Psychiatry Research 104 2001.

99108

Religiosity and religious obsessions in

obsessivecompulsive disorder
Cenk Tek a,U , Berna Ulug b
a

Maryland Psychiatric Research Center and Department of Psychiatry, Uni ersity of Maryland, P.O. Box 21246, Baltimore,
MD 21228, USA
b
Department of Psychiatry, Hacettepe Uni ersity Faculty of Medicine, Ankara, Turkey
Received 27 March 2000; received in revised form 13 February 2001; accepted 6 May 2001

Abstract
Religion has often been thought to play a part in the genesis of some cases of obsessivecompulsive disorder
OCD.. In this study, we explored the relationship between religiosity, religious obsessions, and other clinical
characteristics of OCD. Forty-five outpatients with OCD were evaluated with the YaleBrown ObsessiveCompulsive Scale YBOCS. and the YaleBrown ObsessiveCompulsive Checklist YBOCC. as well as the Religious
Practices Index RPI., which was developed for this study. On the basis of these evaluations, 42% of the patients
were found to have religious obsessions. Despite differences in the frequency of religious obsessions found in this
study compared with others, a factor analysis revealed the symptom dimensions to be similar to those found in other
OCD samples. There was no significant difference in the overall severity of obsessions and compulsions between
patients with and without religious obsessions. RPI scores did not differ significantly between groups. We failed to
find a relationship between RPI scores or religious obsessions and any particular type of obsession or compulsion. A
logistic regression analysis revealed that the sole predictor of the presence of religious obsessions was a higher
number of types of obsessions. In conclusion, we failed to find a conclusive relationship between religiosity and any
other clinical feature of OCD, including the presence of religious obsessions. On the other hand, we showed that the
patients who tend to have a variety of obsessions are more likely also to have religious obsessions. Thus, religion
appears to be one more arena where OCD expresses itself, rather than being a determinant of the disorder. 2001
Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Cross-cultural; Turkey; Obsessions; Compulsions; Religion; Factor analysis

Corresponding author. Tel.: q1-410-642-1066 ext 5831; fax: q1-410-642-1884.

E-mail address: ctek@umaryland.edu C. Tek..

0165-1781r01r$ - see front matter 2001 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 1 6 5 - 1 7 8 1 0 1 . 0 0 3 1 0 - 9

100

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

1. Introduction
Religious obsessions constitute an interesting
component of the phenomenology of obsessive
compulsive disorder OCD.. Perhaps in their most
severe and continuous form, termed as scrupulosity, they attracted the attention of spiritual
authorities long before the definition of obsessional neurosis and at times were correctly recognized as a disease state Greenberg et al., 1987..
Early psychiatric theoreticians like Janet readily
classified these as psychiatric rather than religious
problems cited in Greenberg et al., 1987.. Freud
1961. originally 1907., without limiting his
theory to religious obsessions, went further and
proposed a relationship between obsessivecompulsive symptoms and religious practices, calling
obsessional neurosis an individual religion and
religion a universal obsessional neurosis.
Knowledge about OCD has significantly increased in the last two decades. Despite this,
systematic studies of religious obsessions, and of
the relationship between religious and obsessivecompulsive phenomena, have rarely been
performed. Although the epidemiology of OCD
appears to be stable across cultures Weissman et
al., 1994., patients with religious obsessions may
be over-represented in clinical populations of
Muslim and Jewish Middle Eastern cultures, as
compared with clinical populations from the West,
India and the Far East. The frequency of religious obsessions in clinical populations diagnosed
with OCD is reported to be 10% in the United
States Eisen et al., 1999., 5% in England Dowson, 1977., 11% in India Akhtar et al., 1975., and
7% in Singapore Chia, 1996. as compared with
60% in Egypt Okasha et al., 1994., 50% in Saudi
Arabia Mahgoub and Abdel-Hafeiz, 1991., 50%
in Israel Greenberg, 1984., and 40% in Bahrain
Shooka et al., 1998.. There are exceptions to this
trend such as a recent report of a large US
sample in which 27% of OCD patients suffered
from religious obsessions Mataix-Cols et al.,
1999. and a smaller US study that found 33% of
OCD patients to have religious obsessions
Steketee et al., 1991..
Turkey, with its geographical location bridging

