Research report

Obsessive-compulsive symptoms and personal disposition, family

coherence and school environment in Chinese adolescents:
A resilience approach
Jing Sun
a,n,1
, Zhan-Jiang Li
b,n,1
, Nicholas J. Buys
c
, Eric A. Storch
d,e,f
, Ji-sheng Wang
g
a
Grifth Health Institute and School of Medicine, Grifth University, QLD Q4222, Australia
b
Department of Clinical Psychology, Beijing Key Lab of Mental Disorders; Beijing Anding Hospital, Capital Medical University; Center of Schizophrenia;
Beijing Institute for Brain Disorders, Beijing, China 100088
c
Grifth Health Institute, Grifth University, QLD Q4222, Australia
d
Department of Pediatrics, University of South Florida, United States
e
Rogers Behavioral Health Tampa Bay, Tampa, FL, USA
f
All Children's Hospital Johns Hopkins Medicine, St. Petersburg, FL, USA
g
Institute of Psychology, Chinese Academy of Sciences, Beijing, China
a r t i c l e i n f o
Article history:
Received 10 April 2014
Received in revised form
8 July 2014
Accepted 9 July 2014
Available online 18 July 2014
Keywords:
Obsessive-compulsive disorder
Resilience
Personal disposition
Family coherence
School environment
a b s t r a c t
Background: Risk factors of adolescents with obsessive-compulsive symptoms (OC) have been
extensively examined, but protective resilience factors have not been explored, particularly in Chinese
adolescents.
Aim: This study aimed to investigate the association of resilience factors with the occurrence of OC and
its symptoms in Chinese adolescents.
Method: This study consisted of two phases. The rst phase used a cross-sectional design involving a
stratied clustered non-clinical sample of 3185 secondary school students. A clinical interview procedure
was then employed to diagnose OC in students who had a Leyton Obsessional Inventory yes score of
Z15. The second phase used a case-control study design to analyse the relationship between resilience
factors and OC in a matched sample of 288 adolescents with diagnosed OC relative to 246 healthy
adolescents.
Results: Low personal disposition scores in self-fullment, exibility and self-esteem, and low peer
relation scores in the school environment were associated with a higher probability of having OC.
Canonical correlation analysis indicated that OC symptoms were signicantly associated with personal
dispositions, poor peer relationships and maladaptive social life, but not to family coherence.
Limitations: The study is not prospective in nature, so the causal relationship between OC occurrence and
resilience factors cannot be conrmed. Second, the use of self-report instruments in personal disposition,
family coherence, and school environment may be a source of error.
Conclusions: Resilience factors at both the personal disposition and school environment levels are
important predictors of OC symptoms and caseness. Future studies using prospective designs are needed
to conrm these relationships.
& 2014 Elsevier B.V. All rights reserved.
1. Introduction
Obsessive-compulsive (OC) symptoms are characterised by recur-
rent obsessional thoughts, urges or images that are experienced as
unwanted and intrusive. Obsessions are typically accompanied by
repetitive intentional compulsive behaviours or mental acts that are
functionally linked to obsessions and serve to reduce associated
distress. OC is a relatively common disorder, with a lifetime pre-
valence of 12% across both adults and youth worldwide (American
Psychiatric Association, 2000; Horwath and Weissman, 2000).
Recently there has been an increased focus on paediatric OC and its
causal factors, given its debilitating nature, to better inform treatment
approaches. Past studies in this area have examined the impact of
childhood onset OC on psychological and social functioning, including
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/jad
Journal of Affective Disorders
http://dx.doi.org/10.1016/j.jad.2014.07.015
0165-0327/& 2014 Elsevier B.V. All rights reserved.
n
Corresponding authors.
E-mail addresses: j.sun@grifth.edu.au (J. Sun), lizhj8@ccmu.edu.cn (Z.-J. Li),
n.buys@grifth.edu.au (N.J. Buys), estorch@health.usf.edu (E.A. Storch),
wangjisheng9@126.com (J.-s. Wang).
1
Contribute equally to the paper and rst authors.
Journal of Affective Disorders 168 (2014) 459465
compulsive self-harm or suicide, anxiety, impulsiveness, feelings of
depression and helplessness, poor insight, and aggressive behaviours
(Adam et al., 2010; Aelterman et al., 2011; Geller, 2006; Geller et al.,
1996). They have also examined functional impairment and poor
quality of life due to OC (Micali et al., 2010). However, few studies
have investigated the impact of resilience factors, including personal
dispositions, family coherence and social support, on reducing the
severity of symptoms.
Resilience is a concept that may have utility in the treatment of
OC. It is broadly dened as the ability of an individual to
successfully adapt to or recover from stressful or traumatic
experiences (Masten and Coatsworth, 1998). Previous studies on
mental disorders (e.g., severe depression) using a resilience
approach have focused on protective factors to promote recovery
that include positive personal dispositions (Campbell-Sills et al.,
2006a), personality traits, optimism, self-condence or self-
efcacy (Sun and Stewart, 2007), and an active style of coping
(Campbell-Sills et al., 2006b). They have also examined environ-
mental protective factors such as family warmth, family members
understanding within the family coherence context (Baumrind,
1991), school context such as positive peer relationships and a
sense of belonging within schools (Baker et al., 2003). These
factors act as assets and resources to buffer the effects of adverse
experiences (Luthar, 2003, Rutter, 1985), help people to cope with
adversity, make them better able to deal with stress in the future,
and confer protection from the development of mental disorders
(Skodol et al., 2007).
