Psychiatry Research 127 (2004) 267 – 278

www.elsevier.com/locate/psychres

Perceived parental rearing style in obsessive–compulsive disorder:

relation to symptom dimensions
Pino Alonso a, José M. Menchón a, David Mataix-Cols b, Josep Pifarré a,
Mikel Urretavizcaya a, José M. Crespo a, Susana Jiménez a,
Gema Vallejo a, Julio Vallejo a,*
a
Obsessive – Compulsive Disorder Clinical and Research Unit, Department of Psychiatry, Hospital Prı́ncipes de España,
Ciudad Sanitaria y Universitaria de Bellvitge, c/ Feixa Llarga s/n 08907, Hospitalet de Llobregat, Barcelona, Spain
b
Department of Psychiatry, Imperial College of Science, Technology and Medicine, Charing Cross Hospital, London, UK
Received 2 May 2002; received in revised form 19 November 2002; accepted 18 February 2003

Abstract

Obsessive – compulsive disorder (OCD) runs in families, but the specific contribution of genetic and environmental factors
to its development is not well understood. The aim of this study was to assess whether there are differences in perceived
parental child-rearing practices between OCD patients and healthy controls, and whether any relationship exists between
parental characteristics, depressive symptoms and the expression of particular OCD symptom dimensions. A group of 40
OCD outpatients and 40 matched healthy controls received the EMBU (Own Memories of Parental Rearing Experiences in
Childhood), a self-report measure of perceived parental child-rearing style. The Yale – Brown Obsessive – Compulsive Scale
(Y – BOCS) and the Hamilton Depression Rating Scale (HDRS) were used to assess the severity of obsessive – compulsive
and depressive symptoms. The Y – BOCS Symptom Checklist was used to assess the nature of obsessive – compulsive
symptoms, considering the following five symptom dimensions: contamination/cleaning, aggressive/checking, symmetry/
ordering, sexual/religious and hoarding. Logistic and multiple linear regression analyses were conducted to study the
relationship between parental style of upbringing, depressive symptoms and OCD symptom dimensions. Severe OCD (Y –
BOCS: 27.0 F 7.4) and mild to moderate depressive symptoms (HDRS: 14.0 F 5.4) were detected in our sample. Compared
with healthy controls, OCD patients perceived higher levels of rejection from their fathers. No differences between the
groups with respect to perceived levels of overprotection were detected. The seventy of depressive symptoms could not be
predicted by scores on any perceived parental characteristics. Hoarding was the only OCD symptom dimension that could be
partially predicted by parental traits, specifically low parental emotional warmth. Social/cultural variables such as parental
child-rearing patterns, in interaction with biological and genetic factors, may contribute to the expression of the OCD
phenotype.
D 2004 Published by Elsevier Ireland Ltd.

Keywords: Family; Symptom dimensions; Depression

1. Introduction
* Corresponding author. Tel.: +34-0-93-2607659; fax: +34-0-
93-2607658. Obsessive – compulsive disorder (OCD) is defined
E-mail address: jvallejo@csub.scs.es (J. Vallejo). by the presence of obsessions or compulsions that

0165-1781/$ - see front matter D 2004 Published by Elsevier Ireland Ltd.

doi:10.1016/j.psychres.2001.12.002
268 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

are a significant source of distress or that interfere directly questioning parents, and direct observations
with the patient’s social functioning (American Psy- have been rarely conducted. Other methodological
chiatric Association, 1994). Although current neuro- weaknesses include the use of a great variety of
biological theories of OCD emphasize the implication methods and more or less reliable measures to assess
of dysfunctional corticostriatal circuits in the etiology parental characteristics, small sample sizes and the
of the disorder, behavioral theorists have suggested lack of appropriate comparison groups. Despite all
that social learning factors may also contribute to its these limitations, results indicate that a small but
development in biologically vulnerable subjects. Pa- significant amount of variance in anxiety and de-
rental child-rearing patterns have been proposed as pression may be accounted for by perceived parental
one of these social factors, but no agreement has been rejection and control.
reached either on the exact influence of parental Few studies concerning the influence of early
behaviors in the development of OCD or on its parenting behaviors and attitudes in the development
relationship to specific obsessive – compulsive symp- of OCD have been conducted hitherto. Sub-clinical
toms (Hoover and Insel, 1984; Rasmussen and obsessive – compulsive subjects have been reported to
Tsuang, 1984). perceive their parents as more rejecting, overprotect-
Parental behaviors, especially concerning the abil- ing and less emotionally warm than normal controls
ity to express affection and emotional warmth and to (Ehiobuche, 1988; Kimidis et al., 1992; Cavedo and
avoid excessive protection, control and criticism, Parker, 1994). Results obtained from clinical samples
seem to be important in the development of a are controversial and often contradictory. Hafner
healthy personality. Rejecting and controlling parent- (1988) described high levels of parental overprotec-
ing styles have been described as being associated tion in 81 subjects (mean age 35.7 F 12.5) who were
with a variety of forms of psychopathology, includ- registered as sufferers in the Obsessive –Compulsive
ing depression, schizophrenia, anxiety disorders, Neurosis Support Group of South Australia. Subjects
substance abuse, oppositional child behavior and completed the Parental Bonding Instrument (PBI), a
eating disorders (Parker et al., 1987; Gerlsma and self-report measure of an individual’s perception of
Emmelkamp, 1990; De Rutter, 1994; Rapee, 1997). his or her parents’ rearing practices up to the age of
Most of the studies in this area have focused on the 16 years. Methodological weaknesses of the study
relationship between anxiety disorders and depres- include the fact that subjects were not directly
sion and parental characteristics. In a review of the interviewed by the author, diagnosis was established
literature related to this last issue, Rapee (1997) only on the basis of the results of the Padua Inven-
describes two main child-rearing factors. One, which tory of Obsessions and Compulsions and the Brief
includes behaviors and attitudes related to negative Symptoms Inventory, and the absence of a specific
or hostile feelings toward the child, is termed rejec- control group. Employing the EMBU, a self-report
tion or criticism. The second factor, which refers to measure of an individual’s perception of his or her
behaviors designed to protect the child from possible parent’s rearing style during childhood, Hoekstra et
harm, is called parental control or protection. A al. (1989) compared 119 compulsive checkers and
rearing style characterized by low parental affection cleaners divided into four groups with 277 non-
and high parental control appears to be related to clinical controls. OCD subjects perceived more re-
anxiety disorders and depression, with the most jection and less emotional care from their parents
consistent results obtained for social phobia. Inter- than healthy controls, with higher levels of parental
estingly, some data appear to indicate a somewhat overprotection being reported only by compulsive
stronger relationship between parental rejection and washers. Vogel et al. (1997) employed the PBI to
depression and between parental control and anxiety. compare self-reported patterns of parental bonding in
Nevertheless, investigation in this area suffers from 26 OCD (mean age 35.2 F 12.1), 34 depressed (mean
great methodological limitations. Most studies have age 38.8 F 9.2) and 41 healthy subjects (mean age
employed retrospective self-report measures given to 37.8 F 11.2). Patients with a principal diagnosis of
the offspring, a considerably smaller number of major depressive disorder experienced significantly
studies have examined child-rearing attitudes by lower levels of parental care and significantly higher
 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278 269