between Western Europe and the Middle East,

its strictly secular state, and a predominantly liberal Muslim population, presents a unique opportunity for the study of religious obsessions. Two
studies from the west coast of Turkey reported
that religious obsessions in their clinical populations were 5 and 11.1%, respectively, rates similar to those found in Western studies Egrilmez et
al., 1997; Alptekin, 1991, unpublished dissertation.. We have previously reported a high frequency of OCD patients 48%. suffering from
religious obsessions in Turkeys capital city,
Ankara, which is located in the middle of the
Anatolian peninsula Tek et al., 1998.. It should
be noted that Hacettepe University Hospital,
where this study was conducted, is a tertiary care
center with an overabundance of referrals due to
treatment resistance andror complexity, with
patients from all over the country, although with
more from the middle and eastern parts of Turkey.
Finally, a 34% rate was reported from eastern
Turkey Tezcan and Millet, 1997.. It appears that
the frequency of religious obsessions in Turkish
OCD samples changes depending on the geographical location, possibly becoming higher when
the site of the study is closer to other countries
where high rates of religious obsessions are reported. Certainly many confounding factors may
be at play, as there are vast socioeconomic differences between different parts of Turkey e.g.
the West is richer and more modernized than the
East..
In this study, we attempted to further explore
this interesting clinical phenomenon, especially in
the context of the relationship between religiosity
and the presence of religious obsessions and other
disease variables.

2. Methods
The 45 subjects for this study were recruited
from outpatients at the Psychiatric Clinic of the
Hacettepe University Hospital in Ankara, Turkey.
Subjects were consecutive referrals within a 6month period in 1995 with a primary diagnosis of
OCD. All subjects, after a clinical interview by
two independent psychiatrists, who both used

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

DSM-IV checklists, met the criteria for current

OCD according to DSM-IV. To ensure a pure
obsessivecompulsive sample, exclusion criteria
included history of any psychotic disorder or
bipolar affective disorder, history of alcohol and
substance dependence, and history of any psychotic disorder in first-degree relatives, as determined in the clinical interview. After a complete description of the study to the subjects,
written informed consent was obtained.
All subjects were administered an inventory for
demographic information and psychiatric history,
Turkish versions of YaleBrown Obsessive-Compulsive Scale YBOCS. and the YaleBrown
Obsessive-Compulsive Checklist Y BOCC .
Goodman et al., 1989a,b; Tek et al., 1995., the
Maudsley Obsessive Compulsive Inventory
MOCI., the 17-item Hamilton Depression Rating Scale HAM-D., and the Religious Practices
Index, a scale developed for this study. All the
raters had extensive experience in using these
instruments. Our group Tek et al., 1995. had
previously demonstrated excellent inter-rater reliability in the use of the Turkish version of the
YBOCS.
The presence of religious obsessions was determined by the YBOCC. The definition used
for a religious obsession in this study was either
excessive concern with andror observance of religious teachings and rules, far exceeding what is
practiced in the patients religious reference
group; or excessive concern andror intrusive and
repetitive thoughts about sacrilege, blasphemy,
and sin. Since all the subjects were Muslim, and
Islamic belief involves strong contamination
themes, as well as ritualized cleaning and prayers
in predetermined numbers, some of our patients
had contamination obsessions andror cleaning,
checking, and counting compulsions related but
not necessarily limited to worshipping. These were
rated in their corresponding classes in the
YBOCC rather than as religious obsessions.
The Religious Practices Index RPI. was developed for this study. Since religious practices
change widely from religion to religion, even
within the same religion from order to order or in
different locales, we felt the need to develop a
versatile instrument that would be as objective as