Clinical studies specically focussed on resilience related fac-
tors have been found to be important predictors of OC sympto-
matology in children and adolescents. For example, De Clercq et al.
(2006) found children and adolescents with OC symptoms are
characterised by a spectrum of personal dispositions including
negative emotional experiences (high Neuroticism), impaired
engagement and enjoyment in participation in social activities
(low Extraversion score). This is reected in their high anxiety and
shyness scores, and low self-condence and optimism scores as
measured by the Dimensional Personality Symptom Item Pool
(De Clercq et al., 2006). Family factors including impaired family
functioning, parental blame, family conict and family cohesion
have also been found to be related to OC symptoms (Peris et al.,
2008; Piacentini et al., 2003b; Storch et al., 2007). Parent char-
acteristics, such as over protectiveness are signicantly related to
OC symptoms, while parental care is a protective factor of OC
(Wilcox et al., 2008). Although the substantial evidence indicated
the correlates of family factors and OC symptoms, how these
features relate to treatment outcome is unclear. Some studies
found a signicant link between family factors and responses to
the treatment of OC symptoms, whereas others have not found
this link (Peris et al., 2012). This suggests that OC symptoms
attributed to family factors may need to be examined in a broad or
multiple context that may impact on the family functioning (Peris
et al., 2012). School contextual factors, such as social functioning
and peer relationships are important predictor of OC (Sukhodolsky
et al., 2005). Children with OC have been found to have difculty
to making friends, keeping friends and engaging in age appro-
priate peer activities (Piacentini et al., 2003b). Further, Storch et al.
(2006) found that children with OC were at risk of problematic
peer relations because, in part, their symptoms (washing hands,
cleanliness) were not understood by peers. Cross-sectional studies
have also suggested an inverse relationship between resilience and
OC symptoms. For example, a study of non-clinical Norwegian
adolescents found higher levels of resilience were related to lower
levels of OC symptoms (Hjemdal et al., 2011).
There remains a paucity of research investigating the impact of
resilience on OC caseness and OC symptoms, particularly in
Chinese adolescents. Although clinical studies have identied
individual resilience related factors, no studies have examined
the interaction of combinations of these resilience factors in
relation to the caseness and severity of OC symptoms. The aim
of the present study was to examine the hypothesis that the
combination of personal disposition, family coherence and school
environment factors are strongly correlated with obsessive-
compulsive symptoms in Chinese adolescents.
2. Methods
2.1. Participants
There were two stages in the recruitment of the study sample. In
the rst stage, a cross-sectional study was conducted, using a
stratied clustered sample, to identify Chinese adolescents with OC
among a non-clinical population using the Leyton Obsessional
Inventory (LOI-CV) (Berg et al., 1988) as a screening tool. In the
initial screening test, four schools that had different academic
rankings, based on the quality of teaching and learning outcomes,
were selected in two districts of Beijing, China. Students were then
randomly selected from these four schools. Three classes per grade in
each school were invited to participate in the study, resulting in 3221
potential participants across 72 classes. Of these, 3174 students ages
12 to 18 years agreed to participate in the initial OC screening test,
representing a 98.5% response rate (See Table 1). Children who were
not given permission by their parents and their parents did not
submit consents were excluded from the study. In the initial screen-
ing process, 434 (13%) students had scores of 15 or more on the
validated LOI-CV in Chinese adolescents (Sun et al., 2014), which is
considered as having OC symptoms at the population level.
In the second stage of the study, students with a LOI-CV score
of more than 15 were evaluated using the Structured Clinical
Interview for DSM-IV Axis I Disorders-Patient Edition (SCID-I/P,
Version 2.0) (American Psychiatric Association, 1994), a protocol
which assesses diagnostic criteria for OC and other Diagnostic and
Statistical Manual of Mental Disorders IV (DSM-IV) disorders.
Participants were excluded if they had a documented neurological
impairment or intellectual impairment, or suicidality six months
prior to the start of the study. Ultimately, 288 students who met
criteria on both the SCID and the Maudsley Obessive-Compusive
Inventory (MOCI) were diagnosed to have OC and were recruited
to the clinical group. The control group was also recruited through
the screening process and consisted of 246 adolescents who
scored less than 15 on the LOI-CV and matched the students in
the clinical group in terms of age and gender. These participants
were also free from any psychiatric disorders.
Ethical clearance for this study was obtained through the
Chinese Academy of Sciences, Beijing, China, and informed par-
ental consent was obtained for each student after providing them
with a detailed description of the study. Agreement to conduct the
study was gained from school principals and teachers. All partici-
pants were Chinese, predominantly from a middle socioeconomic
background, and ranged in age from 12 to 18 years (M16.94,
SD1.38). All potential participants were screened in a psychiatric
clinic and diagnosed by two doctoral level clinicians using the
clinical interview procedures described above.
2.2. Measures
Those who met the above inclusion criteria for either the
clinical or control groups were administered a series of measures.
