Child Psychiatry Hum Dev DOI 10.1007/s10578-014-0461-7
ORIGINAL PAPER
Anxiety Disorder Symptoms in Chinese Preschool Children Meifang Wang • Jinxia Zhao
Ó Springer Science+Business Media New York 2014
Abstract The present study investigated anxiety disorder symptoms in Chinese preschool children. A total of 1,854 mothers of children aged 3–6 years completed the 28-item Chinese version of Spence Preschool Anxiety Scale (PAS). Results demonstrated that the structure of anxiety in Chinese preschool children included five factors, and this fivefactor structure applied to different age and gender groups. Inconsistent with the study by Edwards et al. (J Clin Child Adolesc Psychol 39:400–409, 2010) suggesting that obsessive–compulsive disorder did not comprise an independent factor in Australian preschoolers, this symptom can be viewed as an independent factor in Chinese preschoolers. Younger children displayed higher anxiety levels than older children. Anxiety symptoms of Chinese preschoolers were found at a high level compared to other studies. Substantial differences were found with regard to the content of prevalent anxiety symptoms among Chinese and Australian preschool children. The applicability of the five-factor structure of PAS and research implications are discussed. Keywords Anxiety Preschool children Preschool Anxiety Scale China
M. Wang (&) J. Zhao Department of Psychology, Shandong Normal University, No. 88 East Wenhua Road, Jinan 250014, People’s Republic of China e-mail: [email protected]; [email protected] J. Zhao Department of Education, Linyi University, Linyi, People’s Republic of China
Introduction In the past two decades, research efforts in childhood and adolescence anxiety disorders have greatly increased our understanding of anxiety in older children. However, we know relatively little about the anxiety problems in very young children [1, 2]. Research suggests that anxiety disorders have an early age of onset [3, 4], with prevalence rates around 9 % in preschool populations [1]. This prevalence among young children is comparable to those reported for older children [5]. Meanwhile, preschool anxiety problems have demonstrated stability into midchildhood and adolescence [6, 7], and are associated with future onset of other disorders like depression and conduct disorders [8–10]. These early-onset disorders may result in moderate to marked life impairment for young children if left untreated [4, 11]. All these findings highlight the importance of understanding and screening anxiety symptoms in early childhood. Despite early onset and severity of anxiety disorders, it is rare that young children with anxiety problems receive appropriate care and treatment [1, 5]. One of the main barriers is the lack of reliable and valid instruments for assessing preschool anxiety. Thus, psychometrically sound instruments for identifying anxiety symptoms in early childhood are urgently needed. Currently, several measures such as the Child Behavior Checklist [12], the Strengths and Difficulties Questionnaire [13], and the Children’s Moods, Fears and Worries Questionnaire [6] are often used for assessing anxiety of preschool children. Although these measures possess adequate psychometric properties and provide valuable information on psychopathological symptoms of preschool children, they do not identify the specific anxiety symptoms. Instead, they tend to address more global aspects of internalizing behaviors, such as
123
Child Psychiatry Hum Dev
anxiety, depression, and withdrawal. Previous studies suggest that the patterns of anxiety symptoms can already be sub-typed in the preschool period [1, 14]. Therefore, there seems to be a need for symptom-specific and ageappropriate instruments to assess anxiety problems in preschool children. To the best of our knowledge, the Preschool Anxiety Scale (PAS) is the only measure that specifically assesses multiple anxiety symptoms in preschool-aged children [14]. The PAS is a 28-item parent-report measure that was designed to assess anxiety dimensions specified in the DSM-IV. The factor structure and construct validity of the PAS were examined in a large Australian community sample [14]. Confirmatory factor analysis revealed five factors: separation anxiety disorder, physical injury fears, social phobia, obsessive–compulsive disorder, and generalized anxiety disorder. Construct validity of the PAS was also supported by significant correlations with internalizing factor of the CBCL to a greater extent than