101

possible and adaptable to different religions or

cultures. The RPI is a four-point global measure
of religiosity 1 s no religious affiliation except by
culture, no religious activity; 2 s specific religious
affiliation with limited religious activity; 3 s
significant religious activity but religious rules do
not dictate all areas of daily life; and 4 s very
religious.. Specific anchors had been used for
Islam to increase inter-rater reliability. The need
for anchors arose from the fact that the raters
perception of the religious activity of others was
most likely to be associated with their own experiences, which vary from individual to individual. In
developing anchors, we took into account both
proscriptions of the religion and the realities of
the society. For example, abstinence from eating
pork is rather a meaningless anchor in Turkish
society since this type of meat is not regularly
available. However, abstinence from eating pork
might prove to be a useful anchor in another
European country. On the other hand, although
alcohol is definitely prohibited by Islam, alcoholic
beverages are readily available without any serious restrictions, and it is not unusual to serve
alcoholic beverages in special occasions like wedding ceremonies or New Years Eve celebrations;
thus, the occasional recreational use of alcohol
proved to be a useful anchor, disqualifying an
individual from a rating of very religious. This
might not be the case in certain Islamic countries
where alcoholic beverages are actually illegal. At
every step of the RPI interview, we made an
effort to note the exceptions by asking the patients
the practices of their family members and other
people in their specific culture. Unfortunately,
population data are not available for this instrument. However, face validity exists, and in our
application to 50 clinic patients a close to normal
distribution was obtained data available from
authors on request.. The intra-class correlation
coefficient for 15 patients and seven independent
raters with audiotaped interviews was 0.94 P0.001.; the lowest Spearman correlation coefficient between rater pairs was 0.71 P- 0.01..
Clinical and demographic variables of the
groups with and without religious obsessions were
compared by Students t or chi-square tests as
appropriate. We examined the correlation

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

102

between the RPI score and scores on the clinical

rating scales. A principal component analysis with
Varimax rotation was used to obtain major symptom factors as well as to explore the relationship
of religious obsessions to other symptom categories, and factor scores were calculated. To explore the relationship between the RPI score and
any particular type of symptom dimension, linear
regression analysis was used. Finally, to look for
general predictors of the presence of religious
obsessions, we performed a logistic regression
analysis of religious obsessions, on the one hand,
and age, sex, duration of illness, RPI score,
YBOCS score, number of main categories of
obsessions and compulsions separately, as rated
on the YBOCC, on the other hand. Statistical
procedures were done with the aid of SPSS for
Windows, 10.0 software.

3. Results
Nineteen subjects had religious obsessions, as
compared with 26 subjects without them. Table 1
presents relevant socio-demographic variables.
Patients with religious obsessions were significantly younger than patients without them mean
difference in age s 8.3 years.. There were no sex
differences between the two groups, though there

were more females in both groups. Educational

levels, marital status, socioeconomic levels according to physicians evaluation, and residence
in a rural or urban location did not differ significantly in the two groups. Mean RPI scores also
were not significantly different.
As shown in Table 2, YBOCS scores, obsession and compulsion subscores, and MOCI scores
were not significantly different between the two
groups. The patients with religious obsessions reported significantly more obsession categories but
not compulsion categories than the group without
religious obsessions. Time since the first diagnosis
was longer in the group without religious obsessions, though this difference did not reach statistical significance mean difference of 33.1
months.. A family history of OCD characterized
41% of the patients, a percentage that was similar
in the two groups. One patient with religious
obsessions met the criteria for accompanying major depression as compared with six patients in
the group without religious obsessions. Mean
HAM-D scores were not significantly different in
the groups, and symptom composition was also
similar.
Principal component analysis revealed a fivefactor solution, which explained 65.5% of the
variance. The first factor contaminationrcleaning. explained 17.8% of the variance and included

Table 1
Sociodemographic characteristics

Age wmean years S.D..x

Sex w N female %.x
Education wmean years S.D..x
Socioeconomic status 05
wmean S.D..x, range
Marital status
w N married %.x
Location lived
w N urban %.x
Religious practices index RPI.
wmean score S.D..x, range 14
U

NS means P) 0.05.

Patients with
religious obsessions
N s 19

Patients without
religious obsessions
N s 26

Statistic

25.9 7.5.
14 73.7.
11.2 5.0.
3.0 0.9.

34.2 11.9.
21 80.8.
9.6 4.6.
2.7 1.0.

t s 2.66 d.f.s 43.