These measures include structure interview and self-report mea-
sures which were administrated in a psychiatric clinic located in
Beijing, China. The self-report measures included the California
Psychological Inventory (CPI), Maudsley Obessive-Compusive
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 460
Inventory (MOCI), Family Coherence Scale (FCS), and School
Environment Scale (SES), which were validated and presented in
simplied Chinese language format. A high level of agreement was
achieved between the two licensed clinical psychiatrists specializ-
ing in diagnosing childhood OC with intra-class correlation ran-
ging from 0.80 to 0.93 in the measures of clinical interview, CPI,
FCS, and SES. They are described below.
2.2.1. Structured interview
The diagnosis of OC was initially conducted by using the
Structured Clinical Interview for DSM-IV Axis I Disorders-Patient
Edition (SCID-I/P, Version 2.0) (American Psychiatric Association,
1994), a measure which assesses diagnostic criteria for OC and
other Diagnostic and Statistical Manual of Mental Disorders IV
(DSM-IV) disorders. The checklist questions included onset of
OC, obsessive-compulsive symptoms, duration of the symptoms,
and severity of the symptoms. Each participant was reviewed
independently by two psychiatrists, each of whom completed a
Diagnostic Assignment Checklist form (intraclass correlation
40.93 between the two raters).
2.2.2. Maudsley obessive-compusive inventory (MOCI)
MOCI was used to assess severity of OC symptoms following
the SCID procedure. The MOCI (Hodgson and Rachman, 1977) was
also used to assist with diagnosing OC symptoms following the
DSM-IV interview procedure. The MOCI is a 30-item, self-report,
symptom-oriented scale with four dimensions: compulsive
thoughts, cleanliness, checking, and doubting. The MOCI was
translated to Chinese and validated in Chinese adolescents (Xia,
1995). In the current study, the reliability as measured by Cron-
bach's alpha was 0.87, SpearmanBrown Split Half reliability was
0.83, and test and re-test in two months was 0.68 (po0.01).
Conrmatory factor analysis conrmed the validity of the scale
with goodness of t indices ranging from good to excellent levels:

2
/df 537.97/169/730, goodness-of-t index (GFI) 0.94, Com-
parative t index (CFI)0.88, GFI adjusted for degrees of freedom
(AGFI)0.88, and the root mean squared error approximation
(RMSEA)0.05.
2.2.3. California psychological inventory (CPI)
A modied CPI (Yang and Gong, 1993) was used to measure
personal dispositions as it has constructs that best capture the
individual level traits associated with resilience. Ninety items
across ve dimensions which include achievement striving,
responsibility, exibility, independence, and self-esteem were
selected to measure resilience constructs. Higher scores indicated
extreme achievement striving, low levels of responsibility, ex-
ibility, independence and self-esteem. The overall reliability of the
scale was good, with a Cronbach's alpha of 0.84 and a Spearman
Brown split half reliability of 0.85 for the overall scale, with ve
factors explaining 52.8% of the variance in the scale.
2.2.4. Family coherence scale (FES-CV)
FES-CV is a 40 item, self-report scale. Fei et al. (1993) found this
scale to have excellent reliability and validity in a Chinese
adolescent population. The scale was used to provide details of
family functioning and coherence and has six dimensions includ-
ing intimacy, culture entertainment, conict, control, coherence
Table 1
Demographic variables.
Variables Control n (%) Case n (%)
2
or t P value
Gender
Male 146 (59.3%) 163(56.6%)
2
0.41 0.54
Female 100 (40.7%) 125(43.5%)
Age M(SD) 14.7 (1.74) 14.66 (1.73) t0.27 0.98
Ethnicity
Minority 14 (5.7%) 21 (7.3%)
2
0.56 0.49
Han nationality 232 (94.3%) 267 (92.7%)
Education
Year 1 high school 44 (17.9%) 52 (18.1%)
Year 2 high school 40 (16.3%) 48 (16.7%)
Year 3 high school 44 (17.9%) 58 (20.1%)
2
0.67 0.98
Year 4 high school 50 (20.3%) 56 (19.4%)
Year 5 high school 28 (11.4%) 29 (10.1%)
Year 6 high school 40 (16.3%) 45 (15.6%)
Family monthly income
o300 RMB 5 (2.0%) 14 (5.1%)
300700 RMB 108 (44.1%) 127 (46.4%)
7011000 RMB 78 (31.8%) 69 (25.2%)
2
6.39 0.17
10011500 RMB 32 (13.1%) 43 (15.7%)
41500 RMB 22 (9.0%) 21 (7.7%)
Size of the apartment
o8 m
2
30 (12.4%) 29 (10.7%)
814 m
2
90 (37.2%) 108 (39.9%)
1520 m
2
72 (29.8%) 78 (28.8%)
2
0.88 0.92
2126 m
2
26 (10.7%) 32 (11.8%)
426 m
2
24 (9.9%) 24 (8.9%)
Number of children
Only child 186 (90.3%) 206 (90.0%)
Two children and more 20 (9.7%) 23 (10.0%)
2
0.54 0.76
Father's age M(SD) 43.99(3.89) 44.00(3.93)
n
t0.01 0.98
Father's education in years M(SD) 12.53(3.39) 12.26(3.26)
n
t0.92 0.55
Mother's age M(SD) 42.72(3.22) 42.32(3.25)
n
t1.41 0.48
Mother's education in years M(SD) 12.07(2.74) 11.70(2.85)
n
t1.48 0.28
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 461
and independence. The answers for each question were arranged
based on the occurrence of behaviours: yes, or no. Conrmatory
factor analysis conrmed the structure of the scale in the present
sample. The overall reliability of the scale was good, with a
Cronbach's alpha of 0.79 and SpearmanBrown split half reliability
of 0.79 for the overall scale, with four factors explaining 42.2% of
the variance in the scale.