with externalizing factor of the CBCL. Notably, several items relating to obsessive–compulsive disorder in the PAS were reported infrequently by Australian parents. Specifically, only 1.84 % of mothers and .76 % of fathers rated these items as ‘‘quite often true’’ or ‘‘very often true’’. Therefore, in a subsequent study, Edwards et al. [11] slightly modified this measure with another Australian sample, resulting in a revised version of the PAS (PAS-R). Psychometric evaluation of the PAS-R indicated that, except for obsessive– compulsive disorder, the total scale and the other subscales displayed satisfactory internal consistency, test–retest reliability (over a period of 12 months) and good discriminate validity. Given the fact that obsessive–compulsive disorder subscale had quite poor psychometrics (mother/father report: a = .46/.41; 12-month reliability = .57/.51) and relatively infrequent endorsement in Australian samples, this subscale was therefore removed from the PAS-R. These findings seem to indicate that the symptom of obsessive–compulsive disorder is not included in the structure of anxiety among Australian preschoolers. However, there is evidence that obsessive–compulsive disorder represented an independent factor among Dutch preschoolers. Specifically, Broeren and Muris [15] investigated the applicability of the original PAS in Dutch children aged 2–6 years. Exploratory factor analysis revealed five factors that are broadly consistent with the structure of the original PAS, and all the five factors had acceptable internal consistency and construct validity. The aforementioned findings demonstrated that the structure of anxiety symptoms in very young children may differ across cultures. To date, the psychometric properties of the PAS have been examined predominately in Australian and Dutch community samples. Given the reported variation in the
123
structure of anxiety symptoms in different countries, the current study intends to examine the structure of anxiety symptoms in Chinese preschool children. A Chinese translation of the PAS has already been used in at least one study [16]. In that study, the items relating to obsessive– compulsive disorders in the original PAS were reported frequently by the mothers of Chinese preschoolers, from which we can infer that obsessive–compulsive disorder may exist as a distinct factor in Chinese preschool children. Based on the above mentioned reasons, this study will explore the structure of anxiety in Chinese preschoolers using the original PAS containing the symptom of obsessive–compulsive disorder. Meanwhile, this study also evaluates the factorial invariance of the structure of anxiety symptoms in different age (3- and 4-year-olds and 5- and 6year-olds) and gender (males and females) groups, since a wide range of evidence seems to indicate that children’s anxiety symptoms vary significantly between these groups [17, 18]. Additionally, considering the differences in socialization practices between Chinese and Western cultures, the anxiety levels of Chinese children may be different from those of Western peers. For example, in a study using CBCL to examine preschoolers’ behavioral problems, Liu, Cheng, and Leung [19] found that Chinese preschool children scored higher on internalizing problems than their American counterparts. Therefore, the other aim of this study is to compare anxiety symptoms of the current Chinese community sample with those found in previous studies of the PAS in other cultures. In summary, the primary purpose of this study was to examine anxiety disorder symptoms of Chinese preschool children in several ways: (1) to examine the factor structure of anxiety symptoms in Chinese preschool children; (2) to investigate age and gender differences and the prevalence of anxiety symptoms in Chinese preschool children; (3) to compare anxiety symptoms of Chinese preschool children to those of Australian/Dutch counterparts.
Methods Participants Participants consisted of 1,854 Chinese young children aged 3–6 years (M = 4.93 years, SD = .95) and their mothers (M = 33.05 years, SD = 2.62). They were recruited from six public kindergartens in the metropolitan areas of Jinan and Linyi, located in Shandong Province, Eastern China. The kindergartens were selected by a clustered random sampling method. A total of 988 boys and 866 girls from 63 classes and their mothers participated in the present study. The majority of mothers had a high school (20 %) or a college/university education (74 %).
