Annette J. Bouwhuis-Van Keulen, Jurrijn Koelen, Liesbeth Eurelings-
Bontekoe, Christien Hoekstra-Oomen & Gerrit Glas
To cite this article: Annette J. Bouwhuis-Van Keulen, Jurrijn Koelen, Liesbeth Eurelings-
Bontekoe, Christien Hoekstra-Oomen & Gerrit Glas (24 Aug 2023): The evaluation of religious
and spirituality-based therapy compared to standard treatment in mental health care: A
multi-level meta-analysis of randomized controlled trials, Psychotherapy Research, DOI:
10.1080/10503307.2023.2241626
To link to this article: https://doi.org/10.1080/10503307.2023.2241626https://doi.org/10.1080/10503307.2023.2241626
ANNETTE J. BOUWHUIS-VAN KEULEN1∗, JURRIJN KOELEN2∗,
LIESBETH EURELINGS-BONTEKOE3, CHRISTIEN HOEKSTRA-OOMEN4, &
GERRIT GLAS1
1Faculty of Humanities, Department of Philosophy, VUmc Amsterdam, Amsterdam, The Netherlands; 2Faculty of
Behavioural, Management and Social Sciences, Department of Psychology, Health and Technology, University of Twente,
Enschede, The Netherlands; 3Faculty Social Sciences, Department of Clinical Psychology, University of Leiden, Leiden, The
Netherlands & 4Psychology Practice Edelweis, Delft, The Netherlands
(Received 1 December 2022; revised 20 July 2023; accepted 21 July 2023)
ABSTRACT
Objective Psychotherapies are increasingly incorporating spiritual and religious systems of belief and practice, which aligns
with recent developments toward person-centered treatments. The main objective of this meta-analysis was to compare the
efficacy of a religion and spiritually-based (R/S) therapy to non-R/S treatments.
Method A multi-level meta-analysis was conducted to compare randomized controlled studies of the efficacy between R/Sbased
and regular treatments in mental health care setting. Inclusion criteria were diagnosis, psychotherapeutic treatment,
and explicitly religion/spirituality therapy. Outcome was assessed for symptoms and for functioning separately, and
combined. We also examined several moderators, such as type of comparison, outcome domain, and diagnosis.
Results Overall effect sizes obtained from 23 studies and 27 comparison groups indicated that a R/S treatment is moderately
more efficacious compared to regular treatments at posttreatment (g = .52, p < .01) and at follow-up (g = .72, p < .01) (only
available for symptoms). Results were similar for symptoms (g = .44, p < .01) and functioning (g = .62, p < .01).
Conclusion In patients with a strong religious and spiritual affiliation, treatments with a focus on religious and spiritual
issues are more efficacious than non-R/S-based therapy. Limitations as well as future directions are discussed.
Keywords: meta-analysis; randomized controlled trials; religious and spirituality-based therapy; mental health care;
psychotherapy
Clinical or methodological significance of this article: This meta-analysis expands knowledge about R/S-based
psychotherapy and clarifies inconsistent findings from previous meta-analyses. Person-centered therapy with attention to
religious and spiritual language, questions, and subjects are efficacious, especially for depressed patients.
Introduction
The Pew Research Centre (2015) estimates that in
2050 86.8% of the 9.3 billion world population will have a religious/spiritual affiliation. Each religious
tradition has its own worldview and perspective on
human beings, and teaches its followers how to live life in accordance with spiritual values. Religious
knowledge, experiences and activities influence
mental health and psychopathology (Stulp et al.,
2019), personality (Schaap-Jonker et al., 2017),
and coping strategies (Pargament, 2007). Therefore
religious and spiritual affiliations in the domain of
mental health care should be considered important,
which is consistent with person-centered approaches
(see e.g., Glas, 2019, 2021).
Since the end of the last century, attention for the
existential dimension of being human and suffering,
and thus patients’ religious or spiritual world views,
has notably increased in clinical treatment (Glas,
2021; van Os et al., 2019). The American Psychological
Association (APA) has developed guidelines
for dealing with cultural, religious, and spiritual
issues in treatment (APA, 2017). Moreover, scientific
research has revealed the importance of spirituality
to the clinical process, such as the protective effect
of spirituality on mental health (Stulp et al., 2019), as
well as the influence of religious coping strategies on
psychopathology (Ano & Vasconcelles, 2005). As
part of these developments, integrative religious
and spiritual treatment approaches have been developed
that endorse an active search for the sacred, and
how this reflects on a person’s coping with mental
illness (Pargament, 2007; Richards & Bergin,
2005). These religious/spiritual treatments use vocabulary
not used in regular psychotherapy, and they
tend to strongly emphasize the mobilization of
hope, acceptance, and forgiveness, amongst other
concepts. In this more holistic context, with attention
to existential, transcendent, and meaning-making
questions, mental disorders are treated and patients
are met (Glas, 2021).
In this contribution, we focus on psychotherapeutic
treatments that seek to improve the psychological
well-being of patients, by using religious or spiritual
sources related to a transcendent dimension and a
search for the sacred. Following Worthington et al.
(2011), we use the term religious and spiritual
based (R/S) therapy to refer to these treatments,
although other terms are also used, such as spiritually
integrated psychotherapy (Pargament, 2007), faith
adapted therapy (Anderson et al., 2015), body–
mind-spirit therapy (Hsiao et al., 2011), religion
accommodative psychotherapy (Paukert et al.,
2011). Mindfulness-based treatments are not
included because these interventions do not focus
on finding meaning in the transcendent dimension,
but rather at promoting non-judging attention
toward mental states in the current moment, independently
of religious or spiritual content. Our
main aim is to address the question of whether R/S
treatment is more efficacious than non-R/S treatments
for patients suffering from mental disorder.
Previous Meta-Analyses on the Effectiveness
of R/S Treatments
In recent years, several meta-analyses have been published
on the effectiveness of R/S treatments compared
to non-R/S treatments or no treatment
(Anderson et al., 2015; Captari et al., 2018; McCullough,
1999; Smith et al., 2007; Worthington et al.,
2011). Results are somewhat inconsistent, probably
due to the choices of inclusion criteria used, variations
in statistical methods, treatment criteria, and
control groups. Two meta-analyses focused only on
randomized controlled trials (Anderson et al.,
2015; McCullough, 1999), two studies included
also quasi-experimental studies (Captari et al.,
2018; Worthington et al., 2011) and in the metaanalysis
of Smith et al. (2007) single-group pre- to
posttest designs were also allowed. Due to the diversity
of research methods, a considerable variation in
the quality of the included studies was indicated
(Anderson et al., 2015; Paukert et al., 2011), which
may affect the results. Second, the meta-analyses
apply different inclusion criteria in terms of diagnosis
and treatment. McCullough (1999) includes all
psychological diagnosis, while Anderson et al.