2 s 0.32 d.f.s 1.
t s y0.98 d.f.s 43.
t s y0.95 d.f.s 42.

- 0.02
NS
NS
NS

8 42.1.

14 56.0.

2 s 1.90 d.f.s 2.

NS

16 84.2.

21 80.8.

2 s 0.89 d.f.s 1.

NS

2.6 1.1.

2.7 1.0.

t s 0.33 d.f.s 42.

NS

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

103

Table 2
Clinical characteristics

Time since first diagnosis

wmean months S.D..x
Family history of OCD
w N positive %.x
MOCI total score
wmean S.D..x
YBOCS total score
wmean S.D..x
YBOCS obsession score
wmean S.D..x
YBOCS compulsion score
wmean S.D..x
YBOCC number of obsessions
wmain categories mean S.D..x
YBOCC number of compulsions
wmain categories mean S.D..x
HAM-D total score
wmean S.D..x
Major depression
DSM-IV., w N diagnosed %.x

Patients with
religious obsessions
N s 19

Patients without
religious obsessions
N s 26

Statistic

41.1 28.9.

74.2 70.2.

t s 1.42 d.f.s 42.

NS

8 42.1.

10 40.0.

2 s 0.02 d.f.s 1.

NS

20.2 5.95.

21.9 7.1.

t s 0.82 d.f.s 42.

NS

24.8 6.2.

25.0 8.3.

t s 0.11 d.f.s 43.

NS

12.2 3.5.

12.5 4.1.

t s 0.25 d.f.s 43.

NS

12.6 3.4.

12.6 5.2.

t s y0.03 d.f.s 43.

NS

3.84 0.90.

2.19 0.80.

t s y6.48 d.f.s 43.

- 0.001

3.37 1.12.

2.81 1.47.

t s y1.39 d.f.s 43.

NS

7.32 4.83.

10.88 6.87.

t s 1.94 d.f.s 43.

NS

2 s 2.65 d.f.s 1.

NS

1 5.3.

6 23.1.

YBOCS, YaleBrown Obsessive-Compulsive Scale; YBOCC, YaleBrown Obsessive-Compulsive Checklist; MOCI, Maudsley Obsessive-Compulsive Inventory; HAM-D, Hamilton Depression Scale 17 items..
U
NS means P) 0.05.

contamination obsessions, cleaning and repeating

compulsion categories with loadings ranging
between 0.82 and 0.60. The second factor symmetryrordering, 15.2% of the variance. included
symmetry and somatic obsessions and ordering
compulsions with loadings 0.870.5. The third

factor aggressivercounting, 13.7% of the variance. included aggressive obsessions and counting
compulsions with loadings of 0.72 and 0.54, respectively. Hoarding compulsions were also found
in this group, with a weaker 0.19 loading. The
fourth factor sexualrreligious obsessions, 9.7%

Table 3
Major symptom factors
U

Symptom factors

Patients with
religious obsessions
N s 19
Mean S.D..

Patients without
religious obsessions
N s 26
Mean S.D..

Contaminationrcleaning
Symmetryrordering
Aggressivercounting
Sexualrreligious obsessions
Checkingrhoarding compulsions
Level of insight into symptoms
YBOCS item 11, range 04.

2.74 0.56.
1.47 1.17.
1.37 0.68.
1.37 0.5.
0.26 0.45.
1.00 1.05.

2.27 1.08.
0.96 1.0.
1.08 0.8.
0.27 0.45.
0.42 0.58.
1.12 1.18.

t s y1.72 d.f.s 43.

t s y1.58 d.f.s 43.
t s y1.29 d.f.s 43.
t s y7.73 d.f.s 43.
t s 1.00 d.f.s 43.
t s 0.34 d.f.s 43.