Child Psychiatry Hum Dev
Procedure The design of this study was approved by the Institutional Review Board at Shandong Normal University. Permissions to administer questionnaires from the kindergarten principals and maternal consents were obtained prior to data collection. Mothers were invited to participate in a meeting at the kindergarten. Approximately 90 % of mothers attended the meeting. During the meeting, they completed the Chinese version of the PAS and a questionnaire on family demographics. Those mothers who were absent from the meeting on the day of testing (e.g., due to illness or working) received a packet containing questionnaires and instructions from their child’s teacher. They were asked to fill it out at home and return the completed form to their child’s teacher. To examine the test–retest reliability of the original PAS in Chinese culture, a subgroup of mothers (n = 217) were retested over a period of one month after the initial screening. To further examine the convergent and divergent validity of the measure, a subgroup of mothers (n = 1,076) were also asked to complete the Children Behavior Checklist [12] at the first screening. Measures Preschool Anxiety Scale (PAS) As mentioned previously, the PAS is a 28-item parentreport measure of anxiety symptoms for preschool children [14]. The PAS consists of five subscales: separation anxiety disorder (five items), physical injury fears (seven items), social phobia (six items), obsessive–compulsive disorder (five items) and generalized anxiety disorder (five items). Mothers were asked to rate the items of each subscale on a five-point scale ranging from 0 (not at all true) to 4 (very often true). The scores of total scale and each subscale can be calculated by adding the responses of the relevant items. The English version of the PAS was translated into Chinese by the authors, and then back-translated to the original language by a bilingual translator. It was assured that the content of the translated Chinese version was similar to the original English version by the translators. A study with Chinese preschool children suggested that the item ‘‘Is nervous of going swimming’’ was most frequently missing [16]. Thus, a new item ‘‘Is nervous of going across the road’’ was added considering the growing traffic problems and parents’ emphases on children’s security in China. Frequency analyses revealed that, in the present study, the item ‘‘Is nervous of going swimming’’ and the new added item ‘‘Is nervous of going across the road’’ had 10.21 and 1.7 % missing values, respectively. Following the criterion employed by Broeren and Muris
[15], a threshold of 5 % missing values was used to remove unsatisfactory items in this study. The item ‘‘Is nervous of going swimming’’ was therefore excluded from the scale and replaced by the new added item ‘‘Is nervous of going across the road’’ in this study. Children Behavior Checklist (CBCL) The CBCL designed for 4–18 year old children [12] was selected as an indicator of convergent and divergent validity of the PAS. It is a 118-item parent-report measure to assess two broad band factors, Internalizing and Externalizing. Mothers were asked to evaluate on a three-point scale (0 = not true, 1 = somewhat true, 2 = very true) regarding whether the behavior is occurring now or has occurred over the past 6 months. The CBCL was validated in China and demonstrated adequate reliability and validity [20]. In the current sample, Cronbach’s alpha coefficients were .92 for the total scale, .86 for Internalizing factor and .87 for Externalizing factor. Data Analysis To examine the factor structure of anxiety symptoms in Chinese preschool children, CFA was performed using Robust Maximum Likelihood estimation in LISREL 8.70. Since in large sample sizes the v2 statistic is likely to be significant, normed fit index (NFI), non-normed fit index (NNFI), comparative fit index (CFI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR) were calculated in the present study to evaluate the fit of the model. Values of NFI, NNFI and CFI greater than .90 indicate an acceptable fit [21]. RMSEA value below .06 and SRMR value below .08 indicate a relatively good fit [22]. To further examine whether the factor structure of anxiety symptoms is invariant across age and gender groups, we conducted two multi-group confirmatory factor analyses. In each multigroup analysis, a model in which all factor loadings were constrained to be equal (constrained model) across age or gender groups was compared to a model in which these factor loadings were free to vary (unconstrained model) across groups. Following the criterion for evaluating factorial invariance across subgroups described by Cheung and Rensvold [23], the differences in goodness-of-fit indices (DGFIs) between constrained and unconstrained models less or equal to .01 are considered reasonably invariant. Additionally, Cronbach’s alpha coefficients were calculated to evaluate the internal consistency of the PAS total and subscales. Pearson product moment correlations were utilized to assess test–retest reliability and convergent and divergent validity. Age and gender differences in anxiety
123
Child Psychiatry Hum Dev
scores were examined using multivariate analyses of variance (MANOVA) and effect size statistics. The prevalence of anxiety symptoms was evaluated through frequency statistics. Missing values were estimated in SPSS using the estimated means (EM) procedure.