(2015) is limited to depression and anxiety disorders,
and the remaining three meta-analyses aggregate
both psychological, psychosocial, and physiological
diseases. In addition, four meta-analyses use a
broad definition of psychotherapy, allowing for
much variation in treatments (for example one
session therapy, internet therapy, lifestyle training,
and prayer groups). Finally, the inconsistent results
are possibly due to the variation in treatment and
control groups. Table I shows the results of the published
meta-analysis split by, essentially, three comparing
groups. The first comparing group, the
blended group, was used in the meta-analysis of
Captari et al. (2018), McCullough (1999) and
Worthington et al. (2011) and compared a regular
therapy with religion-accommodative regular
therapy, using equal amounts of treatment as
inclusion criteria. Mainstream treatment (i.e. cognitive
behavioral therapy, (CBT)) was adapted from a religious mindset, appropriate within the psychological
theory (i.e. Religious Cognitive Behavioral
therapy (RCBT)). They found an effect size of,
respectively g = .13, d = .18 and d = .13, indicating
that R/S therapy is slightly more effective than
regular therapy. Secondly, Captari et al. (2018) compared
a R/S treatment with a standard therapy, with
both treatments having a completely separate theoretical
background (i.e. Spiritual Integrated treatment
compared to Rational Emotive Therapy). They
found for this comparison with a so-called alternate
treatment group an effect size of Hedges’ g = .33, in
favor of the R/S therapy. Finally, Worthington et al.
(2011), Smith et al. (2007), and Anderson et al.
(2015) merged two or more control groups, including,
as we call it, an additional design. In an
additional treatment group a standard treatment
was compared to a standard treatment plus
additional R/S sessions. This treatment group
differs from the blended group, because of the difference
in treatment duration, and from the alternate
group due to corresponding theoretical orientation.
Medium effect sizes (.26 < d < .59) were found
these combined treatment groups, in favor of the R/
S therapy (Anderson et al., 2015; Smith et al.,
2007; Worthington et al., 2011).
Moderators of Treatment Effects
Because of these inconsistent findings, it is important
to consider potential moderators. First, as mentioned,
the composition of the R/S-based treatment
and the control group plays a significant role
(Captari et al., 2018; Worthington et al., 2011). In
addition to the distinction between treatment and
control group, other treatment aspects can influence
the results, such as the treatment setting (inpatient
vs. outpatient), diagnosis treated, and religious or
spiritual framework. Most meta-analyses use broad
definitions for psychodiagnostics and psychotherapy,
therefore we choose stricter inclusion criteria (a
mental disorder) and opted to include diagnosis as
a potential moderator. Secondly, several quality
aspects seem to influence the ambivalent results,
including methodology and researcher allegiance
(Anderson et al., 2015; Paukert et al., 2011).
Aims of the Current Meta-Analysis
The primary aim of the current meta-analysis was to
compare R/S-based treatments to non-R/S psychotherapies.
Strict inclusion criteria were applied
to research methods, definitions of religiosity/spirituality,
and the definition of psychotherapy. This
meta-analysis focused solely on RCT’s that have been carried out in mental health care settings
among participants with an official (ICD/DSM)
diagnosis, to ensure a more homogeneous study
population. The secondary aim was to systematically
examine several moderator variables, including study
quality, to investigate whether this study can replicate
and extend previous findings. Moreover, in contrast
to most previous meta-analyses, we
systematically examined the type of comparison
that was made in single studies and addressed this
as a potential moderator of effect size magnitude.
The following hypotheses are tested: (i) a R/S treatment
is more efficacious compared to a non-R/S
treatment, in line with results from previous metaanalyses
(Anderson et al., 2015; Captari et al.,
2018; McCullough, 1999; Smith et al., 2007;
Worthington et al., 2011); (ii) the type of comparison
made and study quality moderate the difference in
effect between a R/S and the control group. As
noted, previous meta-analyses have found different
results depending on the type of comparison (Anderson
et al., 2015; Captari et al., 2018; McCullough,
1999; Worthington et al., 2011), therefore we
expect the effect size to be moderated by the type
of comparison. We expect that the effect size of a
“blended” spiritual treatment will be lower from
other types of comparisons, such as those comparing
a regular therapy with the same therapy with separate
R/S therapy added. To our knowledge, this is the first
meta-analysis that addresses these types of comparisons
in more detail. Regarding study quality, in
general, this variable affects the effect sizes of randomized
controlled trials (Thoma et al., 2012) and
for research on R/S-based therapy in particular,
quality aspects seem to influence the ambiguous outcomes
(Anderson et al., 2015; Paukert et al., 2011).
Hence, study quality is expected to moderate the
results, but we have no clear predictions about this given inconsistent findings.
Method
Literature Search and Identification of
Studies
A multiple phase search was conducted in September
2014 and updated in October 2019 and January
2022 to retrieve as many RCTs as possible regarding
the efficacy of religious and spiritual psychotherapy
in mental health care. We used two search strategies
to find eligible studies. First, we conducted searches
in the databases Web of Science, PsycINFO, and
PubMed. Web of Science was searched in September
2014, in October 2019, and in January 2022 using
the terms: (Religio∗ OR Spirit∗ OR Existent∗ OR
Christian OR Muslim OR Buddhis∗ OR Taois∗ OR Jewis∗) AND (Randomi∗ OR Psychotherapy OR
Therapy OR Treatment OR Counsel∗ OR Intervention
∗ OR Outcome OR Patient∗) in the title. In
PsycINFO and PubMed, using the same terms,
only clinical trials (RCTs), reviews, and meta-analyses
were searched. Second, the included studies
in previously published meta-analyses and reviews
(Anderson et al., 2015; Captari et al., 2018;
Hodge, 2006; Hook et al., 2010; Lim et al., 2014;
McCullough, 1999; Paukert et al., 2011; Smith
et al., 2007; Viftrup et al., 2013; Vos et al., 2015;
Worthington et al., 2011) were assessed. We contacted
experts in the field of psychotherapy and religion/
spirituality by e-mail for any ongoing studies,
however, this did not yield any new studies. We did
not search for unpublished manuscripts. Finally, we
searched for unpublished ongoing studies in two
trial registers (www.clinicaltrials.gov and www.
controlled-trials.com), which yielded no results. As
the analysis plan of the meta-analysis and the literature
search took place before 2014, a pre-registration
of the study is missing
Inclusion Criteria
In this study, we set stringent inclusion criteria, to
reduce variability among studies as much as possible.
Included studies had to fulfill the following criteria,
described according to the PICO (population, intervention,
comparator, outcome) framework.
Population. Patients (i) had to be 18 years or
older, (ii) had to meet the official criteria for “Axis
I” mental disorders according to DSM or ICD
manuals, and (iii) were treated in mental health
care settings.
Intervention. Psychotherapeutic and counseling
treatments, meeting the definition of psychotherapy
according to Norcross (1990) were included, and
these treatments were explicitly religious or spiritual,
excluding mindfulness, meditation, yoga, due to its
lack of focus on a transcendent dimension. In
addition, only randomized controlled trials (excluding
cross sectional and quasi experimental studies)
were allowed. There were no restriction on language
and publication date.
Comparator. “Active” control groups were
allowed for comparison, excluding wait list controls.