2.2.5. School environment scale (SES)
The SES is a 40-item, modied scale based on the Scale of
Adaptation of School Life in Adolescents developed by Yuan and
Zhang (1996) with excellent reliability and validity in Chinese
adolescents. The scale was used to provide details of the school
adaptation and environment and has four dimensions including
academic adaptation, peer support, teacher support, and social life
adaptation. Higher scores indicated a poor level of school environ-
ment in each dimension. In the present study, the overall relia-
bility of the scale was good with a Cronbach's alpha, Spearman
Brown Split half reliability, and test-retest reliability of 0.80, 0.77,
and 0.63 respectively in the present study. Further, conrmatory
factor analysis was used to test the concurrent validity. The model
t indices showed good results with
2
/df4201.48/730,
goodness-of-t index (GFI)0.93, Comparative t index (CFI)
0.85, GFI adjusted for degrees of freedom (AGFI)0.92, and
Normed t index (NFI)0.84, and the root mean squared error
approximation (RMSEA)0.05.
2.3. Statistical analysis
STATA statistical package version 13.0 was used for this study.
Multiple logistic regression analysis was used to analyse the
association between resilience and OC in order to control con-
founding effects from other resilience factors. The predictive value
of resilience factors on OC was presented using an odds ratio
through a multiple logistic regression model. Canonical correlation
analysis was used to analyse the degree of relationship between a
group of OC variables and each of the three groups of resilience
factors: personality traits, school environment, and family
coherence.
Canonical correlation analysis was used to examine the rela-
tionships between the four sub-scales of MOCI and the three
groups of resilience factors (subscales of personal dispositions,
family coherence and school environment). This analysis produces
a subset of variables (called a canonical variate) from one set that
is maximally correlated with a subset of variables from a second
set. A second pair of canonical variates, which is uncorrelated with
the rst pair and calculated after the variance from the rst pair is
removed. OC symptoms, as measured by four MOCI subscales,
functioned as the dependent variable and personal dispositions,
family and school environment factors functioned as the indepen-
dent variables in the analysis. The contribution of each variable to
the canonical correlation was carried out using the standardised
coefcients. The square canonical correlation shows the propor-
tion of shared variance between the two groups of variables
relating to OC symptoms and resilience factors. The signicance
levels for all analyses were set at a p value of less than 0.05. A
multiple imputation method with ve imputations was used to
impute the missing values for all variables included in the analysis.
3. Results
In the OC caseness group, there were 163 (56.6%) males and
125 (43.5%) females, while there were 146 (59.3%) males and 100
(40.7%) females in the control group. Students in both groups had
ages ranging from 12 to 18, with an average age of 14 years. There
were approximately equal proportions of students across grades
1 to 6 in the secondary school in both groups. Most OC participants
(94.3%) were of Han Chinese nationality, and 90% were only
children (i.e., had no siblings). The parents of the students with OC
had an average education level of 12 years and an average age of
43 years for fathers and 42 years for mothers (Table 1). There were
no signicant differences between the case group and the control
group in age, gender, ethnicity, grade, living environment, number
of children, family income, size of living accommodations, mother
and father's age and education level, and mother's occupation.
Table 2 shows the means and standard deviation of CPI, FES-CV,
and SES scales and subscales. There are signicant differences
between case group and control group in all subscales of CPL,
conict subscale of FES-CV, and adaptation to study and peer
relationship of SES scale.
Table 3 shows the multiple logistic regression results relating to
whether personality traits, family or school environment factors
were associated with OC caseness. Personal dispositions accounted
for the largest proportion of the variance with 40.9%. The school
environment and family coherence contribution is negligible,
accounting respectively for 2.5% and 1.0% of the variance. Achieve-
ment striving was signicantly and positively related to OC (odds
ratio, 325.12 times; 95% CI, 24.84425.49, po0.01), as was ex-
ibility (odds ratio, 15.45; 95% CI, 2.4198.95, po0.03), and self-
esteem (odds ratio, 5.28, 95% CI, 2.7210.28, po0.001). None of
the family coherence factors were related to the caseness of OC. In
the School Environment Scale, peer relationships and adaption to
Table 2
Mean and standard deviation of CDI, FES-CV and SES.