Results Confirmatory Factor Analysis The correlated five-factor model reported by Spence et al. [14] in the original PAS was tested in current Chinese preschool populations. The results demonstrated that all 28 items had loadings in excess of .31 on their hypothesized factors (see Table 1) and that the values for fit indices fell within the acceptable to good ranges (NFI = .93, NNFI = .94, CFI = .94, RMSEA = .057, SRMR = .056), indicating a satisfactory fit of the five-factor model to the present data. The five factors were found to be strongly intercorrelated, with rs ranging from .52 to .87. Multi-group confirmatory factor analyses were conducted to test the factorial invariance of this model across demographic groups. We first tested whether the model is invariant between younger (3- and 4-year-olds) and older (5- and 6- year-olds) children. Table 2 shows that the model in which the factor loadings were constrained to be equal between younger and older children provided a good fit, very similar to that of the model in which the factor loadings were free to vary across groups. The differences in GFIs below .01 suggested factorial invariance across age groups. After combining the data of younger and older children, we then tested the factorial invariance of the model between boys and girls. Results from model comparisons showed that the differences in GFIs were less than .01 between unconstrained and constrained models (see table 2), which supported the factorial invariance of the model across gender groups. Internal Consistency and Test–Retest Reliability Cronbach’s alpha coefficients were calculated to evaluate the internal consistency of the PAS in Chinese preschool children. Coefficient a values were .87 for the total scale, .55 for separation anxiety disorder, .72 for physical injury fears, .75 for social phobia, .68 for obsessive–compulsive disorder, and .70 for generalized anxiety disorder. To examine test–retest reliability of the Chinese PAS, a total of 217 participants (M = 5.11 years, SD = 1.01; 110 boys and 107 girls) were reassessed one month later. Results demonstrated that the Pearson correlation coefficients (r) were .73 for the total scale, .58 for separation anxiety disorder, .71 for physical injury fears, .64 for social
123
Table 1 Factor loadings of the five-factor model after confirmatory factor analysis Item
SAD
6. Is reluctant to go to sleep without you or to sleep away from home
.31
12. Worries that something bad will happen to his/her parents
.56
16. Worries that something bad might happen to him/her (e.g., getting lost or kidnapped), so he/ she won’t be able to see you again
.58
22. Becomes distressed about your leaving him/her at preschool/ school or with a babysitter
.42
25. Has nightmares about being apart from you
.48
PIF
7. Is scared of heights (high places)
.46
10. Is afraid of crowded or closedin places 13. Is scared of thunder storms
.58
17. Is nervous of going across the road
.53
20. Is afraid of insects and/or spiders
.48
SP
OCD
.56
24. Is frightened of dogs
.44
26. Is afraid of the dark
.55
2. Worries that he/she will do something to look stupid in front of other people
.60
5. Is scared to ask an adult for help (e.g., a preschool teacher)
.56
11. Is afraid of meeting or talking to unfamiliar people
.58
15. Is afraid of talking in front of the class (e.g., show and tell)
.59
19. Worries that he/she will do something embarrassing in front of other people
.62
23. Is afraid to go up to group of children and join their activities
.52
3. Keeps checking that he/she has done things right (e.g., that he/ she closed a door, turned off a tap)
.56
9. Washes his/her hands over and over many times each day
.50
18. Has to have things in exactly the right order or position to stop bad things from happening
.54
21. Has bad or silly thoughts or images that keep coming back over and over
.56