In addition, due to variation in comparison treatment
groups, we made the following four types of comparisons
(Captari et al., 2018; Worthington et al., 2011):
(i) additional comparison: a regular treatment was
compared with the identical treatment except for
the addition of extra religious or spiritual sessions
(regular therapy compared to regular therapy + R/S
therapy); (ii) blended comparison: a religious and
non-religious treatments had the same theoretical
orientation (e.g., CBT) and duration, but in only
one group attention was paid to religious and spiritual
beliefs, cognition, values and actions (for
example: regular rational emotive therapy (RET)
compared to R/S modified version of RET); (iii)
alternative comparison: a R/S treatment was compared
with an alternative treatment with a different
theoretical background (for instance: CBT compared
to spiritual integrated therapy); (iv) medication
comparison: patients who were treated with
medication were compared with patients in a R/S
treatment (medication compared to R/S therapy).
Outcome. Validated questionnaires and quantifiable
outcome data of the main disorder under treatment
were used, with at least posttreatment
measurements. Two outcome measures were distinguished,
depending on the measurements available
from individual studies: (i) psychological symptoms
(split into self-reported and observer-based); and
(ii) functioning/well-being in the broad sense (cognitive,
behavioral, spiritual). Examples of questionnaires
used: Social Adjustment Scale, Automatic
Thoughts Questionnaire, Spiritual Well-Being
Scale, Ways of Religious Coping Scale. We decided
to use follow-up data at 6 months after treatment termination
or nearest to 6 months, because follow-up
data at 6 months were most available.
The screening process consisted of two phases.
Firstly, the first author screened all titles and
abstracts for eligibility. After that, the first author
assessed the full text articles for meeting all inclusion
criteria. When the first author was in doubt, the
studies were discussed in group meetings (GG, LE,
JK, and AB) where relevant parts of the text were
read and assessed until consensus was reached.
Due to limited resources, it was not possible for the
screening and data extraction to take place by two
authors.
Data Extraction
The first author extracted data from each included
study, specifically: basic statistics needed for effect
size calculation, treatment and patient variables,
therapist information, and outcome measures. Treatment
variables were modality (group, individual or
both), setting (inpatient or outpatient), treatment
dosage (intensity of treatment), calculated by
length of treatment in weeks multiplied by number
of sessions multiplied by length of each session (in
hours), and religion/spirituality (Christian, Muslim,
Buddhism, spiritual, Taoist). Patient information
consisted of diagnosis, population (e.g., patients, students,
church members), percentage of women, age,
and dropout (average treatment dropout of the R/S
and regular treatment). Therapist information pertained
to training (psychiatrist, psychologist/counselor,
doctoral students, mental health nurse),
experience, and the number of therapists in the trial.
Quality Assessment
Two independent reviewers (AB and CH) assessed
the quality of the studies using the Psychotherapy
Quality Rating Scale (PQRS), a scale for assessing
randomized clinical trials (Kocsis et al., 2010). This
scale includes 24 items covering six quality domains
and an omnibus item (question 25) to award a final
score to the study. All items are scored on a threepoint
scale (0–2), except item 25 (seven-point
scale: 1–7). Two different quality scores were
derived: a summative total of the individual items
1–24 (range 1–48) and an omnibus score (range 1–
7). The judges were trained by an expert (JK). The
one-way random intraclass correlation coefficient
(ICC) (single measures) was .882 and the average
measures ICC .938 (excellent). In the analyses, the
means of the two scores were used in case of a deviation
of one point, and when the deviation was two
points reviewers discussed the item and reached consensus
about the score.
Data Analysis
Calculation of effect sizes. The analyses were
performed using R-Studio software version
2023.03.0 (Build386). Two functions of R were
used: the rma.mv function of the metafor package
(Viechtbauer, 2015) to fit multi-level meta-analytic
models and the metagen function of the meta
package (Balduzzi et al., 2019) for plots. Because of
dependence of effect sizes of included studies, resulting
in unit-of-analysis problems and double-counting
problems, we used a multi-level meta-analytic model
(Harrer et al., 2021). A multi-level model allows all
the relevant effect sizes reported in the individual
studies to be included. Assink and Wibbelink
(2016) describe in a step-by-step tutorial how to
perform a three-level meta-analytic model with the
rma.mv function in R. In addition, using the
metagen package, separate effect sizes were calculated
for symptoms and functioning posttreatment
and follow-up, and prediction intervals were
assessed. A prediction interval quantifies the uncertainty
associated with making individual predictions and indicates what effects are to be expected for
future patients (IntHout et al., 2016). It provides
an estimated range within which the true value of
an individual observation is likely to fall, taking into
account both the inherent variability in the data
and the uncertainty in the prediction model. We
also generated forest plots and funnel plots with
this package.
Between group effect sizes (Hedges’ g) were calculated
by comparing means, standard deviations and
sample sizes of the R/S-based treatments with the
regular treatments at post-treatment and follow-up.
In studies that compared two or more therapies,
the sample size of the shared group was split, to
avoid the problem of double counting (Harrer
et al., 2021). To calculate the pooled effect sizes
per domain, the n was divided by the number of
questionnaires used in the study. Effect sizes of
< .30 are considered small, .50 medium, and >.80
large (Cohen, 1977).
A random effects model was used to compute
weighted mean effect sizes, because samples, treatment
interventions, population, and methodologies
were expected to vary across studies. The random
effects model results in more conservative results
and broader 95%-confidence intervals (95% CI)
than the fixed effects model. To estimate the parameters
in the multi-level model, the Restricted
Maximum Likelihood method (REML) was used
(Viechtbauer, 2005), and a Knapp-Hartung adjustment
(Knapp & Hartung, 2003) of the confidence
interval was applied. The following effect sizes were
calculated: overall (symptoms + functioning) effect
sizes posttreatment (in rma.mv), psychological
symptoms posttreatment and follow-up (in
metagen), and posttreatment effect sizes for the functioning
outcomes (in metagen).
Heterogeneity. In order to test for heterogeneity,
one-sided log-likelihood-ratio-tests were performed
in rma.mv to test variance (Assink & Wibbelink,
2016). Three sources of variance are distinguished
in the multi-level model: Level 1: sample variance
of the effect sizes; Level 2: variance between effects
sizes within one study (e.g., due to subgroups
within a study or different outcome domains);
Level 3: variance between studies. Significant
results on the second and third level indicate heterogeneous
distribution of the effect sizes. In that case,
differences in effect sizes could be explained by
study characteristics, which can be further investigated
by moderator analyses.
Moderator analysis. Instructions from Assink
and Wibbelink (2016) regarding moderator analysis
were used. Dummy variables were created for the
categorical variables (assessment perspective,
domains, type of comparison, religion, diagnosis,
population, modality, and setting) and the dimensional
variables (study quality, mean age, percentage
gender, percentage dropout, treatment doses) were
centered around the mean. The results of the moderator
analyses include: (i) an omnibus test of moderators,
based on the F distribution, to assess
whether a variable moderates the difference in
effect size between a R/S and non-R/S treatment,
i.e. to test significant differences between the levels
of a moderator, and (ii) a model result showing
whether a level (reference category) within a moderator
individually deviate from zero, and whether the
other levels differ from the reference category,
based on the t distribution.