levels of maternal overprotection than healthy con- who induce excessive fears of making mistakes in
trols, while no significant difference between OCD their children. However, this hypothesis has been
and normal subjects was detected. Since the presence only partially supported by later studies. While
of an additional diagnosis of depression in the OCD Turner et al. (1979) reported no significant differ-
group (46% of the patients) was associated with ences between washers and checkers regarding fear
significantly lower levels of parental care and higher of criticism, Steketee et al. (1985) found that check-
levels of parental overprotection, the authors sug- ers more often perceived their mothers as meticulous
gested that this parental rearing style may act as a and demanding than washers did. Neither study
vulnerability factor more specifically related to the found any significant differences regarding overpro-
development of depressive disorders than to OCD. tection between washers and checkers.
The principal limitation of this study comes from the Nevertheless, on studying influences of parental
reduced number of OCD patients included in the behaviors in the development of obsessive – compul-
analyses. Finally, Turgeon et al. (2002) have recently sive symptoms, one must not forget that OCD can
employed the PBI and the EMBU to compare have a devastating effect on the quality of family life
recalled parental behaviors among 43 out-patients (Steketee and Pruyn, 1998). Many families become
with OCD, 38 out-patients with panic disorder with dysfunctional as a result of a family member’s OCD
agoraphobia (PDA) and 120 non-anxious controls. symptoms. Frequently, parents and siblings become
Patients with OCD and PDA did not significantly involved in the sufferer’s avoidance behaviors and
differ on mean scores on any of the PBI and EMBU compulsions in an effort to relieve the fear and
scales. Participants with anxiety disorders compared anxiety that the patient is feeling. Family and lei-
with the non-anxious group rated both their mothers sure-time routines and activities are frequently mod-
and fathers as more protective. No differences were ified to accommodate the OCD sufferer. All these
found between the anxious and non-anxious groups efforts often lead relatives to experience severe feel-
on the Emotional Warmth, Rejection and Care scales. ings of frustration, anger, guilt and loneliness. Child-
Limitations of this study include the fact that patients rearing patterns may play a role in the development
were not recruited from hospital settings but through of OCD, but one should also consider that the
advertisements in the media, so they constitute a self- primary presence of obsessive – compulsive symp-
selected sample, which may not represent general toms in a child may also elicit certain parental
OCD patients. Another possible weakness may come behaviors and attitudes, especially a tendency to
from the recruitment of the control group, since greater rejection and/or protection towards the affect-
healthy comparison subjects were not directly inter- ed child.
viewed by the researchers but just psychiatrically Thus, the role of parental influences in the devel-
screened by a telephone interview. opment of OCD and the relationship between paren-
The specific influence of child-rearing patterns on tal child-rearing traits and OCD subtypes are still
the development of different obsessive – compulsive controversial topics. Previous studies employed cat-
symptoms has also been proposed with inconclusive egorically defined and mutually exclusive OCD
results. Rachman and Hodgson (1980), who reported subgroups, and only differences between washers
that parents of OCD patients are frequently described and checkers, the most frequent OCD subtypes, were
as overprotecting, overcontrolling and overcritical by examined. To our knowledge, no previous studies
their children, maintained that a distinction could be have addressed the influence of perceived parental
established between washers and checkers on this characteristics in the development of other frequent
topic. According to these authors, a different fear obsessive –compulsive symptoms such as hoarding,
structure, related to upbringing styles, would underlie sexual/religious themes, symmetry or ordering. Since
the most common forms of ritualistic behavior in a possible influence of current mood state on the
OCD: washing behavior would emerge from over- perception of parental rearing style has been postu-
protective and overcontrolling families that produce lated and parental rejection has been described as
fearful dependent children, while checking behavior being associated with depression, we decided to
would be related to overcritical and rejecting parents study whether parental rearing patterns were related
270 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