Variables OC case group (n288) Control group (n246) t-value P
Personal dispositions Achievement motivation 0.53 (0.13) 0.39(0.13) 12.80 o0.001
Responsibility 0.39(0.13) 0.31(0.12) 7.19 o0.001
Flexibility 0.60(0.13) 0.56(0.15) 3.58 o0.001
Independence 0.53(0.12) 0.43(0.13) 8.77 o0.001
Self-esteem 2.01(0.52) 1.54(0.38) 12.00 o0.001
Family coherence Cohesion 0.10(0.20) 0.18(0.19) 1.19 0.24
Culture and recreation 0.48(0.24) 0.49(0.24) 0.48 0.63
Conict 0.37(0.29) 0.32(0.28) 2.13 0.03
Control 0.41(0.33) 0.40(0.32) 0.56 0.57
Coherence 0.78(0.22) 0.80(0.24) 0.83 0.41
Independence 0.27(0.20) 0.25(0.22) 0.99 0.32
School environment Adaption to study 2.22(0.50) 2.08(0.49) 3.27 o0.001
Peer relationship 1.61(0.41) 1.39(0.31) 6.72 o0.001
Relationship with teacher 2.63(0.59) 2.55(0.56) 1.56 0.12
Adaption to social life 1.83(0.62) 1.79(0.56) 0.94 0.35
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 462
social life were signicantly related to the caseness of OC. The
probability of having OC in the caseness group was 3.37 times (95%
CI, 1.336.95, po0.001) that of the control group when the poor
peer relationship scores increased by one unit. The probability of
having OC in the OC students was reduced by 38% (95% CI, 0.38
0.98, po0.05) compared to the non-OC students when the ability
to adapt to social life is increased. The results of the regression
indicated that the ve predictors explained 44.4% of the variance
(R
2
0.444,
2
159.54, po0.001, classication rate 75%).
Table 3 shows the results relating to the canonical correlation
between two groups of variables relating to OC symptoms and
personal dispositions, family coherence, and school environment.
The canonical correlation coefcient among Personality Traits and
MOCI subscales was 0.40, which represents 16.3% of the associated
variance (Wilks0.83, F5.47; df 10, 562, po0.001). The sub-
scales standardized coefcients for the canonical correlation are
shown in Table 3. In the case of MOCI, the two subscales with the
greatest coefcients were Obsession (0.67) and Compulsion (0.45).
In the case of the Personality Traits, these were the Interpersonal
(0.48) and Disorganised (0.39) subscales.
The canonical correlation coefcient among School Environment
and MOCI subscales was 0.2229, which represents 5.0% of the
associated variance (Wilks0.93, F2.79; df8, 564; po0.01).
The subscales standardized coefcients for the canonical correlation
are shown in Table 4. In the case of MOCI, the subscale with the
greatest coefcients was compulsions (0.65). In the case of the School
Environment, this was the peer relations (0.71) subscale. The
canonical correlation coefcient among MOCI subscales and Family
Coherence was 0.1356, which represents 1.8% of the associated
variance (Wilks0.97, F0.66, p0.79). In the case of MOCI, the
subscale with the greatest coefcient was obsessions (1.23,
p0.03). In the Family Coherence, none of the subscales were
statistically signicant.
4. Discussion
The present study demonstrates a signicant relationship
between resilience factors and the caseness of OC. Using a
resilience approach to explore symptoms of OC may contribute
to a better understanding of the relationship between psycho-
pathology and protective factors. Chinese youth with high resi-
lience scores on personal dispositions and reported availability of
social support in school are less likely to have OC. Lower levels of
these protective factors were positively associated with OC
symptoms.
In analysing the occurrence of OC in relation to resilience
factors the amount of total variance in OC symptoms was up to
44.4%. Personal dispositions contributed to the largest proportion
of the variance with 40.9%. The school environment and family
coherence's contribution is negligible with 2.5% and 1.0% of the
variance. These results suggest that personal dispositions such as
unrealistic achievement motivation, inexibility, and low levels of
Table 4
Caronical correlation analysis on the relationship of OC to resilience factors.
Standard coefcient t P 95% CI Wilk F R
2
OC 1st Obsession 0.67 4.00 0.00 0.71 2.07 5.47nnn 16.3%
set Compulsion 0.45 2.67 0.01 0.32 2.13
2nd Self-fullment 0.42 2.12 0.04 0.24 6.21
Personal Set Responsibility 0.09 0.52 0.61 1.99 3.40
dispositions Flexibility 0.03 0.24 0.81 1.93 2.46
Independence 0.25 1.57 0.12 0.53 4.67
Self-esteem 0.51 2.95 0.00 0.33 1.63
OC 1st Obsession 1.23 2.23 0.03 4.80 0.30 0.66 1.8%
set Compulsion 0.90 1.64 0.10 0.50 5.47
2nd Cohesion 0.25 0.51 0.61 3.54 6.00
set Environment 0.51 1.12 0.27 5.84 1.61
Family Conict 0.27 0.55 0.58 4.23 2.38
Coherence Control 0.06 0.13 0.89 2.90 2.53
Organisation 0.18 0.36 0.72 5.20 3.60
Independence 0.87 1.90 0.06 0.16 8.93
OC 1st Obsession 046 1.44 0.15 0.36 2.28 2.78nn 5.0%
set Compulsion 0.65 2.02 0.04 0.05 3.54
Academic 0.52 1.65 0.10 0.20 2.29
School 2nd Teacher 0.36 1.19 0.24 1.60 0.39
environment Set Peer relations 0.71 2.29 0.02 0.24 3.18
Social life 0.04 0.14 0.89 1.04 0.90
Statistical signicance: po0.05; Figures in bold refers to statistical signicance results.
Table 3
Relationship of personality traits, family coherence and school environment to the
occurrence of OC.