Publication bias. Publication bias denotes the
problem that non-significant studies could be unpublished,
potentially resulting in biased meta-analytic
findings. Several methods were used to test for publication
bias. Funnel plots were visually inspected,
and Egger’s regression and the Duval and Tweedie’s
trim and fill procedure were calculated. Publication
bias was tested with the metagen package for psychological
symptoms and functioning posttreatment.
Results
Study Selection
Figure 1 shows the flow chart of the selection procedure:
9688 articles were screened and, of these,
9453 were discarded based on title and 248 based
on abstract. A total of 235 full-text articles were
assessed for eligibility, of which 206 were excluded,
17 were included and 12 studies were discussed in
group meetings. A description of the discussion and
decision-making process can be found in Supplement
4.
Descriptive Characteristics of Studies
Treatments of the included studies. Twentythree
studies were included in the meta-analysis
with a total of 1499 clients (R/S treatment: n = 768;
non-R/S treatment: n = 731). In total, 27 comparisons
could be made from which k = 83 posttreatment
effect sizes were extracted and k = 28 follow-up effect
sizes. Razali et al. (1998) and Zhang et al. (2002)
evaluated two studies in their articles and in the
studies of Ebrahimi et al. (2013) and Richards
et al. (2006) a spiritual group was compared with 2
alternative treatments. Six treatments had a Muslim tradition (n = 245), eight Christian (n = 159), one
Taoist (n = 95), one Buddhist (n = 58) and seven
general spiritual tradition (n = 223).
Detailed information about the treatment offered
can be found in Supplement 1. In summary, 16
studies pertained to outpatient treatments, four
studies examined inpatients, and three studies
lacked information about the setting. Of these 16
studies involving outpatients, five treatments were
offered in group therapy (average of 8.2 sessions),
eight studies pertained to individual sessions
(average of 10.88 sessions), two studies combined
individual and group sessions and one study lacked
information. The amount of studies in each type of
comparison was: (i) additional: 15 studies (n =
581); (ii) blended: 6 studies (n = 85); (iii) alternative:
3 comparisons (n = 39); and (iv) medication: 3
studies (n = 97). Patients in the blended and alternative
comparison group received equal dosage of
therapy sessions.
Patient characteristics in the included
studies. A variety of disorders were treated in the
studies: depression (s = 10); dysthymic disorder (s
= 3); generalized anxiety disorder (s = 7); eating disorders
(s = 3); substance abuse (s = 2), and 2
studies treated multiple disorders. The average age
of participants was 34.5 years (5 studies lacked
data) and 57% were women (7 studies missing
data). Patients were involved in 74.1% of the comparison
groups, the remainder of participants were
drawn from student samples and the general population
(25.9%).
Therapist characteristics in the included
studies. In four studies the therapist was a psychiatrist,
in four a doctoral student provided the
treatment, in two studies a psychologist was
involved, in three studies a mental health nurse
and in two a counselor was involved. In only five
articles the therapists were described as experienced,
and four studies reported that the therapists
were trained and supervised in the provided treatment.
In five studies, the therapist treats both
treatment groups. The number of therapists
involved was usually one or two, however, three
studies involved more than three therapists. Two
studies provided an evaluation of the treatment
and evaluation of the therapeutic relationship was
absent in all studies. Seven studies did not
contain any information about the therapists
involved. A minority of studies (17%) paid attention
to allegiance of therapists to the type of treatment.
In these cases, the therapists treated both
treatment groups and were not involved in the development of the R/S treatment. However, in all
these studies the authors or supervisors developed
the R/S therapy.
Methodological quality of the included
studies. The mean methodological quality across
the 23 studies according to the PQRS was 19. This
indicates an overall “rather poor” study quality and
falls below the score of 24 reflecting “adequate
quality.” Six of the included studies had a score
above 23.
Effect Sizes
Overall R/S versus regular treatment. The
first aim of this meta-analysis was to compare the efficacy
of a R/S treatment to religious/spiritual patients
with the efficacy of a regular treatment post-treatment
(Table II). The overall effect size was Hedges’ g = .52
(SE = .14, t(82) = 3.68, p < .001), 95% CI = 0.24–
0.80, PI = (−0.76–1.81) at posttreatment, indicating
that a R/S treatment is moderately more efficacious
compared to a regular therapy. Results of the log-likelihood-
ratio-tests, assessing heterogeneity, revealed significant variability between studies (level 3),
55.09% of the total variance could be attributed to
differences between studies (p < .001). Within-study
variance was mainly found at level 1 (sampling variance)
(44.91%).
Psychological symptoms R/S versus regular
treatment. Significant effects on psychological
symptoms were found at posttreatment (g = .44,
SE = .13, t(48) = 3.36, p < .01, 95% CI = 0.17–
0.70), of which 7 outliers were detected in six
studies (Barron, 2007; Chida et al., 2016; Razali
et al., 2002; Shanke et al., 2017; Tonkin, 2005;
Zhang et al., 2002). After removing these outliers,
the significant result remained (g = .43, t(41) =
6.20, p < .001, 95% CI = 0.29–0.57). Figure 2
shows the funnel plot of the symptom results, and
the appendix contains the forest plot (Supplement
2) including the prediction interval (PI = −1.14–
2.01). The superiority of a R/S treatment was maintained
over longer time frames, considering the
follow-up results (g = .72, SE = .24, t(27) = 2.98, p
< .01, 95% CI = 0.22–1.22).
Functioning R/S versus regular treatment.
Posttreatment effect sizes for functioning, with one
outlier (Miller et al., 2008), also demonstrated the
efficacy of a R/S-based therapy over a non-R/S treatment
(g = .62, SE = .09, t(35) = 6.82, p < .01, 95% CI = 0.44–0.81) (see Figure 3 the funnel plot and
in the appendix the forest plot (S3)), including the
prediction intervals, which does not contain zero
(PI = 0.03–1.21). Follow-up results were not available
for this outcome domain.
Moderators. Tables III and IV show the results of
the univariate moderator analyses comparing R/S
treatment with regular treatment. The results of the
omnibus tests of moderators showed no significant
differences for any of the potential moderator variables
(p values of the F distributions > .05), meaning
that the overall effect size is not moderated by any of
the categorical and continuous moderator variables.
The model results of the moderator analyses did
show differences within variables, that is, for some
levels within a variable significant effect sizes (βo and
to) were found between a R/S-based therapy and
regular therapy. For example, for assessment perspective,
the level observer-based symptoms and functioning
were both significantly different from zero, but this
was not the case for self-reported symptoms. This
implies that the reported differences between the R/
S and non-R/S treatment groups applies only to observer-
based symptoms and functioning. Both domains,
symptoms, and functioning, deviated significantly
from zero, meaning that there was no distinction
between symptoms and functioning concerning the
superiority of R/S-based therapy. Only the additional
type of comparison group showed a significant deviation,
indicating that the reported differences only
apply to the additional comparison group. Moreover,
for the moderator diagnosis, effect sizes of anxiety and
eating disorders were not significant while studies that
treated depressed patients the R/S treatment differed
significantly from a non-R/S therapy. This implies
that our results only apply to depressed patients. All
religions deviated significantly from zero, meaning
that our results concerning the superiority of R/Sbased
therapy apply to all religions tested. Finally, R/S-based therapy outperformed regular therapy in
case of patients (compared to general population),
outpatients (compared to inpatients), and treatments
combining individual and group therapy (compared
to individual therapy).