to the severity of depressive symptoms in our Interview for DSM-III-R: Non-Patient Version (SCID-
sample. The three-fold purpose of the present study NP) (Spitzer et al., 1989) and the guidelines estab-
was to examine whether (1) there are differences in lished by Shtasel et al. (1991) to exclude psychiatric
perceived parental child-rearing patterns between disorders.
OCD patients and healthy controls, (2) any relation-
ship exists between perceived parental characteristics 2.2. Clinical assessment
and previously identified OCD symptom dimensions,
and (3) perceived parental traits are related to the Information was obtained on both sociodemo-
presence of depressive symptoms in OCD. graphic—age, sex, years of education, years living
at parents’ home and socioeconomic level following
the Hollingshead and Redlich (1958) classification—
2. Methods and clinical variables (age at onset of OCD defined as
age when symptoms became a significant source of
2.1. Subjects distress and interfered with the patient’s social func-
tioning). The severity of OCD was assessed using a
Study participants were 40 outpatients consecu- clinician-administered version of the Yale – Brown
tively admitted to the OCD Clinic of Bellvitge Obsessive –Compulsive Scale (Y – BOCS) (Goodman
University Hospital (Barcelona, Spain) between et al., 1989), which establishes the following severity
1997 and 1999. All patients met DSM-IV criteria levels: subclinical (scores of 0 – 7), mild (8 – 15),
for OCD (American Psychiatric Association, 1994). moderate (16 – 23), severe (24 – 31) and extreme
Diagnosis was independently assigned by two psy- (32 –40). A clinician-administered version of the 21-
chiatrists with extensive clinical experience in OCD, item Hamilton Depression Rating Scale (HDRS)
who separately interviewed the patients using the (Hamilton, 1960) was used to assess the severity of
Structured Clinical Interview for DSM-IV Axis I depressive symptoms (scores of 0 – 63).
Disorders-Clinician Version (SCID-CV) (First et al., The nature of OCD symptoms was ascertained
1997). Patients were eligible when both research via a clinician-administered version of the Y –BOCS
examiners agreed on all criteria. All patients gave Symptom Checklist (Goodman et al., 1989). This is
written informed consent after complete description a comprehensive list of more than 50 examples of
of the study. Exclusion criteria were the presence of obsessions and compulsions that can be grouped into
any other comorbid axis I disorder and/or any 13 major categories. Despite some differences, recent
neurological disorder. During the selection period, factor-analytic studies have been fairly consistent in
67 outpatients of those referred for examination at reducing the symptoms of OCD into a few clinically
the Department of Psychiatry of our hospital fulfilled meaningful dimensions (Baer, 1994; Leckman et al.,
DSM-IV criteria for OCD and were evaluated by the 1997; Mataix-Cols et al., 1999; Summerfeldt et al.,
examiners. Of these patients, 27 were ruled out in 1999) that at least in adult patients, tend to remain
accordance with the exclusion criteria: 18 (26.8%) stable over time (Mataix-Cols et al., 2002b). These
because of concomitant major depression or dysthy- dimensions are the following: (1) symmetry obses-
mia, seven (10.4%) because of comorbid anxiety sions and repeating, counting and ordering compul-
disorders other than OCD and two (2.9%) because sions; (2) hoarding obsessions and compulsions; (3)
of fulfilling criteria for eating disorders. contamination obsessions and cleaning compulsions;
Forty normal comparison subjects, recruited from (4) aggressive obsessions and checking; and (5)
residents of the local community, were matched with sexual/religious obsessions. Following the methodol-
patients for gender, age, years of education and ogy of previous studies (Baer, 1994; Mataix-Cols et
socioeconomic status. They were asked to participate al., 1999), for each of these categories, if a patient
in a study on psychological health with no payment identified at least one of the specific symptoms
offered. They had no past or current history of under that category as a principal or major problem,
psychiatric or neurological diagnoses as determined that category was assigned a score of 2. If a patient
in a brief interview based on the Structured Clinical endorsed at least one of the specific symptoms but
 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278 271

did not consider it to be a major problem, that independently (Arrindell and Van der Ende, 1984).
category was assigned a score of 1. Finally, a score Nevertheless, because it was not possible to reach a
of 0 was assigned if a patient did not endorse any of sufficient level of cross-national constancy in the
the symptoms under that category. In this study, the favoring subject dimension, only the other three fac-
patients’ scores on the five symptom dimensions tors were considered in the translated versions. The
identified in a previous study (Mataix-Cols et al., EMBU has been widely used in studies of rearing style
1999), namely ‘Symmetry/ordering’, ‘Hoarding’, and different psychopathological conditions in several
‘Contamination/cleaning’, ‘Aggression/checking’ countries, and its psychometric properties have been
and ‘Sexual/Religious obsessions’, were computed found to be adequate (Anasagasti and Denia, 1988;
by summing the scores of the symptom categories Benjaminsen et al., 1990; Khalil and Stark, 1992).
under each dimension and then used in all subse- Although results on the EMBU dimensions are
quent analyses. usually independently determined for the father and
the mother, we decided to consider a global parental
2.3. Measurement of parental rearing-style score (defined as the sum of father’s and mother’s
scores) to assess not only the rearing style associated
The EMBU (Egna Minnen av Barndoms Uppfos- with one parent or another, but the influence of
tran or Own Memories of Parental Rearing Experi- parental rearing practices jointly.
ences in Childhood) was used to assess the study
participants’ memories about their parents’ rearing 2.4. Statistical analysis
practices. The EMBU is an 81-item self-report mea-
sure of an individual’s perception of his or her parent’s Differences between the OCD and control groups
rearing style during childhood (no specific reference is in demographic and clinical variables—including the
made to any time frame for which subjects are three subscales of the EMBU for both parents jointly
requested to remember their parent’s attitudes). All and each parent separately—were investigated with
items are separately scored for the father and the one-way analyses of variance (ANOVAs) for contin-
mother on 4-point scales ranging from 1 (‘no, never’) uous variables and chi-square tests for categorical
to 4 (‘yes, most of the time’). The EMBU, initially variables. Mann – Whitney U tests were used for
developed in Sweden by Perris et al. (1980), has been continuous variables when the Levene test for homo-
adapted for use in over 25 countries and validated for geneity of variances was significant. To control for
different national contexts (Arrindell et al., 1986). The error derived from multiple comparisons, the Bonfer-
validated Spanish version of the EMBU was roni correction was employed (significance level was
employed in the present study (Arrindell et al., 1988). established at 0.016 when comparing both groups on
The EMBU consists of 14 subscales, each of the three main dependent variables, i.e. the subscales
which contains items that give an indication of the of the EMBU).
degree to which each parent was described as abusive, Multiple linear regression analyses (stepwise meth-
depriving, punitive, shaming, rejecting, overprotec- od) were conducted to assess whether certain parental
tive, overinvolved, tolerant, affectionate, performance child-rearing patterns predicted the presence of spe-
oriented, guilt engendering, stimulating, favoring sib- cific obsessive –compulsive symptom dimensions. In
lings and favoring the subject. In addition to these a these models, the patients’ scores on each of the
priori dimensions, the EMBU contains two general subscales of the EMBU were entered as independent
questions that are also separately scored for the father variables and the scores on the five previously iden-
and the mother: one concerned with the degree of tified OCD symptom dimensions (Mataix-Cols et al.,
consistency in parental rearing behavior and the other 1999) as dependent variables. To control for the effect
with the degree of strictness of parental rearing style. of symptom severity and depression, all analyses were
Factor analysis of the original EMBU version identi- repeated entering the total Y – BOCS and HDRS
fied the following four primary dimensions: rejection, scores first in the models (enter method).
emotional warmth, overprotection and favoring Correlations between scores on perceived parental
subject, which are determined for both parents child-rearing patterns and clinical variables such as
272 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