Resilience factors OR 95% CI
Personal dispositions
Achievement motivation 325.12 (24.84425.49)
nnn
Responsibility 0.78 (0.069.76)
Flexibility 15.45 (2.4198.95)
nn
Independence 1.60 (0.1814.19)
Self-esteem 5.28 (2.7210.28)
nnn
Family coherence
Cohesion 0.59 (0.154.10)
Culture and recreation 0.40 (0.181.92)
Conict 1.86 (1.289.67)
Control 0.90 (0.361.94)
Organisation 0.92 (0.466.18)
Independence 0.98 (0.315.00)
School environment
Adaption to study 0.62 (0.321.19)
Peer relationship 3.37 (1.336.95)
nnn
Teacher/student relationship 1.08 (0.671.75)
Adaption to social life 0.62 (0.380.98)
n
Statistical signicance.
Figures in bold refers to statistical signicance results.
n
Po0.05.
nn
Po0.01.
nnn
Po0.001.
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 463
self-esteem may play a major part in the occurrence of OC. Striving
for high levels of achievement may reect unrealistic expectations
among Chinese adolescents with OC, and appears to result in
negative emotions and maladaptive cognitions related to uncer-
tainty and change. This result is consistent with Aelterman et al.
(2011) who found that adolescents with high OC scores also had
higher scores in extreme achievement striving. Within the current
study low levels of exibility found in adolescents with OC is
consistent with a number of studies that found paediatric OC
patients had worse abstraction-exibility and mental set-shifting
(Ornstein et al., 2010; Taner et al., 2011; Britton et al., 2010). The
lower self-esteem scores are similar to those obtained in previous
studies of OC adolescents, which suggests that this population are
also characterised by extreme scores on several maladaptive
lower-order dimensions, including lower scores on self-
condence and optimism scores (Aelterman et al., 2011). This
result is also in accordance with the low self-esteem scores
observed in adult OC patients (Clark, 2005, 2007).
Although school environment explained a small proportion of
variances in the occurrence of OC, lack of peer relations and
inability to adapt to school social life were found to be related to
the occurrence of OC. This suggests that adolescents with OC have
decits in social functioning and interpersonal relationships. This
is consistent with previous ndings that many youth with OC
report a variety of social impairments such as difculty making
friends, difculty keeping friends, and difculty engaging in age-
appropriate activities with peers (Borda et al., 2013; Langley et al.,
2010; Piacentini et al., 2003a; Storch et al., 2006). The above
ndings suggest that measuring low levels of resilience and
maladaptive behaviour is important and may add a signicant
contribution to future studies exploring the relationship between
mental health and OC symptoms.
Canonical correlation analysis further explored the degree of
relationship between OC symptoms and resilience factors in the
clinical group. The large number of clinically diagnosed OC group
participants provided a unique opportunity to assess whether OC
symptoms were related to resilience factor levels. MOCI shared a
considerable percentage of the total variance. The subscales that
seemed to play a more predominant role when explaining the
associated variance between OC and low levels of resilience, were
the obsession and compulsion subscales of the MOCI and the
extreme achievement striving and self-esteem subscales of the
personal disposition scale. MOCI shared a considerable percentage
of the variance with school environment, in particular, the MOCI
compulsion subscale and the peer relations subscale of the school
environment scale explained signicant variance in relation to the
relationship between OC symptoms and resilience. These results
imply future prevention of OC symptoms and directions that
efforts to help OC adolescents recover, should focus on the
improvement of personal dispositions including developing
appropriate levels of achievement expectation and self-esteem,
and positive and supportive peer relations.
Surprisingly, we found no relationship between OC symptoms
and family coherence factors in either multiple logistic and
canonical correlation analysis. This nding contradicts those from
previous studies (Peris et al., 2012), and may be due to the fact that
the age of participants in the present study was much older than
in Peris et al. and that peers and school factors play an important
role in the adolescent's life. Another explanation may be that
previous studies did not put family coherence in a multiplicity of
variables context, so the confounding effects of other possible
explanatory variables such as its relationship with personal dis-
positions and school environment within the resilience framework
were not examined.
This study has several limitations. First, the study is not
prospective in nature, so the causal relationship between OC
occurrence and resilience factors, and OC symptoms and resilience
factors cannot be conrmed. Second, the use of self-report instru-
ments in personal disposition, family coherence, and school
environment may be a source of error associated with this type
of assessment. Despite the limitations of the study, the DSM-IV
clinical interview procedure in conjunction with the MOCI scale
used in diagnosing adolescents with OC provides valid data for the
study. Adding a matched control group has further validated the
results of this study. The use of robust clinical interviews in this
large sample of nonclinical adolescents as well as the use of a
control group of non-OC adolescents provides considerable
strength to the study, and underlies the importance of its ndings.
Taken together with evidence from previous research, the
primary implication of this study is that clinicians and health care
workers should recognise the important role that resilience factors
play in adolescents, as well as obsessions and compulsions.
Providing resources to increase protective factors may enhance
mental health as well as strengthening an individual's ability to
buffer adversity or risk factors. A person-centred approach is
recommended for adolescents with OC, which focuses on con-
necting OC adolescents to peers and social activities. Furthermore,
patients who have low motivation and self-esteem need to engage
in strengthening personal traits and have individual mastery
experiences. Further research is needed in this area.