Publication Bias
Publication bias was tested for psychological symptoms
and functioning at posttreatment. First, visual
inspection of the funnel plots did not indicate
potential bias, although asymmetry is visible on the
right side of the two plots. Secondly, Egger’s
Regression Test (Egger et al., 1997) showed that
the intercept of the variable symptoms (βo = −.39,
p = .59) and functioning (βo = −.83, p = .13) did not
deviate significantly from zero, implying that the
funnel plots are indeed symmetric. Duval and Tweedie’s
trim and fill method showed that symmetry in
the funnels plots was restored by imputing 17
symptom effect sizes from 11 studies, and 9 effect
sizes (6 studies) for functioning. All effect sizes
were imputed on the right side of the plots,
meaning an overrepresentation of below average
findings. In sum, there appears to be some publication
bias, however, from studies with below
average results.
Discussion
The first purpose of this meta-analysis was to
compare a R/S-based therapy to non-R/S treatments.
The hypothesis that religious or spiritual patients
benefit more from a R/S therapy compared to non-
R/S therapy was confirmed in this meta-analysis.
Moderate effect sizes were found overall (g = .52)
and for symptom reduction (g = .44), comparable
to previous meta-analytic findings that reported
effects between d = .13 – d = .59 for reducing complaints
(Anderson et al., 2015; Captari et al., 2018;
McCullough, 1999; Smith et al., 2007; Worthington
et al., 2011). R/S treatments not only reduce complaints
more, but also enhance general functioning
(g = .62) (to compare: d = .41 in Worthington et al.
[2011] and g = .43 in Captari et al. [2018]). Interestingly,
as indicated by the prediction intervals in both
domains, the estimate of functioning (PI = .03–1.21)
was more robust than that of symptoms (PI = -.76–
1.81). In the case of functioning, the interval does
not contain zero, which implies that R/S interventions
will most likely be efficacious in this domain
for future patients. This fits with the more personalized
approach which is inherent in R/S treatment,
where the focus does not lie primarily on symptoms.
In other words, this meta-analysis affirms the beneficial
effects of person-centered treatments in
which religious and spiritual experiences, language,
and worldviews of patients are recognized and explicitly
addressed. We suspect that an explanation can
be found in the person-centered approach that
focuses on the person as a whole and sees the experience
and alleviation of psychological distress as a personal
journey, or narrative, with an inherent
existential dimension (Glas, 2021). Our meta-analysis
corroborates the view that openness and willingness
to address the spiritual, religious, and existential aspects of the patient’s suffering are
associated with feelings of being recognized,
accepted, and understood as a person. Although
the underlying mechanisms are not well understood,
this open and encouraging attitude invites using a
spiritually more inclusive vocabulary which might
positively affect the therapeutic relationship (van Os
et al., 2019). It also emphasizes the importance to
foster awareness in clinicians of their R/S identity in
training and supervision, as this is an area where psychotherapists
feel they lack competence (Magaldi &
Trub, 2018). In fact, dealing with self-disclosure of
(parts of) their own R/S identity may enhance openness
and thus strengthen the therapeutic relationship.
Yet, our study does not answer the question
how the various and complex ways in which worldviews
of both patient and clinician interact and how
this interaction may influence treatment results.
This topic deserves further study.
Implication of the Moderator Findings
The moderator analyses yielded interesting exploratory
results, but these results should be evaluated
taking into account the (limited) statistical power
per moderator. First, that the type of comparison
would be a significant moderator of effect, as
suggested by previous meta-analyses, could not be
replicated in the current meta-analysis. The
omnibus moderator test yielded no significant
result. However, of all type of comparisons, only
the effect size of the additional group differed significantly.
As was shown in this and three previously
published meta-analyses, a “blended” treatment
does not appear to show a significant effect compared
to its non-blended counterparts (Captari et al., 2018;
McCullough, 1999; Worthington et al., 2011; see
however Anderson et al., 2015). However, there
seems to be more differentiation in the types of comparisons
that were merged by earlier meta-analyses
(Table I), including the distinction between an
additional and alternative design. This may explain
some of the inconsistent results from previous
meta-analyses. Importantly, however, it should be
noted that statistical power is a major problem in
this comparison, as the alternative groups consists
of only 69 patients. Further research is needed to
gain more clarity on whether the method of integrating
religion and spirituality into treatments affects the
efficacy of psychotherapy. Secondly, the overall
effect size comparing a R/S-based therapy to a
general therapy seems especially applicable to treatments
of depressed patients. This exploratory analysis
suggests that depressive patients in
particular benefit from treatments in which religious and spiritual subjects are part of a mental health
treatment. Aspects of R/S are closely intertwined
with risk of depressive symptoms, coping with, and
the course and recovery from depression (Braam &
Koenig, 2019). Positive aspects of R/S, including
prayer, church attendance and social support, as
well as discussing spiritual struggles, reflecting on
life values and existential meaning, can promote
recovery from depression (Hittner & Swickert,
2010). Secondly, the overall effect size of selfreported
symptoms is not significant, while observer-
based symptoms and functioning are. Several
explanations are given for the discrepancy between
observer and self-reported symptoms, for example,
personality factors, personality characteristics, and
psychiatric history (Schat et al., 2017). Finally, we
hypothesized that study quality would moderate the
difference in effect between R/S and non-R/S treatment.
This was not confirmed in this meta-analysis;
an overall study quality measure did not explain the
difference in efficacy. In conclusion, the results of
the moderator analyses point toward some directions
for further research.
Limitations
The results of this meta-analysis have to be interpreted
within the context of several theoretical and
procedural limitations. First, the mean methodological
quality of the included studies was “rather poor.”
This may influence the results (Thoma et al., 2012),
although no association was found between the
overall study quality and treatment results. Highquality
randomized controlled trials are needed to
increase the reliability of the efficacy of R/S-based
treatments. Second, the moderators are not always
unambiguous and therefore reliable. For example,
studies lack information about the moderators, allegiance
bias is present, the absence of various religions
(Jewish, Buddhism), the approach of integrating religion
in psychotherapy, and some moderators appear
to overlap, such as type of comparison and religion.
In addition, potential mediating variables that
explain the effectiveness of psychotherapy in
general, and a R/S treatment in particular, are
missing, such as client factors (motivation and involvement
in therapy), patients’ preference, and quality
of the therapeutic relationship (Cooper, 2008). We
recommend researchers to include these variables
in further research to explain the efficacy of R/S treatments.