age at onset of OCD, severity of OCD and presence of Patients’ scores on the HDRS were in the mild to
depressive symptoms were examined with Pearson moderate range. No patient met DSM-IV criteria for
correlation coefficients. In order to further examine major depression, since the presence of a comorbid
the possible relationship between perceived parental axis I disorder was an exclusion criterion. Frequencies
child-rearing patterns and the presence of depressive of the major symptom dimensions of the Y –BOCS
symptoms in OCD, patients were classified into the Symptom Checklist are listed in Table 2.
following two groups according to their scores on the As shown in Table 3, OCD patients perceived their
HDRS: those with at least moderate depression fathers as more rejecting (U = 458.5, Z = 3.0, P =
(HDRS z 17; n = 14) and those with sub-clinical or 0.003) than controls. Lower paternal emotional warmth
mild depression (HDRS < 17; n = 26). This dichoto- was also described by OCD patients, but differences
mous variable was then used as the dependent vari- did not reach statistical significance since the use of the
able in a logistic regression analysis (stepwise Bonferroni correction raised the level of significance to
method) where patients’ scores on each of the sub- 0.016. When considered jointly, OCD patients per-
scales of the EMBU were entered as independent ceived their parents as less emotionally warm than
variables. controls did, although this difference was not statisti-
The significance level was set at 0.05, and all cally significant after application of the Bonferroni
analyses were conducted using the SPSS statistical correction. No difference between the groups could
package (version 10.0). be detected regarding parental overprotection.
In the OCD group, multiple linear regression
analyses revealed strong negative partial correlations
3. Results between scores on the hoarding dimension and per-
ceived parental emotional warmth (R 2 = 0.24,
The demographic and clinical variables of the beta = 0.49, t = 3.46, P = 0.001). These results
patient and control groups are shown in Table 1. remained unchanged when total Y – BOCS and HDRS
There were no differences between the two groups scores were forced first into the models. None of the
with respect to sex, age, years of education, years other OCD symptom dimensions were significantly
living in the parental home or socioeconomic level. related to perceived parental child-rearing patterns.
Patients’ scores on the Y –BOCS suggested the No significant correlations were observed between
presence of severe OCD symptoms in our sample. scores on perceived parental traits and age at onset of

Table 1
Demographic and clinical characteristics of OCD patients and healthy comparison subjects
OCD (n = 40) Controls (n = 40)
Variable N % N % v2 d.f. P
Sex, male 20 50.0 20 50.0 0.00 1 1.00
Socioeconomic level 0.00 2 1.00
Low-medium 12 30.0 12 30.0
Medium 18 45.0 18 45.0
Medium-high 10 25.0 10 25.0

Mean S.D. Range Mean S.D. Range F d.f. P

Age, years 29.2 9.6 17 – 55 31.0 7.7 19 – 53 0.77 1.78 0.38
Education, years 10.9 2.7 11.0 2.8 0.02 1.78 0.87
Living at parents’ home, years 25.4 6.4 25.2 5.2 0.02 1.78 0.87
Illness onset, years 16.8 6.2 6 – 38
Y – BOCS, total 27.0 7.4
Y – BOCS, obsessions 13.6 4.1
Y – BOCS, compulsions 13.3 4.4
HDRS 14.0 5.4
 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278 273

Table 2 ing symptoms was the only clinical dimension that

Frequencies of the major symptom dimensions of the Yale – Brown could be partially predicted by perceived parental
Obsessive – Compulsive Checklist in a group of 40 OCD patients
traits. Specifically, patients with high scores on the
Absent Present Major
hoarding dimension perceived their parents as being
symptom symptom symptom
less emotionally warm than patients with other symp-
N % N % N %
toms. The severity of depressive symptoms in OCD
Hoarding 30 75.0 8 20.0 2 5.0 patients could not be predicted by scores on any
Aggressive/checking 10 25.0 15 37.5 15 37.5
perceived parental characteristics.
Contamination/cleaning 20 50.0 11 27.5 9 22.5
Sexual/religious 26 65.0 8 20.0 6 15.0 Previous research on the contribution of parental
Symmetry/ordering 20 50.0 12 30.0 8 20.0 rearing practices to the development of OCD had
yielded mixed results. While some authors described
high levels of parental overprotection in OCD patients
OCD, OCD severity or presence of depressive symp- (Hafner, 1988; Merkel et al., 1993; Turgeon et al.,
toms (Table 4). Logistic regression analysis showed 2002), others reported more rejection and less caring
that the severity of depressive symptomatology could than in normal controls (Hoekstra et al., 1989) or no
not be predicted by perceived parental child-rearing significant differences between patients and healthy
patterns. subjects (Vogel et al., 1997). Our results support
previous reports on the existence of differences in
perceived parental styles of upbringing in OCD,
4. Discussion mainly concerning rejection and emotional care, but
do not support the previously reported relationship
Patients with OCD perceived higher levels of between parental overprotection and OCD. The influ-
rejection from their fathers than healthy controls. ence of parental overprotection in the development of
Lower levels of emotional warmth from their fathers anxiety disorders or depression is still a controversial
and both parents considered jointly were also de- topic. An affectionless, controling rearing style (low
scribed by OCD sufferers, although these differences parental affection and high parental control) has been
did not reach statistical significance. No significant reported to be associated with different anxiety dis-
difference regarding parental overprotection was orders and depression (Gerlsma and Emmelkamp,
detected between the groups. The presence of hoard- 1990). Some studies have suggested that a specific