27. Has to keep thinking special thoughts (e.g., numbers or words) to stop bad things from happening
.59
GAD
Child Psychiatry Hum Dev
Age and Gender Differences
Table 1 continued Item
SAD
PIF
SP
OCD
GAD
1. Has difficulty stopping him/ herself from worrying
.53
4. Is tense, restless or irritable due to worrying
.66
8. Has trouble sleeping due to worrying
.56
14. Spends a large part of each day worrying about various things
.58
28. Asks for reassurance when it does not seem necessary
.53
SAD separation anxiety disorder, PIF physical injury fears, SP social phobia, OCD obsessive–compulsive disorder, GAD generalized anxiety disorder
A MANOVA on the total score and subscales of the PAS was conducted to examine age and gender differences of anxiety symptoms in preschool children. Based on the Wilks’ lambda criterion, results of the MANOVA indicated that the combined dependent variables were significantly different between younger (3- and 4-year-olds) and older (5- and 6-year-olds) groups [F(5,1846) = 7.81, p \ .001, g2 = .021], but not for gender groups [F(5,1846) = 1.52, p [ .05, g2 = .004] or gender by age interactions [F(5,1846) = .15, p [ .05, g2 = .000]. Specifically, younger children were found to score significantly higher on the total score and all the subscales of the PAS as compared to the older children (see table 3). Prevalence of Anxiety Symptoms
phobia, .63 for obsessive–compulsive disorder, and .59 for generalized anxiety disorder. Convergent and Divergent Validity To further explore the convergent and divergent validity of the Chinese PAS, correlations between the PAS and CBCL subscales were calculated in a subgroup of 1,076 participants. The total score and subscales of the PAS correlated significantly with the CBCL Internalizing score (r = .31 to .59) and with the CBCL Externalizing score (r = .21 to .40). Following the recommendation of Meng et al. [24] for significant correlations, a series of Z tests were conducted to determine whether correlations between PAS scales and Internalizing score (convergent relationship) are significantly stronger than correlations between PAS scales and Externalizing score (divergent relationship). The results indicated that the correlations with the Internalizing score are significantly greater than the correlations with the Externalizing score for all scales of the PAS, Z-scores ranged from 3.48 to 10.91, ps \ .001. Overall, this finding supported the convergent and divergent validity of the PAS.
Following previous studies [14, 25], the most prevalent symptoms of Chinese preschool children were examined by calculating the percentage of mothers who rated each item as either 3 (quite often true) or 4 (very often true) for their children. As shown in Table 4, the most prevalent symptoms in Chinese preschool children were related to separation anxiety (‘‘Is reluctant to go to sleep without you or to sleep away from home’’), physical injury fears (dark, dogs, insects and/or spiders, thunder storms, crowded or closedin places) and social phobia (‘‘Is afraid of talking in front of the class or preschool group’’). Additionally, three items (‘‘Has things in exactly the right position’’, ‘‘Washes his/ her hands over and over many times each day’’, and ‘‘Keeps checking that he/she has done things right’’) relating to obsessive–compulsive disorder were also among the top 10 presenting problems. Comparisons to Previous Studies To investigate whether the means of anxiety symptoms experienced by the current Chinese community sample differed from those found in previous studies of the PAS
Table 2 Tests of factorial invariance across age (younger and older) and gender (boys and girls) groups v2
df
p
NFI
NNFI
CFI
RMSEA
SRMR
Unconstrained model
2,758.94
680
\.001
.922
.933
.940
.057
.058
Constrained model
2,808.56
703
\.001
.920
.934
.939
.057
.058
Model Invariance across age groups
Invariance across gender groups Unconstrained model
2,791.22