Thirdly, although this meta-analysis tried to
create a homogeneous research group, the studies
differ on various variables, which raises the question
whether comparing these studies with each other
does full justice to their individual aspects. The studies differ, for example, in the severity of the complaints,
psychotherapy orientations, content of the R/
S treatment, control group and location. Not all of
these factors could be included as moderators.
Further, the meta-analysis did not include a
measure of the importance of spirituality to patients
and their religious motivation, while studies show,
for example, that the onset and course of psychopathology
differ in patients who are intrinsically or
extrinsically religious motivated (Park, 2021). It is
likely that patients sought help at institutions with
religious affiliation with good reason, and this may
have introduced an important bias. Specifically, it
may have resulted in underestimated effects for the
control groups due to patients’ unfulfilled needs to
talk about religion. This topic requires further attention
as religious motivation may affect the outcome
of a R/S treatment. Fifth, this meta-analysis is
based on a traditional mental health care system,
with categorical diagnosis, an official DSM diagnosis
and evidence-based treatments. However, the
approach within mental health care is increasingly
broader, with more emphasis on dimensionalizing
of mental illnesses and transdiagnostic factors
(Krueger & Eaton, 2015). This broader approach
may yield a more comprehensive picture, also with
respect to severity of illness.
Regarding the procedural limitations, all studies
were searched and included by one author, which
may have compromised the reliability of the search.
Second, in the absence of data on the studies, the
authors were not contacted to complete this information
or were not given access to all potential eligible
research reports. Further, to assess study
quality, we took the sum scale score, and this
overall rating may miss key elements, such as allegiance
(Jüni et al., 1999). Finally, pre-registration
of the meta-analysis was lacked.
Despite some limitations mentioned, this metaanalysis
yields new insights, compared to previous
meta-analyses, by providing a more homogeneous
base of primary studies, (with strict criteria regarding
a DSM diagnosis, psychotherapy, and spiritual
topics of a R/S-based treatment). Second, we evaluated
treatment effects in several domains. Third, we
included a number of moderators, for example by
looking in some detail into varieties of control groups.
Conclusions
This meta-analysis supports the hypothesis that a R/
S-based treatment is more efficacious than regular
treatment in mental health care settings for patients
with religious affiliations. This type of person-centered
treatment focused on the religious, spiritual
and existential dimensions of human beings is efficacious
in reducing psychological symptoms and enhancing
general functioning, in particular for depressed
patients. We, therefore, advocate that clinicians and
mental health care providers should discuss patients’
worldviews, expressed in religious and spiritual subjects,
in psychotherapy. More research is needed
into the role of the type of comparison and the way
in which religious and spiritual topics should be integrated
in psychotherapy. We encourage future
research on R/S psychotherapy to adhere to developments
toward person-centered and holistic therapeutic
approaches. Finally, we recommend that
future studies focus on mechanisms of change in R/
S psychotherapeutic treatment, as this is an essential
step toward improving psychotherapy further.
Funding
This research did not receive any specific grant from
funding agencies in the public, commercial, or notfor-
profit sectors.
Disclosure Statement
No potential conflict of interest was reported by the
author(s).
Supplemental data
Supplemental data for this article can be accessed
online at https://doi.org/10.1080/10503307.2023.
2241626.
Contributors
A.J. Bouwhuis-van Keulen: conceptualization,
writing – original draft, literature search, data curation,
quality assessment, methodology, formal analysis,
visualization. J. Koelen: conceptualization,
writing – review and editing, supervision, methodology,
formal analysis. E.H.M. Eurelings: conceptualization,
writing – review, and editing. C.
Hoekstra-Oomen: quality assessment, writing –
review and editing. G. Glas: conceptualization,
writing – review, and editing.
References
References marked ∗ are studies included in the
meta-analysis.
American Psychological Association. (2017). Multicultural guidelines:
An ecological approach to context, identity, and intersectionality.
Retrieved March 4, 2021, from http://www.apa.org/about/ policy/multicultural-guideline.pdf.
Anderson, N., Heywood-Everett, S., Siddiqi, N., Wright, J.,
Meredith, J., & McMillan, D. (2015). Faith-adapted psychological
therapies for depression and anxiety: Systematic
review and meta-analysis. Journal of Affective Disorders, 176,
183–196. https://doi.org/10.1016/j.jad.2015.01.019
Ano, G. G., & Vasconcelles, E. B. (2005). Religious coping and
psychological adjustment to stress: A meta-analysis. Journal of
Clinical Psychology, 61(4), 461–480. https://doi.org/10.1002/ jclp.20049
Assink, M., & Wibbelink, C. J. M. (2016). Fitting three-level
meta-analytic models in R: A step-by-step tutorial. The
Quantitative Methods for Psychology, 12(3), 154–174. https:// doi.org/10.20982/tqmp.12.3.p154
∗Azhar, M. Z., & Varma, S. L. (1995a). Religious psychotherapy
as management of bereavement. Acta Psychiatrica
Scandinavica, 91(4), 233–235. https://doi.org/10.1111/j.1600- 0447.1995.tb09774.x
∗Azhar, M. Z., & Varma, S. L. (1995b). Religious psychotherapy
in depressive patients. Psychotherapy Psychosomatics, 63(3–4),
165–168. https://doi.org/10.1159/000288954
∗Azhar, M. Z., Varma, S. L., & Dharap, A. S. (1994). Religious
psychotherapy in anxiety disorder patients. Acta Psychiatrica
Scandinavica, 90(1), 1–3. https://doi.org/10.1111/j.1600- 0447.1994.tb01545.x
Balduzzi, S., Rücker, G., & Schwarzer, G. (2019). How to
perform a meta-analysis with R: A practical tutorial. Evidence
Based Mental Health, 22(4), 153–160. https://doi.org/10.1136/ ebmental-2019-300117
∗Barron, L. W. (2007). Effect of religious coping skills training with
group cognitive behavioral therapy for treatment of depression
[Unpublished doctoral dissertation], Northcentral University,
Prescott, Arizona.
Braam, A. W., & Koenig, H. G. (2019). Religion, spirituality and
depression in prospective studies: A systematic review. Journal
of Affective Disorders, 257, 428–438. https://doi.org/10.1016/j. jad.2019.06.063
Captari, L. E., Hook, J. N., Hoyt, W., Davis, D. E.,
McElroy-Heltzel, S. E., & WorthingtonJr.E. L. (2018).
Integrating clients’ religion and spirituality within psychotherapy:
A comprehensive meta-analysis. Journal of Clinical Psychology, 74
(11), 1938–1951. https://doi.org/10.1002/jclp.22681
∗Chida, Y., Schrempft, S., & Steptoe, A. (2016). A novel religious/
spiritual group psychotherapy reduces depressive symptoms in
a randomized clinical trial. Journal of Religion and Health, 55(5),
1495–1506.https://doi.org/10.1007/s10943-015-0113-7
Cohen, J. (1977). Statistical power analysis for the behavioural
sciences. Academic Press.
Cooper, M. (2008). Essential research findings in counselling and psychotherapy:
The facts are friendly. British Association for
Counselling and Psychotherapy.