NS
NS
NS
- 0.001
NS
NS

NS means P) 0.05

104

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

of the variance. included sexual and religious

obsessions with loadings of 0.69 and 0.55, respectively. This factor was the only pure obsessive
factor. The fifth factor checkingrhoarding compulsions, 9.2% of the variance. included checking
and hoarding compulsions with loadings of 0.88
and 0.18, respectively, and was the only pure
compulsive factor. This five-factor solution is very
similar to that found in the much larger study of
Mataix-Cols et al. 1999. and other similar studies
Baer, 1994; Leckman et al., 1997.. Table 3 presents scores derived from this solution as well as
the level of insight about symptoms for the groups
with and without religious obsessions.
There were no correlations between RPI scores
and YBOCS total scores and subscores, MOCI
total scores, or the total number of categories of
obsessions or compulsions highest r s 0.25,
lowest Ps 0.10.. Interestingly, the RPI score was
negatively correlated with years of education r s
y0.56, P- 0.001. and socioeconomic level r s
y0.33, P- 0.05.. For this reason we felt the need
to take into account demographic variables age,
sex, educational level, and socioeconomic level.
before we explored the relationship between religiosity and major symptom factors. A linear regression analysis with the RPI score as the dependent variable, and the four demographic variables
and the five symptom factors as the independent
variables, confirmed a significant negative association between educational level and RPI score
s y0.50, partial r s y0.46, d.f.s 9,34, P0.005.. There was a strong but non-significant
association between the contaminationrcleaning
factor and the RPI score s 0.30, partial r s
0.32, Ps 0.055.. The remaining independent variables, including the religiousrsexual obsessions
factor, were not significantly associated with RPI
scores highest s 0.19, lowest Ps 0.17..
The presence of religious obsessions, on the
one hand, and age, sex, educational level, socioeconomic level, time since first diagnosis, RPI
score, YBOCS score, and number of categories
of obsessions and compulsions, separately as rated
on the YBOCC, on the other hand, entered a
logistic regression analysis. The total number of
categories of obsessions emerged as the single
significant predictor for the presence of religious

obsessions R s 0.32, Wald 2 s 8.00, d.f.s 1,

P- 0.005.. The remaining variables, including the
total number of compulsions and the RPI score,
were not significant predictors highest R s 0.09,
highest Wald 2 s 2.43, d.f.s 1, lowest Ps 0.12.

4. Discussion
Our study failed to demonstrate a conclusive
relationship between religiosity and religious
obsessions or other disease variables. Patients
with religious obsessions were younger and more
likely to suffer from a greater variety of obsessions but not compulsions. Apart from this,
patients with and without religious obsessions
seemed to have similar clinical characteristics.
Consistent with previous studies, five symptom
dimensions emerged from a factor analysis, and
religious obsessions loaded on a pure obsessions
factor together with sexual obsessions. However,
the level of religiosity was not associated with this
dimension. Instead, perhaps reflecting the strong
contamination themes and cleaning rituals in Islam, there was a non-significant but strong association between the contaminationrcleaning factor
and religiosity.
There are a limited number of studies about
religious obsessions. Previously, Lewis 1994. tried
to test Freuds observation of the resemblance
between religious practices and obsessive actions
in a community Catholic sample and failed to find
a relationship between obsessionality and religious practices. Our study was not designed to
test Freuds ideas on this matter. Nevertheless,
our findings do not present a hint of a relationship between religious practices and obsessive
compulsive psychopathology. Greenberg, who
presented various case studies on religious obsessions, argues that religious obsessions can be separated readily from religious rites, especially by
the distress and resistance they cause, and are not
necessarily related to religion itself Greenberg,
1984; Greenberg et al., 1987.. It is not clear if
different religions have different impacts on
obsessivecompulsive psychopathology. All religions by their nature involve rituals to a certain
extent, perhaps some religions more than others.

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

Islam is a very ritualistic religion. However,

Okasha et al. 1994., who presented a large Egyptian sample with religious obsessions noted that
their Christian patients, who constituted 10% of
the sample, seemed to suffer as much as their
Muslim patients from obsessions with religious
themes. In a study from the United States, Steketee et al. 1991. did not report any differences
between Catholic and other OCD patients in
terms of religious obsessions. However, they did
find an association between religious obsessions,
overall illness severity, and self-reported level of
religiosity. We failed to show a similar association
in this study between religious obsessions and
religiosity or overall illness severity. Since both
studies used very similar instruments, with the
exception that their religiosity measure was a
self-reported one as compared with a clinician-reported one in our study, it may be that patients
with religious obsessions have a self-perception of