5. Conclusion
The ndings of this study suggest that resilience is signicantly
related to the caseness of OC and its symptoms in a sample of
Chinese adolescents. In particular, personal disposition, such as
extreme achievement striving, inexibility, and low levels of self-
esteem, as well as school environmental factors such as poor peer
relationships and maladaptive adjustment to school life are
signicantly related to the occurrence of OC. These factors pre-
dominantly explained the variances of OC symptoms in this
population. Study results have treatment and prevention implica-
tions. Early interventions that are designed to foster personal traits
and develop interpersonal skills and a sense of belonging could
have a benecial effect on adolescents diagnosed with OC.
Furthermore, youth programs promoting personality competence,
social and emotional competencies may help prevent the devel-
opment of OC in vulnerable adolescents.
Role of funding source
The funding body did not inuence the study design, research process and
publications.
Conict of interest
Authors declare that there is no conict of interest.
Acknowledgement
The authors wish to thank Institute of Psychology, Chinese Academy of Science
to support data collection work.
References
Adam, B.L., Bergman, R.L., Tara, S.P., Susanna, C., James, T.M., John, P., 2010.
Correlates of insight among youth with obsessive-compulsive disorder. J. Child
Psychol. Psychiatry 51, 603611.
Aelterman, N., De Clercq, B., De Bolle, M., De Fruyt, F., 2011. General and
maladaptive personality dimensions in pediatric obsessive-compulsive symp-
toms. Child Psychiatry Hum. Dev. 42, 2441.
American Psychiatric Association, 2000. Diagnostic and Statistical Manual of
Mental DisordersText Revision: DSM-IV-TR. American Psychiatric Association,
Washington (DC).
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 464
Baker, J.A., Dilly, L.J., Aupperlee, J.L., Patil, S.A., 2003. The developmental context of
school satisfaction: school as psychologically healthy environment. Sch. Psy-
chol. Q. 18, 206221.
Baumrind, D., 1991. The inuence of parenting style on adolescent competence and
substance use. J. Early Adolesc. 11, 5695.
Berg, C.Z., Whitaker, A., Davies, M., Flament, M.F., Rapoport, J.L., 1988. The survey
form of the Leyton Obsessional Inventory-Child Version: norms from an
epidemiological study. J. Am. Acad. Child Adolesc. Psychiatry 27, 759763.
Borda, T., Feinstein, B.A., Neziroglu, F., Veccia, T., Perez-Rivera, R., 2013. Are children
with obsessive-compulsive disorder at risk for problematic peer relationships?
J. Obsessive-Compuls. Relat. Disord. 2, 359365.
Britton, J.C., Rauch, S.L., Rosso, I.M., Killgore, W.D., Price, L.M., Ragan, J., Chosak, A.,
Hezel, D.M., Pine, D.S., Leibenluft, E., Pauls, D.L., Jenike, M.A., Stewart, S.E., 2010.
Cognitive inexibility and frontal-cortical activation in pediatric obsessive-
compulsive disorder. J. Am. Acad. Child Adolesc. Psychiatry 49, 944953.
Campbell-Sills, L., Barlow, D.H., Brown, T.A., Hofmann, S.G., 2006a. Effects of
suppression and acceptance on emotional responses of individuals with
anxiety and mood disorders. Behav. Res. Ther. 44, 12511263.
Campbell-Sills, L., Cohan, S.L., Stein, M.B., 2006b. Relationship of resilience to
personality, coping, and psychiatric symptoms in young adults. Behav. Res.
Ther. 44, 585599.
Clark, L.A., 2005. Temperament as a unifying basis for personality and psycho-
pathology. J. Abnorm. Psychol. 114, 505521.
Clark, L.A., 2007. Assessment and diagnosis of personality disorder: perennial
issues and an emerging reconceptualization. Annu. Rev. Psychol. 58, 227257.
De Clercq, B., De Fruyt, F., Van Leeuwen, K., Mervielde, I., 2006. The structure of
maladaptive personality traits in childhood: a step toward an integrative
developmental perspective for DSM-V. J. Abnorm. Psychol. Inq. 115, 639657.
Fei, L.P., Zhen, Y.P., Zhou, D.H., 1993. Validation of family environment scale in
Chinese. Chin. J. Ment. Health, S93S101 (Supplementary).
Geller, D.A., 2006. Obsessive-compulsive and spectrum disorders in children and
adolescents. Psychiatr. Clin. North Am. 29, 353370.
Geller, D.A., Biederman, J., Grifn, S., Jones, J., Lefkowitz, T.R., 1996. Comorbidity of
juvenile obsessive-compulsive disorder with disruptive behavior disorders.
J. Am. Acad. Child Adolesc. Psychiatry 35, 16371646.
Hjemdal, O., Vogel, P.A., Solem, S., Hagen, K., Stiles, T.C., 2011. The relationship
between resilience and levels of anxiety, depression, and obsessive-compulsive
symptoms in adolescents. Clin. Psychol. Psychother. 18, 314321.
Hodgson, R.J., Rachman, S., 1977. Obsessional-compulsive complaints. Behav. Res.
Ther. 15, 389395.
Horwath, E., Weissman, M.M., 2000. The epidemiology and cross-national pre-
sentation of obsessive-compulsive disorder. Psychiatr. Clin. North Am. 23,
493507.