∗Ebrahimi, A., Neshatdoost, H. T., Mousavi, S. G., Asadollahi,
G. A., & Nasiri, H. (2013). Controlled randomized clinical
trial of spirituality integrated psychotherapy, cognitive-behavioral
therapy and medication intervention on depressive symptoms
and dysfunctional attitudes in patients with dysthymic
disorder. Advances Biomedical Research, 2(1), 1–7. https://doi. org/10.4103/2277-9175.114201
Egger, M., Davey-Smith, G., Schneider, M., & Minder,
C. (1997). Bias in meta-analysis detected by a simple, graphical
test. British Medical Journal, 315, 629–634. https://doi.org/10. 1136/bmj.315.7109.629
Glas, G. (2019). Person-centered care in psychiatry. Self-relational,
contextual, and normative perspectives. Routledge.
Glas, G. (2021). Models of integration of christian worldview and
psychiatry. In J. R. Peteet, H. S. Moffic, A. Hankir, & H. G.
Koenig (Eds.), Christianity and psychiatry (pp. 163–180).
Springer International Publishing.
Harrer, M., Cuijpers, P., Furukawa, T. A., & Ebert, D. D. (2021).
Doing meta-analysis with R. A hands-on guide. Chapmann &
Hall/CRC Press. Welcome! | Doing Meta-Analysis in R
(bookdown.org).
Hittner, J. B., & Swickert, R. (2010). Discriminant ability of the
sense of coherence scale: Manageability, meaningfulness, and
comprehensibility as classifiers of depression. Individual
Differences Research, 8(3), 171–175.
Hodge, D. R. (2006). Spiritually modified cognitive therapy: A
review of the literature. Social Work, 51(2), 157–166. https:// doi.org/10.1093/sw/51.2.157
Hook, J. N., WorthingtonJr.E. L., Davis, D. E., JenningsII, D. J.,
Gartner, A. L., & Hook, J. P. (2010). Empirically supported
religious and spiritual therapies. Journal of Clinical Psychology,
66(1), 46–72. https://doi.org/10.1002/jclp.20626
∗Hsiao, F. H., Jow, G. M., Lai, Y. M., Chen, Y. T., Wang, K. C.,
Ng, S. M., Ho, R. T. H., Chan, C. L. W., & Yang, T. T.
(2011). The long-term effects of psychotherapy added to pharmacotherapy
on morning to evening diurnal cortisol patterns in
outpatients with major depression. Psychotherapy and
Psychosomatics, 80(3), 166–172. https://doi.org/10.1159/ 000321558
IntHout, J., Ioannidis, J. P. A., Rovers, M. M., & Goeman, J. J.
(2016). Plea for routinely presenting prediction intervals in
meta-analysis. BMJ Open, 6(7), e010247. https://doi.org/10. 1136/bmjopen-2015-010247
∗Johnson, W. B., DeVries, R., Ridley, C. R., Pettorini, D., &
Peterson, D. R. (1994). The comparative efficacy of
Christian and secular rational-emotive therapy with Christian
clients. Journal of Psychology and Theology, 22(2), 130–140.
https://doi.org/10.1177/009164719402200206
∗Johnson, W. B., & Ridley, C. R. (1992). Brief Christian and
non-Christian rational-emotive therapy with depressed
Christian clients: An exploratory study. Counseling and
Values, 36(3), 220–229. https://doi.org/10.1002/j.2161- 007X.1992.tb00790.x
Jüni, P., Witschi, A., Bloch, R., & Egger, M. (1999). The hazards
of scoring the quality of clinical trials for meta-analysis. The
Journal of the American Medical Association, 282(11), 1054
Knapp, G., & Hartung, J. (2003). Improved tests for a random
effects meta-regression with a single covariate. Statistics in
Medicine, 22(17), 2693–2710. https://doi.org/10.1002/sim. 1482
Kocsis, J. H., Gerber, A. J., Milrod, B., Roose, S. P., Barber, J.,
Thase, M. E., Perkins, P., & Leon, A. C. (2010). A new
scale for assessing the quality of randomized controlled
clinical trials of psychotherapy. Comprehensive Psychiatry,
51(3), 319–324. https://doi.org/10.1016/j.comppsych.2009. 07.001
∗Koszycki, D., Bilodeau, C., Raab-Mayo, K., & Bradwejn, J.
(2014). A multifaith spiritually based intervention versus supportive
therapy for generalized anxiety disorder: A pilot randomized
controlled trial. Journal of Clinical Psychology, 70(6),
489–509. https://doi.org/10.1002/jclp.22052
∗Koszycki, D., Raab, K., Aldosary, F., & Bradwejn, J. (2010). A
multifaith spiritually based intervention for generalized
anxiety disorder: A pilot randomized trial. Journal of Clinical
Psychology, 66(4), 430–441. https://doi.org/10.1002/jclp. 20663
Krueger, R. F., & Eaton, N. R. (2015). Transdiagnostic factors of
mental disorders. World Psychiatry, 14(1), 27–29. https://doi. org/10.1002/wps.20175
Lim, C., Sim, K., Renjan, V., Sam, H. F., & Quah, S. L. (2014).
Adapted cognitive-behavioral therapy for religious individuals
with mental disorder: A systematic review. Asian Journal of
Psychiatry, 9, 3–12. https://doi.org/10.1016/j.ajp.2013.12.011
Magaldi, D., & Trub, L. (2018). (What) do you believe?:
Therapist spiritual/religious/non-religious self-disclosure.
Psychotherapy Research, 28(3), 484–498. https://doi.org/10. 1080/10503307.2016.1233365
∗Margolin, A., Beitel, M., Schuman-Olivier, Z., & Avants, S. K.
(2006). A controlled study of a spirituality-focussed intervention
for increasing motivation for HIV prevention among
drug users. AIDS Education and Prevention, 18(4), 311–322.
https://doi.org/10.1521/aeap.2006.18.4.311
McCullough, M. E. (1999). Research on religion-accommodative
counseling: Review and meta-analysis. Journal of Counseling
Psychology, 46(1), 92–98. https://doi.org/10.1037/0022-0167. 46.1.92
∗Miller, W. R., Forcehimes, A., O’Leary, M. J., & LaNoue, M. D.
(2008). Spiritual direction in addiction treatment: Two clinical
trials. Journal of Substance Abuse Treatment, 35(4), 434–442.
https://doi.org/10.1016/j.jsat.2008.02.004
Norcross, J. C. (1990). An eclectic definition of psychotherapy. In
J. K. Zeig, & W. M. Munion (Eds.), What is psychotherapy?
Contemporary perspectives (pp. 218–220). Jossey-Bass.
Pargament, K. I. (2007). Spiritually integrated psychotherapy.
Understanding and addressing the sacred. The Guilford Press.
Park, C. L. (2021). Intrinsic and extrinsic religious motivation:
Retrospect and prospect. The International Journal for the
Psychology of Religion, 31(3), 213–222. https://doi.org/10. 1080/10508619.2021.1916241
Paukert, A. L., Phillips, L. L., Cully, J. A., Romero, C., & Stanley,
M. A. (2011). Systematic review of the effects of religiousaccommodative
psychotherapy for depression and anxiety.