680
\.001
.921
.932
.939
.058
.057
Constrained model
2,826.64
703
\.001
.920
.934
.939
.058
.059
NFI normed fit index, NNFI non-normed fit index, CFI non-normed fit index, RMSEA root mean square error of approximation, SRMR standardized root mean square residual
123
123
Discussion
* p \ .05; ** p \ .01; *** p \ .001
SAD separation anxiety disorder, PIF physical injury fears, SP social phobia, OCD obsessive–compulsive disorder, GAD generalized anxiety disorder
1.43 ± 2.19
2.95 ± 2.91 18.84 ± 11.57
with Australian [14] and Dutch community samples [15], a series of t tests were conducted. As presented in Table 3, Chinese preschool children in the current sample were found to score significantly higher on the total scale and subscales of the PAS than Australian [ts(1,853) [ 8.70, ps \ .001] and Dutch samples [ts(1,853) [ 19.43, ps \ .001] in previous studies, with exception of the social phobia subscale on which no significant difference was found between Chinese preschool children and their Dutch peers (t = .21, p [ .05). To compare the most prevalent anxiety symptoms of Chinese preschoolers with those found in prior research of the PAS, Table 4 also presents the results reported by Spence et al. [14] in Australian community samples. The contents of the most prevalent anxiety symptoms in Chinese preschool children and Australian samples showed substantial differences, although they also displayed many similarities. Specifically, three items referring to obsessive–compulsive disorder mentioned above, which were rare in Australian preschoolers, were evidently reported frequently in Chinese preschool children.
2.15 ± 2.57 17.28 ± 11.83
1.23 ± 2.08 3.98 ± 3.21 .58
.039 -.02
.65 .19
-.034 4.04 ± 3.24
4.21 ± 2.75 27.62 ± 13.00 6.26* 18.86***
5.54* .109
.115 .201
3.81 ± 3.16 4.16 ± 3.25
4.33 ± 2.92 29.08 ± 13.48
OCD
GAD Total
4.00 ± 2.78 26.44 ± 12.83
3.93 ± 3.19
4.10 ± 2.95 27.87 ± 13.46
4.16 ± 2.85 27.73 ± 13.21
6.69 ± 4.11
5.43 ± 4.22 4.66 ± 3.74
6.51 ± 4.37 8.91 ± 4.91
.07 .013
5.45 ± 3.90
3.01 -.082 9.12 ± 4.99 8.72 ± 4.83
5.47 ± 3.89 11.09**
6.03* .114
.152 5.16 ± 3.75
9.19 ± 4.90
5.75 ± 4.04
PIF
SP
8.63 ± 4.91
5.42 ± 3.92
.45 .031 5.19 ± 2.91 5.28 ± 2.93 36.55*** .28 4.84 ± 2.71 5.65 ± 3.07
Cohen’s d 5–6 years
SAD
Cohen’s d Girls Boys 3–4 years
F
Total by gender Total by age
Table 3 Means and standard deviations of the PAS (mean ± SD) for current and previous community samples [14, 15]
F
5.24 ± 2.92
Current sample
2.73 ± 2.85
Australian sample
2.34 ± 2.41
Dutch sample
Child Psychiatry Hum Dev
Although a large body of research has consistently shown that anxiety disorders have an early onset in childhood [1, 4, 14], little research has focused on the anxiety problems of very young children. The present study contributes to the existing literature by examining the nature of anxiety symptoms in Chinese preschool children. Confirmatory factor analysis indicated that the correlated five-factor model (including separation anxiety disorder, physical injury fears, social phobia, obsessive– compulsive disorder, and generalized anxiety disorder) in the original PAS fit the current data well. Further multigroup analyses supported the factorial invariance of this model across age and gender groups in Chinese preschool populations. These findings demonstrated that the anxiety symptoms of Chinese preschoolers reflected a five-factor structure, and that both younger and older children, and both boys and girls presented the same pattern of anxiety symptoms. Notably, the items representing obsessive– compulsive disorder comprised an independent factor in Chinese preschool children, which is inconsistent with the study by Edwards and colleagues [11] suggesting that obsessive–compulsive disorder did not comprise an independent factor in Australian young children. The reliability and validity of the Chinese PAS provided empirical support for the five-factor structure of anxiety (particularly for obsessive–compulsive disorder as an independent factor of anxiety) in Chinese preschool populations. Specifically, the current study found that, except
