically unexplained symptoms, somatisation disorder and hypochondriasis: course and prognosis. A systematic review. Journal of Psychosomatic Resea rch, 66, 363e377.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. London: Academic Press.
Hedman, E., Andersson, E., Andersson, G., Lindefors, N., Lekander, M., Rck, C., et al.
(2013). Mediators in internet-based cognitive behavior therapy for severe
health anxiety. PLoS ONE, 8(10), e77752.

* Hedman , E., Anderss on, E., Andersson, G. , Ljo' tsson, B. , Rc k , C., Asmun dso n,
G., et al. (2011). Internet -ba se d c ognit ive- be havioural the rapy for severe
health anxi et y:
randomi ze d
c ont ro ll ed
t ri a l .
Britis h
Jour nal
of
Psychiatry, 198 ,
23 0e236 .
Hedman, E., Lindefors, N., Andersson, G., Andersson, E., Lekander, M., Rck, C., et al.
(2013). Predictors of outcome in Internet-based cognitive behavior therapy for
severe health anxiety. Behaviour Research and Therapy, 51, 711e717.
Hofmann, S. G., & Smits, J. A. (20 08). Cognitive-behavioral therapy for adult anxiety
disorders: a meta-analysis of randomized placebo-controlled trials. Journal of
Clinical Psychiatry, 69, 621e632.
Hollon, S. D., & Beck, A. T. (1986). Cogniti ve and cognitive -behavior therapies. In
S. L. Gartield, & A. E. Bergin (Eds.), Handbook of psychotherapy and behavior
change (Vol. III); (pp. 4 43e482). New York: Wiley.
House, A. (1989). Hypochondriasis and related disorders: assessment and management of patients referred for a psychiatric opinion. General Hospital Psychiatry, 11, 156e165.
Katon, W. J., & Walker, E. A. (1998). Medically unexplained sympt oms in primary
care. Journal of Clinical Psychiatry, 59, 15e21.
Ladee, G. A. (1966). Hypochondriacal syndromes. Amsterdam: Elsevier.
Lesse, S. (1980). Masked depression-the ubiquitous but unappreciated syndrome.
Psychiatric Journal of the University of Ottawa, 5, 268e273.
Marcus, D. K., Gurley, J. R., Marchi, M. M., & Bauer, C. (20 07). Cognitive and
perceptual variables in hypochondriasis and health anxiety: a systematic review. Clinical Psychology Review, 27, 127e139.
* McManus, F., Surawy, C., Muse, K., Vazques-Montes, M., & Williams, M. G. (2012).
A randomized clinical trial of mindfulness-based cognitive therapy versus unrestricted services for health anxiety (hypochondriasis). Journal of Consulting
and Clinical Psychology, 80, 817e828.
Nakao, M., Shinozaki, Y., Ahern, D. K., & Barsky, A. J. (2011). Anxiety as a predictor
of improvements in somatic symptoms and health anxiety associated with
cog- nitiveebehavioral intervention in hypochondriasis. Psychotherapy and
Psycho- somatics, 80, 151e158.
Noyes, R., Kathol, R. G., Fisher, M. M., Phillips, B. M., Suelzer, M. T., & Woodman, C. L.
(1994). Psychiatric comorbidity among patients with hypochondriasis. General
Hospital Psychiatry, 16, 78e87.
Olatunji, B., Davis, M., Powers, M., & Smits, J. (2013). Cognitive-behavioral therapy
for obsessiveecompulsive disorder: a meta-analysis of treatment outcome and
moderators. Journal of Psychiatric Research, 47, 33e41.
Olatunji, B. O., Deacon, B. J., & Abramowitz, J. S. (20 09). Is hypochondriasis an
anxiety disorder? British Journal of Psychiatry, 194, 481e482.
Orwin, R. G. (1983). A fail safe N for effect size in meta-analysis. Journal for
Educational Statistics, 8, 157e159.
Paul, G. L. (1967). Strategy of outcome research in psychotherapy. Journal of
Consulting
Psycholog y,
31,
109e118.
Rosa-Alca'zar, A. I., Sa'nchez-Meca, J., Go' mez-Conesa, A., & Marn-Martnez, F.
(20 08).
Psychological treatment of obsessiveecompulsive disorder: a meta-analysis.
Clinical Psyc hology Review, 28, 1310e1325.
Rosenthal, R. (1979). The le drawer problem and tolerance for null results.
Psychological Bulletin, 86, 638e641.
Rosenthal, R. (1991). Meta-analytic procedures for social research (rev. ed.). Thousand
Oaks, CA, US: Sage Publications, Inc.
* Seivewright, H., Green, J., Salkovskis, P., Barrett, B., Nur, U., & Tyrer, P. (2008).