Langley, A.K., Lewin, A.B., Bergman, R.L., Lee, J.C., Piacentini, J., 2010. Correlates of
comorbid anxiety and externalizing disorders in childhood obsessive compul-
sive disorder. Eur. Child Adolesc. Psychiatry 19, 637645.
Luthar, S.S. (Ed.), 2003. Resilience and Vulnerability: Adaptation in the Context of
Childhood Adversities. Cambridge University Press, New York.
Masten, A.S., Coatsworth, J.D., 1998. The development of competence in favorable
and unfavorable environments: lessons from research on successful children.
Am. Psycholo. 53, 205220.
Micali, N., Heyman, I., Perez, M., Hilton, K., Nakatani, E., Turner, C., Mataix-Cols, D.,
2010. Long-term outcomes of obsessive-compulsive disorder: follow-up of 142
children and adolescents. Br. J. Psychiatry 197, 128134.
Ornstein, T.J., Arnold, P., Manassis, K., Mendlowitz, S., Schachar, R., 2010. Neurop-
sychological performance in childhood OC: a preliminary study. Depress.
Anxiety 27, 372380.
Peris, T.S., Bergman, R.L., Langley, A., Chang, S., McCracken, J.T., Piacentini, J., 2008.
Correlates of family accommodation of childhood obsessive compulsive dis-
order. J. Am. Acad. Child Adolesc. Psychiatry 47, 11731181.
Peris, T.S., Sugar, C.A., Bergman, R.L., Chang, S., Langley, A., Piacentini, J., 2012.
Family factors predict treatment outcome for pediatric obsessive-compulsive
disorder. J. Consult. Clin. Psychol. 80, 255263.
Piacentini, J., Bergman, R.L., Keller, M., McCracken, J., 2003a. Functional impairment
in children and adolescents with obsessive-compulsive disorder. J. Child
Adolesc. Psychopharmacol. 13 (Suppl. 1), S61S69.
Piacentini, J., Bergman, R.L., Keller, M., McCracken, J., 2003b. Functional impairment
in children and adolescents with obsessive-compulsive disorder. J. Child
Adolesc. Psychopharmacol. 13, s61s69.
Rutter, M., 1985. Resilience in the face of adversity: protective factors and
resistance to psychiatric disorder. Br. J. Psychiatry 147, 598611.
Skodol, A.E., Bender, D.S., Pagano, M.E., Shea, M.T., Yen, S., Sanislow, C.A., Grilo, C.M.,
Daversa, M.T., Stout, R.L., Zanarini, M.C., McGlashan, T.H., Gunderson, J.G., 2007.
Positive childhood experiences: resilience and recovery from personality
disorder in early adulthood. J. Clin. Psychiatry 68, 11021108.
Storch, E.A., Geffken, G.R., Merlo, L.J., Jacob, M.L., Murphy, T.K., Goodman, W.K.,
Larson, M.J., Fernandez, M., Grabill, K., 2007. Family accommodation in
pediatric obsessive-compulsive disorder. J. Clin. Child Adolesc. Psychol. 36,
207216.
Storch, E.A., Ledley, D.R., Lewin, A.B., Murphy, T.K., Johns, N.B., Goodman, W.K.,
Geffken, G.R., 2006. Peer victimization in children with obsessive-compulsive
disorder: relations with symptoms of psychopathology. J. Clin. Child Adolesc.
Psychol. 35, 446455.
Sukhodolsky, D.G., do Rosario-Campos, M.C., Scahill, L., Katsovich, L., Pauls, D.L.,
2005. Adaptive, emotional, and family functioning of children with obsessive-
compulsive disorder and comorbid attention decit hyperactivity disorder. Am.
J. Psychiatry 162, 11251132.
Sun, J., Boschen, M.J., Farrell, L.J., Buys, N., Li, Z.J., 2014. Obsessive-compulsive
symptoms in a normative Chinese sample of youth: prevalence, symptom
dimensions, and factor structure of the Leyton Obsessional Inventory-Child
Version. J. Affect. Disord. 164, 1927.
Sun, J., Stewart, D., 2007. Development of population based resilience measures in
the primary school setting. Health Educ. 107, 575599.
Taner, Y.I., Bakar, E., Oner, O., 2011. Impaired executive functions in paediatric
obsessive-compulsive disorder patients. Acta Neuropsychiatr. 23, 272281.
Wilcox, H.C., Grados, M., Samuels, J., Riddle, M.A., Bienvenu III, O.J., Pinto, A., Cullen,
B., Wang, Y., Shugart, Y.Y., Liang, K.Y., Nestadt, G., 2008. The association between
parental bonding and obsessive compulsive disorder in offspring at high
familial risk. J. Affect. Disord. 111, 3139.
Xia, C.Y., 1995. Validation of obsessive-compulsive scale in secondary school and
university students. Chin. J. Clin. Psychol. 3, 173174.
Yang, J., Gong, Y.X., 1993. The revision of the California psychological inventory in
China. Chin. J. Clin. Psychol. 1, 1115.
Yuan, L.X., Zhang, H.C., 1996. Secondary school student school adaptive ability scale
development. Beijing Norm. Univ. Bull. 32, 9095.
J. Sun et al. / Journal of Affective Disorders 168 (2014) 459465 465