105

religious devotion greater than what is objectively

observed from the extent of their religious practices.
Despite problems of statistical power attributable to the small number of subjects, the
five-factor symptom structure found in our study
resembles that found in previously published
studies with much larger sample sizes Baer, 1994;
Leckman et al., 1997; Mataix-Cols et al., 1999..
Sexual and religious obsessions consistently load
on the same factor across studies from different
cultures, a factor that generally represents a pure
obsessions symptom dimension. In the current
study, patients with religious obsessions seemed
to have a more varied number of types of obsessions, but not compulsions, compared with the
patients without religious obsessions. Moreover,
the number of different types of obsessions, but
not compulsions, emerged as a good predictor of
the presence of religious obsessions. All these

Table 4
Common themes of obsessions in different cultures % frequency.

USA Rhode Island.

Rasmussen and Eisen, 1992.
USA Massachusetts .
Mataix-Cols et al., 1999.
USA Connecticut.
Holzer et al., 1994.
England
Dowson, 1977.
Turkey West.
Egrilmez et al., 1997.
Turkey
Ankaracurrent study.
Turkey East.
Tezcan and Millet, 1997.
Israel
Greenberg, 1984.
Bahrain
Shooka et al., 1998.
Egypt
Okasha et al., 1994.
Saudi Arabia
Mahgoub and Abdel-Hafeiz, 1991.
India
Akhtar et al., 1975.
Singapore
Chia, 1996.

Contamination

Aggression

Religious

Sexual

Symmetryr exactness

50

31

10

24

32

59

76

27

19

45

62

77

26

23

67

54

32

12

53

11

13

78

84

42

31

51

51

21

34

18

40

20

50

10

10

38

40

32

60

41

60

47

53

41

21

50

46

29

11

10

10

35

16

10

11

106

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

findings suggest that the group with religious

obsessions may represent a predominantly pure
obsessive group among OCD patients. In that
case, since the emphasis is on the type of symptom i.e. obsessions vs. compulsions. rather than
on the subject matter, the lack of a significant
association between religiosity and religious
obsessions is not surprising. Other support for
this possibility comes from the fact that religious
obsessions seem to respond to conventional treatment for OCD as well as other types of obsessions do Fallon et al., 1990; Mataix-Cols et al.,
1999.. Thus, our findings support the notion that
apart from being a subject for obsessions, religion
itself does not seem to be directly related to
obsessional symptoms. The hint of an association
between the level of religiosity and the contaminationrcleaning symptom dimension is not necessarily contrary to this notion. Contamination
themes and cleaning rituals are very dominant in
Islamic practice. A practicing Muslim has to pray
five times during the day, and before each prayer
there is a cleaning ritual involving washing certain
body parts in predetermined numbers. The clean
state can easily be annulled by certain thoughts
and actions, which might necessitate the repetition of the whole cleaning process. Thus, the
Islamic religion in a way prepares the perfect
cultural setting for expression of symptoms included in this factor, namely contamination
obsessions, cleaning and repeating compulsions.
The age difference between the groups with
and without religious obsessions might have suggested a waning of religious obsessions over the
years. In that case, though, we would have expected to see a predictive relationship between
age and the religious obsessions, which was not
the case.
It should be noted that this study has some
shortcomings. Our sample size is small and, as
pointed out in Section 1, we could not control for
a sampling bias towards more chronic and complicated cases. To our surprise, many patients, when
they were administered the YBOCC, reported
types of symptoms that they had not revealed
during their tenure in our clinic, revealing a
greater variety of symptoms than many other
studies reported. However, the results of the fac-