Journal of Contemporary Psychotherapy, 41(2), 99–108. https:// doi.org/10.1007/s10879-010-9154-0
∗Pecheur, D. R., & Edwards, K. J. (1984). A comparison of
secular and religious versions of cognitive therapy with
depressed Christian college students. Journal of Psychology
and Theology, 12(1), 45–54. https://doi.org/10.1177/ 009164718401200106
Pew Research Centre. (2015, April 2). The future of world religions:
Population growth projections, 2010-2050. Department Religion.
https://www.pewresearch.org/religion/2015/04/02/religiousprojections- 2010-2050/#fn-22652-1.
∗Propst, L. R., Ostrom, R., Watkins, P., Dean, T., & Mashburn,
D. (1992). Comparative efficacy of religious and nonreligious
cognitive-behavioral therapy for the treatment of clinical
depression in religious individuals. Journal of Consulting and
Clinical Psychology, 60(1), 94–103. https://doi.org/10.1037/ 0022-006X.60.1.94
∗Razali, S. M., Aminah, K., & Khan, U. A. (2002). Religious-cultural
psychotherapy in the management of anxiety patients.
Transcultural Psychiatry, 39(2), 130–136. https://doi.org/10. 1177/136346150203900106
∗Razali, S. M., Hasanah, C. I., Aminah, K., & Subramaniam, M.
(1998). Religious-sociocultural psychotherapy in patients with
anxiety and depression. Australian and New Zealand Journal of
Psychiatry, 32(6), 867–872. https://doi.org/10.3109/ 00048679809073877
∗Rentala, S., Fong, T. C. T., Nattala, P., Chan, C. L. W., &
Konduru, R. (2015). Effectiveness of body-mind-spirit intervention
on well-being, functional impairment and quality of
life among depressive patients – A randomized controlled
trial. Journal of Advanced Nursing, 71(9), 2153–2163. https:// doi.org/10.1111/jan.12677
Richards, P. S., & Bergin, A. E. (2005). A spiritual strategy for counseling
and psychotherapy (2nd ed). American Psychological
Association.
∗Richards, P. S., Berrett, M. E., Hardman, R. K., & Eggett, D. L.
(2006). Comparative efficacy of spirituality, cognitive, and
emotional support groups for treating eating disorder inpatients. Eating Disorders, 14(5), 401–415. https://doi.org/10.1080/10640260600952548
Schaap-Jonker, H., van der Velde, N., Eurelings-Bontekoe,
E. H. M., & Corveleyn, J. M. T. (2017). Types of god representation
and mental health: A person-oriented approach. The
International Journal for the Psychology of Religion, 27(4), 199–
Schat, A., van Noorden, M. S., Giltay, E. J., Noom, M. J.,
Vermeiren, R. R. J. M., & Zitman, F. G. (2017).
Concordance between self-reported and observer-rated
anxiety severity in outpatients with anxiety disorders: The
Leiden routine outcome monitoring study. Psychology and
Psychotherapy: Theory, Research and Practice, 90(4), 705–719.
https://doi.org/10.1111/papt.12134
∗Shanke, A., Kalal, K., Save, D., & Sarve, P. (2017). Evaluation of
the effect of spiritual care on patients with generalized anxiety
and depression: A randomized controlled study. Psychology,
Health & Medicine, 22(10), 1186–1191. https://doi.org/10. 1080/13548506.2017.1290260
Smith, T. B., Bartz, J., & Richards, P. S. (2007). Outcome of religious
and spiritual adaptation to psychotherapy: A meta-analytic
review. Psychotherapy Research, 17(6), 643–655. https:// doi.org/10.1080/10503300701250347
∗Sreevani, R., Reddemma, K., Chan, C. L. W., Leung, P. P. Y.,
Wong, V., & Chan, C. H. Y. (2013). Effectiveness of integrated
body-mind-spirit group intervention on the well-being of
Indian patients with depression: A pilot study. The Journal of
Nursing Research, 21(3), 178–185. https://doi.org/10.1097/jnr. 0b013e3182a0b041
Stulp, H. P., Koelen, J., Schep-Akkerman, A., Glas, G. G., &
Eurelings-Bontekoe, E. H. M. (2019). God representations
and aspects of psychological functioning: A meta-analysis.
Cogent Psychology, 6(1), 1647926. https://doi.org/10.1080/ 23311908.2019.1647926
Thoma, N. C., McKay, D., Gerber, A. J., Milrod, B. L., Edwards,
A. R., & Kocsis, J. H. (2012). A quality-based review of randomized
controlled trails of cognitive-behavioral therapy for
depression: An assessment and metaregression. The American Journal of Psychiatry, 169(1), 22–30. https://doi.org/10.1176/appi.ajp.2011.11030433
∗Tonkin, K. M. (2005). Obesity, bulimia, and binge-eating disorder:
The use of a cognitive behavioral and spiritual intervention
[Unpublished doctoral dissertation], Bowling Green State
University, OH.
van Os, J., Guloksuz, S., Vijn, T. W., Hafkenscheid, A., &
Delespaul, P. (2019). The evidence-based group-level
symptom-reduction model as the organizing principle for
mental health care: Time for change? World Psychiatry, 18(1),
88–96. https://doi.org/10.1002/wps.20609
Viechtbauer, W. (2005). Bias and efficiency of meta-analytic variance
estimators in the random-effects model. Journal of
Educational and Behavioral Statistics, 30(3), 261–293. https:// doi.org/10.3102/10769986030003261
Viechtbauer, W. (2015). Conducting meta-analyse in R with
the metafor package. Journal of Statistical Software, 36(3),
1–48.
Viftrup, D. T., Hvidt, N. C., & Buus, N. (2013). Spiritually and
religiously integrated group psychotherapy: A systematic literature
review. Evidence-Based Complementary and Alternative
Medicine, 2013-10-31(article ID 274625-12), https://doi.org/ 10.1155/2013/274625
Vos, J., Craig, M., & Cooper, M. (2015). Existential therapies: A
meta-analysis of their effects on psychological outcomes.
Journal of Consulting and Clinical Psychology, 83(1), 115–128.
https://doi.org/10.1037/a0037167
Worthington, E. L., Hook, J. N., Davis, D. E., & McDaniel,
M. A. (2011). Religion and spirituality. In J. C. Norcross
(Ed.), Psychotherapy relationships that work: Evidence based
responsiveness (2nd ed.) (pp. 402–421). Oxford University
Press.
∗Zhang, Y., Young, D., Lee, S., Li, L., Zhang, H., Xiao, Z., Hao,
W., Feng, Y., Zhou, H., & Chang, D. F. (2002). Chinese
Taoist cognitive psychotherapy in the treatment of generalized
anxiety disorder in contemporary China. Transcultural
Psychiatry, 39(1), 115–129. https://doi.org/10.1177/ 136346150203900105