Table 3
Parental rearing style assessed by the EMBU in OCD patients and healthy controls
OCD (n = 40) Controls (n = 40)
Variable Mean S.D. Mean S.D. F/U d.f./Z P
Father
Rejectiona 39.8 15.4 31.0 7.6 458.5 3.0 0.003
Emotional warmth 43.1 11.8 49.2 10.6 5.82 1.78 0.019b
Overprotectiona 32.6 10.2 30.0 5.5 711.5 0.4 0.62

Mother
Rejectiona 38.2 12.2 33.7 6.7 685.0 1.1 0.26
Emotional warmth 49.8 11.7 53.5 9.0 2.40 1.78 0.12
Overprotectiona 34.1 9.6 34.3 5.5 732.0 0.6 0.51

Parents
Rejectiona 78.0 26.9 65.1 11.1 580.0 1.8 0.07
Emotional warmth 92.9 21.9 102.1 18.4 3.99 1.78 0.04b
Overprotectiona 66.7 19.4 64.8 8.7 713.0 0.47 0.63
a
Mann Whitney U-tests were used when variances were not homogeneous.
b
Application of the Bonferroni correction raised the significance level to 0.016.
274 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

Table 4
Correlations between scores on perceived parental child-rearing patterns assessed by the EMBU and clinical variables (age at onset of OCD,
severity of OCD and depression) in a group of 40 OCD patients
Age at onset of OCD Y – BOCS HDRS
r P r P r P
Father
Rejection 0.18 0.26 0.11 0.48 0.03 0.81
Emotional warmth 0.07 0.66 0.28 0.07 0.25 0.10
Overprotection 0.01 0.9 0.01 0.92 0.09 0.57

Mother
Rejection 0.24 0.12 0.01 0.94 0.07 0.66
Emotional warmth 0.19 0.23 0.18 0.24 0.25 0.11
Overprotection 0.04 0.78 0.001 0.99 0.12 0.45

Parents
Rejection 0.21 0.18 0.07 0.66 0.01 0.94
Emotional warmth 0.06 0.69 0.25 0.11 0.27 0.08
Overprotection 0.03 0.84 0.008 0.96 0.10 0.50

relationship may exist between particular child-rear- social disability compared with non-hoarding OCD
ing patterns and specific emotional traits: parental and other anxiety disorders (Frost et al., 2000; Mataix-
control may be more closely related to anxiety where- Cols et al., 2000). Furthermore, these patients are less
as parental rejection may be more specifically related likely to be married (Frost and Gross, 1993) and a
to depression. Other authors (Parker, 1979) have substantial number of treatment-seeking hoarders are
suggested that the interaction between these two socially phobic (Steketee et al., 2000). Hoarding OCD
child-rearing factors may be more important in the has also been associated with poorer treatment re-
development of anxiety or depression than the sepa- sponse to serotonergic agents and cognitive-behavior-
rate influence of each one of them. Further studies are al therapy (Black et al., 1998; Mataix-Cols et al.,
needed to clarify whether rejection and overprotection 2002a). Alsobrook et al. (1999) have recently reported
play a distinct role in the origins of anxiety and that a significant greater genetic component can be
depression. established in OCD patients with symmetry/ordering
Regarding the relationship between parental child- symptoms. The results of the current study may
rearing patterns and the development of specific OCD suggest that social variables such as parental child-
symptoms, our results do not support previous hy- rearing style could especially contribute to the devel-
pothesized differences between washers and checkers opment of other OCD symptoms such as hoarding
(Rachman and Hodgson, 1980). We found no signif- obsessions and compulsions.
icant associations between contamination/cleaning or As previously discussed, a relationship between
aggressive/checking dimensions and any of the three parental child-rearing style and the development of
parental rearing style factors defined by the EMBU. disorders other than OCD has also been proposed.
The presence of hoarding obsessions or compulsions Perceived parental overprotection and rejection have
was the only OCD symptom dimension that could be been linked to the development of agoraphobia (De
significantly predicted by perceived parental traits, Rutter, 1994), and Parker et al. (1987) reported that
specifically by low parental emotional warmth. neurotic depressives perceived their parents as less
The presence of hoarding obsessions and compul- caring and more protective than melancholic depres-
sions in OCD appears to be related to some particular sives and healthy subjects did, and that this ‘affec-
clinical characteristics. Hoarding OCD is associated tionless control’ style of upbringing was highly
with higher levels of comorbidity (i.e. anxiety, de- discriminating for neurotic depression. We found no
pression, personality disorders), as well as work and significant relationship between perceived parental
 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278 275