tor analysis and the order of frequencies seemed

similar to those found in other samples from
various cultures, despite the differences in frequency of various symptoms Table 4..
Although our hypothesis that the extent of
religious practices is a measure of religiosity
has face validity, religiosity is an extremely difficult concept to measure. The Religious Practices
Index RPI. has not been tested in the community from which our sample is derived, so population norms are not available. Therefore, it is
difficult to draw inferences from findings such as
a negative correlation between educational level
or socioeconomic status and the RPI score.
Nevertheless, very high interrater reliability
among raters from the same community supports
the face validity of the RPI.
Why the prevalence of religious obsessions is
higher in certain communities remains an unanswered question. It is possible that the types of
symptoms are culturally determined since the
variance is not confined to the religious obsessions, but symptomatic presentation varies widely
from culture to culture as can be seen in Table 4.
Religion may have a greater impact in certain
cultures than in others, contributing to the expression of religious obsessions or causing family
and peer pressure for mental health care due to
the very nature of the symptoms; thus, religious
obsessions may be overrepresented in clinical
samples from some cultures. Yet another reason
for the possible overrepresentation of religious
obsessions in certain areas of the world might be
limited access to mental health care. If that is the
case, perhaps patients with a greater number of
symptoms are presenting to the clinics in these
areas, a variable that is associated with the presence of the religious obsessions in this study.
There is also the possibility of underdiagnosis of
religious obsessions in certain cultures. A recent
study by Mataix-Cols et al. 1999. supports the
possibility that religious obsessions could have
been underdiagnosed or underreported in earlier
studies from the United States, since most often
they did not represent the major symptom.
In conclusion, despite the similarities between
obsessivecompulsive psychopathology and religious phenomena as observed by Freud, there

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

does not seem to be a strong relationship between

religion and OCD. Instead of being a determinant of the disorder, religion appears to be just
another arena where OCD expresses itself. Certain symptoms may be more prevalent in different
cultures. However, similar results of factor-analytic studies across cultures provide evidence that
the overall phenomenology of the illness does not
vary from culture to culture. Clinicians should be
sensitive to the fact that religious obsessions may
be more prevalent in certain cultures with which
they may not be well acquainted. Nevertheless,
religious obsessions should be treated as obsessions rather than religious phenomena.

Acknowledgements
The authors are grateful for invaluable assistance from Aylin Ulusahin, M.D., and Ahsen
Orhon, M.D. The study is dedicated to the memory of Professor Isik Savasir, Ph.D.; without her
help, this study would never have materialized.
References
Akhtar, S., Wig, N.N., Varma, V.K., Pershad, D., Verma, S.K.,
1975. A phenomenological analysis of symptoms in obsessive-compulsive neurosis. British Journal of Psychiatry 127,
342348.
Alptekin, K., 1991. Clinical Characteristics and Magnetic Resonance Imaging of the Brain in Obsessive Compulsive
Disorder in Turkish.. Unpublished Dissertation, Dokuz
Eylul
University, Izmir.
Baer, L., 1994. Factor analysis of symptom subtypes of obsessive compulsive disorder and their relation to personality
and tic disorders. Journal of Clinical Psychiatry 55 Suppl..,
1823.
Chia, B.H., 1996. A Singapore study of obsessive compulsive
disorder. Singapore Medical Journal 37 4., 402406.
Dowson, J.H., 1977. The phenomenology of severe
obsessivecompulsive neurosis. British Journal of Psychiatry 131, 7578.
Egrilmez, A., Gulseren, L., Gulseren, S., Kultur, S., 1997.
Phenomenology of obsessions in a Turkish series of OCD
patients. Psychopathology 30 2., 106110.
Eisen, J.L., Goodman, W.K., Keller, M.B., Warshaw, M.G.,
DeMarco, L.M., Luce, D.D., Rasmussen, S.A., 1999. Patterns of remission and relapse in obsessive-compulsive disorder: a 2-year prospective study. Journal of Clinical Psychiatry 60, 346351.