Cognitive-behaviou ral therapy for health anxiety in a genitourinary medicine
clinic: randomised controlled trial. British Journal of Psychiatry, 193, 332e337.
Simon, G. E., Gureje, O., & Fullerton, C. (20 01). Course of hypochondriasis in an
international primary care study. Gene ral Hospital Psychiatry, 23, 51e55.

Smits, J. A. J., Hofmann, S. G., Roseneld, D., DeBoer, L. B., Costa, P., Simon, N. M.,
et al. (2013). d-cycloserine augmentation of cognitive behavioral group therapy
of social anxiety disorder: prognostic and prescriptive variables. Journal of
Consulting & Clinical Psychology, 81, 110 0e1112.
Smits, J. A. J., Minhajuddin, A., Thase, M. E., & Jarrett, R. J. (2012). Outcomes of acute
phase cognitive therapy in outpatients with anxious versus nonanxious
depression. Psychotherapy & Psychosomatics, 81, 153e160.
* Sorensen, P., Birket-Smith, M., Wattar, U., Buemann, I., & Salkovskis, P. (2011).
A randomized clinical trial of cognitive behavioural therapy versus short-term
psychodynamic psyc hotherapy versus no intervention for patients with hypochondriasis. Psychological Medicine, 41, 431e441.
* Speckens, A. E. M., van Hemert, A. M., Spinhoven, P., Hawton, k. E., Bolk, J. H., &
RooiJmans, H. (1995). Cognitive behavioural therapy for medically unexplained
physical symptoms: a randomized controlled trial. BMJ, 311, 1328e1332.
Speckens, A. E. M., Spinhoven, P., van Hemert, A. M., Bolk, J., & Hawton, K.
(1997). Cognitive behavioural therapy for unexplained physical symptoms;
process and prognostic factors. Behavioural and Cognitive Psychotherapy, 25,
291e294.
Stewart, S. E., Yen, C. H., Egan, S. D., & Jenike, M. (20 06). Outcome predictors for
severe obsessiveecompulsi ve patients in intensive residential treatment. Journal of Psychiatric Research, 40, 511e519.
* Sumathipala, A., Siribaddana, S., Abeysingha, M. R. N., De Silva, P., Dewey, M.,
Prince, M., et al. (20 08). A randomised controlled trial of cognitive behavioural
therapy versus structured care for medically unexplained symp toms: a randomized controlled trial. British Journal of Psychiatry, 193, 51e59.
Taylor, S., & Asmundson, G. (20 04). Treating health anxiety: A cogniti ve-behavioral
approach. New York: Guilford Press.
Taylor, S., Asmundson, G., & Coons, M. (20 05). Current directions in the treatment of
hypochondriasis. Journal of Cogniti ve Psychotherapy: An International Quarterly,
19, 285e304.
Thomson, A. B., & Page, L. A. (20 07). Psychotherapies for hypochondriasis. Cochrane
Database of Systematic Reviews, 4, 1e43.
* Visse r, S., & Bouman, T. K. (20 01). The treatment of hypochondriasis: exposure
plus response prevention vs. cognitive thera py. Behaviour Research and Therapy,
39, 423e442.
Walker, J., Vincent, N., Furere, P., Cox, B., & Kjernisted, K. (1999). Treatment
pref- erence in hypochondriasis. Journal of Behavior Therapy and Experimental
Psy- chiatry, 30, 251e258.
* Warwi ck, H. M. C. (1996). A controlled trial of cognitive-behavioural treatment of
hypochondriasis. British Journal of Psychiatry, 169, 189e195.
Warwick, H. M. C., Clark, D. M., Cobb, A. M., & Salkovskis, P. M. A. (1996). Controlled
trial of cognitive-behavioural treatment of hypochondriasis. British Journal of
Psychiatry, 169, 189e195.
Warwick, H., & Salkovskis, P. (1990). Hypochondriasis. Behaviour Research and
Therapy,
28,
105e117.
Warwick, H., & Salkovskis, P. (20 01). Cognitive-behavioral treatment of hypochondriasis. In V. Starcevic, & D. R. Lipsitt (Eds.), Hypochondriasis: Modern perspectives on an ancient malady (pp. 314e328). New York: Oxford University
Press.
Weck, F., Neng, J. M., Richtberg, S., & Stangier, U. (2012). Dysfunctional beliefs about
symptoms and illness in patients with hypochondriasis. Psychosomatics, 53,
148e154.
Westen, D., Novotny, C. M., & Thompson-Brenner, H. (20 04). The empirical status of
empirically supported psychotherapies: assumptions, ndings, and reporting in
controlled clinical trials. Psychological Bulletin, 130, 631e663.