characteristics and the presence of depressive symp- reported a close association between hoarding symp-
toms in OCD patients. Nevertheless, our negative toms and comorbid axis II diagnoses, especially from
results may be explained by the fact that patients in the anxious– fearful cluster (Frost et al., 2000; Mataix-
this study were ‘pure’ obsessive – compulsive sub- Cols et al., 2000). Therefore, the association between
jects; none of them satisfied DSM-IV criteria for perceived parental emotional warmth and hoarding
major depressive disorder or dysthymia, and they only detected in our study may be confounded by the
showed mild or moderate depressive symptoms sec- presence of abnormal personality traits. Future studies
ondary to the distress caused by their OCD symptoms. would benefit from the assessment of personality
From a clinical perspective, these results support disorders to control for their effect.
the importance of involving family members in the This study was only based on retrospective reports
treatment of obsessive – compulsive children. OCD of parental rearing style, so a memory bias cannot be
often produces severe stress on family members of disregarded. Direct observational studies of parent –
affected children, because of their involvement in the child interactions or studies combining data from
patient’s compulsions or avoidance behaviors as well offspring, parents and siblings on perceived child-
as because of modification of family and leisure time rearing practices may increase validity of the results in
routines to accommodate the patient. Parents of OCD this area, although each source of information has its
patients often feel confused and anxious when faced own limitations and biases. As specific mood-congru-
with their children’s obsessive – compulsive behav- ent memory biases associated with depression have
iors, and their responses to OCD symptoms are been described, the presence of depressive symptoms
frequently inconsistent or erratic. Rigid, demanding and overall illness severity were taken into account in
and highly critical families generate feelings of guilt, the statistical analyses, and results were not confound-
increase anxiety in affected children, and discourage ed by correlation with either of the two factors. On the
them from engaging in active treatment for OCD. other hand, previous findings with the EMBU suggest
Thus, altering family communication style and learn- that evaluating retrospective data does not threaten the
ing alternative responses to the patient’s OCD symp- reliability and validity of the information obtained
toms may be an important issue especially to facilitate since it can be interpreted as a measure of the
gains of cognitive-behavioral treatment (Steketee and phenomenological impact of parental behaviors
Van Noppen, 1998). (Arrindell et al., 1983).
Several limitations of the present study need to be Although normal comparison subjects were care-
addressed. Although the dimensional approach adop- fully selected and screened to rule out any past or
ted in this study has the potential advantage of current history of psychiatric or neurological disorder,
overcoming the difficulty of recruiting a sufficient the presence of subthreshold obsessive –compulsive
sample size of each OCD clinical subtype, some symptoms was not assessed in the control group.
symptom dimensions (i.e. sexual/religious, symme- Several studies have reported that a high percentage
try/ordering, hoarding) were present in a reduced of the normal population have some obsessions and
proportion of patients. So, the sample size might have compulsions, and it has been postulated that obses-
been insufficient to detect a significant relationship sive –compulsive phenomena form a continuum with
between some of these clinical dimensions and paren- few symptoms and minimal severity at one end, and
tal rearing factors. The results of the current study many symptoms and severe impairment at the other
need to be replicated in larger samples to address this (Rachman and DeSilva, 1978). Therefore, some of our
issue as well as to confirm the stability of the detected healthy control subjects may exhibit subclinical ob-
association between hoarding and perceived parental sessive – compulsive symptoms, which could influ-
emotional warmth. ence the results of the study.
On the other hand, the presence of personality Finally, the presence of obsessive – compulsive
disorders was not specifically assessed in our study. traits or any other psychiatric conditions in the parents
Comorbid personality disorders have been reported to of the OCD group was not studied. Clinical and sub-
be present in approximately 50% of OCD patients clinical obsessional features as well as other anxious
(Baer et al., 1990). Moreover, two recent studies have and affective disorders have been reported in parents
276 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

of OCD patients (Rasmussen and Tsuang, 1986), and (1999FI-00726). DM-C was funded by a Marie Curie
they may play an important role in parental rearing grant from the EU.
practices. Future studies should take into account this
factor, since the study of the relationship between
parental psychopathology and child-rearing patterns References
would increase our understanding of family influences
on the development of OCD. Alsobrook, J.P., Leckman, J.F., Goodman, W.K., Rasmussen, S.A.,
In conclusion, OCD patients perceive their fathers Pauls, D.L., 1999. Segregation analysis of obsessive – compul-
as being more rejecting than control subjects. Patients sive disorder using symptom-based factor scores. American
Journal of Medical Genetics (Neuropsychiatric Genetics) 88,
with hoarding symptoms perceive their parents as less
669 – 675.
emotionally warm than patients without these symp- American Psychiatric Association. 1994. Diagnostic and Statistical
toms. Whether this reflects actual rearing practices or Manual of Mental Disorders, 4th ed. Author, Washington, DC.
the patients’ biased perception needs to be further Anasagasti, J.I., Denia, M., 1988. Opiate addicts and their perceived
investigated using more objective measures. If repli- parental rearing. Acta Psychiatrica Scandinavica 78 (Suppl.
344), 121 – 126.
cated, the current findings would suggest that parental
Arrindell, W.A., Emmelkamp, P.M.G., Brilman, E., Monsma, A.,
style of upbringing, as well as other social variables, 1983. Psychometric evaluation of an inventory for assessment of
may interact with genetic and biological factors to parental rearing practices. Acta Psychiatrica Scandinavica 67,
shape the OCD phenotype in vulnerable subjects. 163 – 177.
Recent published articles have focused on the rela- Arrindell, W.A., Perris, C., Eisemann, M., Perris, H., Van der Ende,
J., Ross, M., Benjaminsen, S., Gaszner, P., del Vecchio, M.,
tionship between functional and anatomical variability
1986. Cross-national generalizability of patterns of parental
of different brain areas and behavioral styles (Sugiura rearing behavior: invariance of EMBU dimensional representa-
et al., 2000; Pujol et al., 2002). Parental child-rearing tions of healthy subjects from Australia, Denmark, Hungary,
practices may contribute to the development of dis- Italy and The Netherlands. Personality and Individual Differ-
torted beliefs about responsibility, threat estimation, ences 7, 103 – 112.
Arrindell, W.A., Perris, H., Denia, M., Van der Ende, J., Perris, C.,
perfectionism, control or tolerance for ambiguity,
Kokkevi, A., Anasagasti, J.I., Eisemann, M., 1988. The constan-
which are frequently exhibited by patients with cy of structure of perceived parental rearing style in Greek and
OCD (Steketee et al., 1998). Parental behaviors and Spanish subjects as compared with the Dutch. International
attitudes may also be related to the development of Journal of Psychology 23, 3 – 23.
patterns of temperament and character described in Arrindell, W.A., Van der Ende, J., 1984. Replicability and invari-
ance of dimensions of parental rearing behavior: further Dutch
OCD (high harm avoidance, low novelty-seeking and
experiences with the EMBU. Personality and Individual Differ-
cooperativeness) (Lyoo et al., 2001). Studies on the ences 5, 671 – 682.
relationship between parental bonding styles and Baer, L., 1994. Factor analysis of symptom subtypes of obsessive
cognitive domains or personality characteristics in compulsive disorder and their relation to personality and tic
OCD patients, and between these aspects and the disorders. Journal of Clinical Psychiatry 55 (Suppl. 3), 18 – 23.
Baer, L., Jenike, M.A., Ricciardi, J.N., Holland, A.D., Seymour,
morphology or function of different brain regions,
R.J., Minichiello, W.E., Buttolph, M.L., 1990. Standardized
may constitute a way to analyze the interaction of assessment of personality disorders in obsessive – compulsive
social/cultural variables and biological factors in disorder. Archives of General Psychiatry 47, 826 – 830.
OCD. Further research on this interaction could help Benjaminsen, S., Krarup, G., Lauritsen, R., 1990. Personality,
us to understand the complex etiology and heteroge- parental rearing behavior and parental loss in attempted sui-
cide: a comparative study. Acta Psychiatrica Scandinavica 82,
neity of OCD.
389 – 397.
Black, D.W., Monahan, P., Gable, J., Blum, N., Clancy, G., Baker,
P., 1998. Hoarding and treatment response in 38 non-depressed
Acknowledgements subjects with obsessive – compulsive disorder. Journal of Clini-
cal Psychiatry 59, 420 – 425.
Cavedo, L.C., Parker, G., 1994. Parental Bonding Instrument: ex-
This study was supported in part by grant FIS 99/
ploring links between scores and obsessionality. Social Psychi-
1260 from the Spanish Ministerio de Sanidad y atry and Psychiatric Epidemiology 29, 78 – 82.
Consumo and grant 010210 from Fundació La Marató De Rutter, C., 1994. Anxious attachment in agoraphobia and ob-
TV3. PA was funded by the Generalitat of Catalonia sessive – compulsive disorder: a literature review and treatment
 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278 277