107

Fallon, B.A., Liebowitz, M.R., Hollander, E., Schneier, F.R.,

Campeas, R.B., Fairbanks, J., Papp, L.A., Hatterer, J.A.,
Sandberg, D., 1990. The pharmacotherapy of moral or
religious scrupulosity. Journal of Clinical Psychiatry 51 12.,
517521.
Freud, S., 1961. Obsessive actions and religious practices.. In:
Strachey, J. Ed.., The Standard Edition of the Complete
Psychological Works of Sigmund Freud original work published in 1907., 9. Hogarth Press, London, pp. 117127.
Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C.,
Fleischmann, R.L., Hill, C.L., Heninger, G.R., Charney,
D.S., 1989a. The YaleBrown Obsessive Compulsive Scale.
I. Development, use, and reliability. Archives of General
Psychiatry 46 11., 10061011.
Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C.,
Delgado, P., Heninger, G.R., Charney, D.S., 1989b. The
YaleBrown Obsessive Compulsive Scale. II. Validity.
Archives of General Psychiatry 46 11., 10121016.
Greenberg, D., 1984. Are religious compulsions religious or
compulsive: A phenomenological study. American Journal
of Psychotherapy 38 4., 524532.
Greenberg, D., Witztum, E., Pisante, J., 1987. Scrupulosity:
religious attitudes and clinical presentations. British Journal of Medical Psychology 60, 2937.
Holzer, J.C., Goodman, W.K., McDougle, C.J., Baer, L., Boyarsky, B.K., Leckman, J.F., Price, L.H., 1994. Obsessive
compulsive disorder with and without tic disorder. British
Journal of Psychiatry 164, 469473.
Leckman, J.F., Grice, D.E., Boardman, J., Zhang, H., Vitale,
A., Bondi, C., Alsobrook, J., Peterson, B.S., Cohen, D.J.,
Rasmussen, S.A., Goodman, W.K., McDougle, C.J., Pauls,
D.L., 1997. Symptoms of obsessivecompulsive disorder.
American Journal of Psychiatry 154 7., 911917.
Lewis, C.A.J., 1994. Religiosity and obsessionality: the relationship between Freuds religious practices. Journal of
Psychology 128 2., 189196.
Mataix-Cols, D., Rauch, S.L., Manzo, P.A., Jenike, M.A.,
Baer, L., 1999. Use of factor-analyzed symptom dimensions
to predict outcome with serotonin reuptake inhibitors and
placebo in the treatment of obsessivecompulsive disorder.
American Journal of Psychiatry 156 9., 14091416.
Mahgoub, O.M., Abdel-Hafeiz, H.B., 1991. Pattern of obsessivecompulsive disorder in eastern Saudi Arabia. British
Journal of Psychiatry 158, 840842.
Okasha, A., Saad, A., Khalil, A.H., el Dawla, A.S., Yehia, N.,
1994. Phenomenology of obsessivecompulsive disorder: a
transcultural study. Comprehensive Psychiatry 35 3.,
191197.
Rasmussen, S.A., Eisen, J.L., 1992. The epidemiology and
clinical features of obsessive compulsive disorder. Psychiatric Clinics of North America 15 4., 743758.
Shooka, A., al-Haddad, M.K., Raees, A., 1998. OCD in
Bahrain: a phenomenological profile. International Journal
of Social Psychiatry 44 2., 147154.
Steketee, G., Quay, S., White, K., 1991. Religion and guilt in
OCD patients. Journal of Anxiety Disorders 5, 359367.

108

C. Tek, B. Ulug r Psychiatry Research 104 (2001) 99108

Tek, C., Ulug, B., Rezaki, B.G., Tanriverdi, N., Mercan, S.,
Demir, B., Vargel, S., 1995. YaleBrown Obsessive Compulsive Scale and US National Institute of Mental Health
Global Obsessive Compulsive Scale in Turkish: reliability
and validity. Acta Psychiatrica Scandinavica 91 6., 410413.
Tek, C., Ulug, B., Ulusahin, A., Orhon, A., 1998. Phenomenology of OCD in a Turkish sample. Poster presentation. Annual Meeting of the American Psychiatric Association, Toronto, Canada.

Tezcan, E., M illet, B., 1997. Phenomenology of

obsessivecompulsive disorders. Forms and characteristics
of obsessions and compulsions in East Turkey article in
French.. Encephale Sep-Oct 23 5., 342350.
Weissman, M.M., Bland, R.C., Canino, G.J., Greenwald, S.,
Hwu, H.G., Lee, C.K., Newman, S.C., Oakley-Browne, M.A.,
Rubio- Stipec, M., Wickramaratne, P.J., 1994. The cross
national epidemiology of obsessive compulsive disorder.
Journal of Clinical Psychiatry 55 Suppl.., 510.