implications. In: Perris, C., Arrindell, W.A., Eisemann, M. outcome with serotonin reuptake inhibitors and placebo in the
(Eds.), Parenting and Psychopathology. John Wiley and Sons, treatment of obsessive – compulsive disorder. American Journal
Chichester, UK, pp. 281 – 301. of Psychiatry 156, 1409 – 1416.
Ehiobuche, I., 1988. Obsessive – compulsive neurosis in relation to Mataix-Cols, D., Baer, L., Rauch, S.L., Jenike, M.A., 2000. Rela-
parental child-rearing patterns among Greek, Italian and Anglo- tion of factor-analyzed symptom dimensions of obsessive – com-
Australian subjects. Acta Psychiatrica Scandinavica 78 (Suppl pulsive disorder to personality disorders. Acta Psychiatrica
344), 115 – 120. Scandinavica 102, 199 – 202.
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1997. Mataix-Cols, D., Marks, I.M., Greist, J.H., Kobak, K.A., Baer, L.,
Structured Clinical Interview for DSM-IV Axis I Disorders-Cli- 2002a. Obsessive – compulsive symptom dimensions as predic-
nician Version (SCID-CV). American Psychiatric Press, Inc, tors of compliance with and response to behaviour therapy:
Washington, DC. results from a controlled trial. Psychotherapy and Psychosomat-
Frost, R.O., Gross, R.C., 1993. The hoarding of possessions. Be- ics 71, 255 – 262.
haviour Research and Therapy 31, 367 – 381. Mataix-Cols, D., Rauch, S.L., Baer, L., Eisen, J.L., Shera, D.M.,
Frost, R.O., Steketee, G., Williams, L.F., Warren, R., 2000. Mood, Goodman, W.K., Rasmussen, S.A., Jenike, M.A., 2002b. Symp-
personality disorder symptoms and disability in obsessive – tom stability in adult obsessive-compulsive disorder: data from a
compulsive hoarders: a comparison with clinical and non-clin- naturalistic two-year follow-up study. American Journal of Psy-
ical controls. Behaviour Research and Therapy 38, 1071 – 1081. chiatry 159, 263 – 268.
Gerlsma, C., Emmelkamp, P.M.G., 1990. Anxiety, depression and Merkel, W.T., Pollard, C.A., Wiener, R.L., Staebler, C.R., 1993.
perception of early parenting: a meta-analysis. Clinical Psychol- Perceived parental characteristics of patients with obsessive –
ogy Review 10, 251 – 277. compulsive disorder, depression and panic disorder. Child Psy-
Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., Del- chiatry and Human Development 24, 49 – 57.
gado, P., Heninger, G.R., Charney, D.S., 1989. The Yale – Parker, G., 1979. Reported parental characteristics of agoraphobics
Brown Obsessive – Compulsive Scale (Y – BOCS). Part I: devel- and social phobics. British Journal of Psychiatry 195, 555 – 560.
opment, use and reliability. Archives of General Psychiatry 46, Parker, G., Kiloh, L., Hayward, L., 1987. Parental representations
1006 – 1011. of neurotic and endogenous depressives. Journal of Affective
Hafner, R.J., 1988. Obsessive – compulsive disorder: a question- Disorders 13, 75 – 82.
naire survey of a self-help group. International Journal of Social Perris, C., Jacobsson, L., Lindström, H., von Knorring, L., Perris,
Psychiatry 34, 310 – 315. H., 1980. Development of a new inventory for assessing mem-
Hamilton, M., 1960. A rating scale for depression. Journal of Neu- ories of parental rearing behavior. Acta Psychiatrica Scandinav-
rology, Neurosurgery and Psychiatry 23, 56 – 61. ica 61, 265 – 274.
Hoekstra, R.J., Visser, S., Emmelkamp, P.M.G., 1989. A social Pujol, J., Lopez, A., Deus, J., Cardoner, N., Vallejo, J., Capdevila,
learning formulation of the etiology of obsessive – compulsive A., Paus, A., 2002. Anatomical variability of the anterior cin-
disorders. In: Emmelkamp, P.M.G., et al. (Eds.), Fresh Perspec- gulate gyrus and basic dimensions of human personality. Neuro-
tives on Anxiety Disorders. Swets and Zeitlinger, Amsterdam, image 15, 847 – 855.
pp. 115 – 123 (Annual Series of European Research in Rachman, S., DeSilva, P., 1978. Abnormal and normal obsessions.
Behavior Therapy, vol 4). Behaviour Research and Therapy 16, 233 – 248.
Hollingshead, A.B., Redlich, F.C., 1958. Social Class and Mental Rachman, S., Hodgson, R., 1980. Obsessions and Compulsions.
Illness. Wiley, New York. Prentice Hall Inc, Englewood Cliffs, NJ.
Hoover, C., Insel, T., 1984. Families of origin in obsessive – com- Rapee, R.M., 1997. Potential role of childrearing practices in the
pulsive disorder. Journal of Nervous and Mental Disease 172, development of anxiety and depression. Clinical Psychology
207 – 215. Review 17, 47 – 67.
Khalil, N., Stark, P.M., 1992. Do perceived parental rearing patterns Rasmussen, S., Tsuang, M., 1984. The epidemiology of obses-
influence social behavior dimensions and disease severity in sive – compulsive disorder. Journal of Clinical Psychiatry 45,
schizophrenia? Acta Psychiatrica Scandinavica 86, 146 – 152. 450 – 457.
Kimidis, S., Minas, I.H., Ata, A.W., Stuart, G.W., 1992. Construct Rasmussen, S., Tsuang, M., 1986. Clinical characteristics and fam-
validation in adolescents of the brief current form of the Parental ily history in DSM III obsessive – compulsive disorder. Ameri-
Bonding Instrument. Comprehensive Psychiatry 33, 378 – 383. can Journal of Psychiatry 143, 317 – 322.
Leckman, J.F., Grice, D.E., Boardman, J., Zhang, H., Vitale, A., Shtasel, D.L., Gur, R.E., Mozley, P.D., Richards, J., Taleff, M.M.,
Bondi, C., Alsobrook, J., Peterson, B.S., Cohen, D.J., Rasmus- Heimberg, C., Gallacher, F., Gur, R.C., 1991. Volunteers for
sen, S.A., Goodman, W.K., McDougle, C.J., Pauls, D.L., 1997. biomedical research. Recruitment and screening of normal con-
Symptoms of obsessive compulsive disorder. American Journal trols. Archives of General Psychiatry 48, 1022 – 1025.
of Psychiatry 154, 911 – 917. Spitzer, R.L., Williams, J.B.W., Gibbon, M., 1989. Structured Clin-
Lyoo, I.K., Lee, D.W., Kim, Y.S., Kong, S.W., Kwon, J.S., 2001. ical Interview for DSM-III-R: Non-Patient Version (SCID-NP).
Patterns of temperament and character in obsessive – compulsive New York State Psychiatric Institute, New York.
disorder. Journal of Clinical Psychiatry 62, 637 – 641. Steketee, G.S., Grayson, J.B., Foa, E.B., 1985. Obsessive – compul-
Mataix-Cols, D., Rauch, S.L., Manzo, P.A., Jenike, M.A., Baer, L., sive disorder: differences between washers and checkers. Be-
1999. Use of factor-analyzed symptom dimensions to predict haviour Research and Therapy 23, 197 – 201.
278 P. Alonso et al. / Psychiatry Research 127 (2004) 267–278

Steketee, G.S., Frost, R.O., Rheaume, J., Wilhelm, S., 1998. Goto, R., Sato, K., Ono, S., Schormann, T., Zilles, K., Fukuda,
Cognitive theory and treatment of obsessive – compulsive dis- H., 2000. Correlation between human personality and neural
order. In: Jenike, M.A., et al. (Eds.), Obsessive – Compulsive activity in cerebral cortex. Neuroimage 11, 541 – 546.
Disorders. Practical Management, 3rd ed. Mosby, Inc, St. Summerfeldt, L.J., Richter, M.A., Antony, M.M., Swinson, R.P.,
Louis, pp. 368 – 399. 1999. Symptom structure in obsessive – compulsive disorder: a
Steketee, G., Frost, R.O., Wincze, J., Greene, K.A.I., Douglass, H., confirmatory factor-analytic study. Behaviour Research and
2000. Group and individual treatment of compulsive hoarding: Therapy 37, 297 – 311.
a pilot study. Behavioral and Cognitive Psychotherapy 28, Turgeon, L., O’Connor, K.P., Marchand, A., Freeston, M.H., 2002.
259 – 268. Recollections of parent – child relationships in patients with ob-
Steketee, G.S., Pruyn, N.A., 1998. Families of individuals with sessive – compulsive disorder and panic disorder with agorapho-
obsessive – compulsive disorder. In: Swinson, R.P., et al. bia. Acta Psychiatrica Scandinavica 105, 310 – 316.
(Eds.), Obsessive – Compulsive Disorder. Theory, Research Turner, R.M., Steketee, G.S., Foa, E.B., 1979. Fear of criticism in
and Treatment. The Guilford Press, New York, pp. 120 – 140. washers, checkers and phobics. Behaviour Research and Ther-
Steketee, G.S., Van Noppen, B.L., 1998. Group and family treat- apy 17, 79 – 81.
ment for obsessive – compulsive disorder. In: Jenike, M.A., Vogel, P.A., Stiles, T.C., Nordahl, H.M., 1997. Recollections of
et al. (Eds.), Obsessive – Compulsive Disorders. Practical Man- parent – child relationships in OCD out-patients compared to
agement, 3rd ed. Mosby, Inc, St. Louis, pp. 443 – 468. depressed out-patients and healthy controls. Acta Psychiatrica
Sugiura, M., Kawashima, R., Nakagawa, M., Okada, K., Sato, T., Scandinavica 96, 469 – 474.