Child Psychiatry Hum Dev Table 4 The percentage of children receiving mother ratings of 3 or 4 (quite often true or very often true) for current and previous Australian samples [14] Item
Table 4 continued Item
Current sample (%)
Australian sample (%)
14. Spends a large part of each day worrying about various things
1.6
.4
Current sample (%)
Australian sample (%)
6. Is reluctant to go to sleep without you or to sleep away from home
44
12.7
21. Has bad or silly thoughts or images that keep coming back over and over
1.1
1.2
26. Is afraid of the dark
26
17.8
1.2
20.7
17.8
25. Has nightmares about being apart from you
.9
24. Is frightened of dogs 20. Is afraid of insects and/or spiders
16.8
10.6
18. Has to have things in exactly the right order or position to stop bad things from happening
14.3
1.6
9. Washes his/her hands over and over many times each day
14.1
1.4
13. Is scared of thunder storms
12.7
13.2
10. Is afraid of crowded or closed-in places
12.2
1.8
15. Is afraid of talking in front of the class (preschool group) e.g., show and tell
11.6
5.9
3. Keeps checking that he/she has done things right (e.g., that he/she closed a door, turned off a tap)
9.9
4.0
7. Is scared of heights (high places)
9.0
4.0
11. Is afraid of meeting or talking to unfamiliar people
8.8
9.8
5. Is scared to ask an adult for help (e.g., a preschool or school teacher)
7.1
2.4
4. Is tense, restless or irritable due to worrying
6.7
1.2
28. Asks for reassurance when it doesn’t seem necessary
6.1
2.4
19. Worries that he/she will do something embarrassing in front of other people
5.6
1.0
17. Is nervous of going across the road
5.3
–
22. Becomes distressed about your leaving him/her at preschool/school or with a babysitter
5.1
4.5
1. Has difficulty stopping him/herself from worrying
5.1
3.9
12. Worries that something bad will happen to his/her parents
4.7
2.8
16. Worries that something bad might happen to him/her (e.g., getting lost or kidnapped), so he/she won’t be able to see you again
4.6
1.4
2. Worries that he/she will do something to look stupid in front of other people 23. Is afraid to go up to group of children and join their activities
4.4
4.3
3.3
7.7
27. Has to keep thinking special thoughts (e.g., numbers) to stop bad things from happening
2.0
1.0
8. Has trouble sleeping due to worrying
1.9
1.6
for separation anxiety subscale having an a of .55, the PAS total scale and subscales displayed satisfactory levels of internal consistency reliability, test–retest reliability and convergent and divergent validity. These results supported the applicability of the five-factor structure of PAS in Chinese preschool populations. Inconsistent with the study by Edwards and colleagues suggesting that the obsessive– compulsive disorder subscale had quite poor psychometrics [11], this subscale had acceptable internal consistency and test–retest reliability in the present study, which suggested that obsessive–compulsive disorder was an important section of the five-factor structure of anxiety in very young Chinese children. The prevalence of anxiety symptoms provided further support for obsessive–compulsive disorder as an independent factor in the structure of anxiety for Chinese preschoolers. In specific, comparisons of the prevalent PAS items (rated as ‘‘quite often true’’ or ‘‘very often true’’) indicated that there were substantial differences regarding the contents of the prevalent anxiety symptoms among Chinese and Australian preschoolers. As mentioned above, three items (‘‘Has things in exactly the right position’’, ‘‘Washes his/her hands over and over many times each day’’, and ‘‘Keeps checking that he/she has done things right’’) representing obsessive–compulsive disorder, which were rare in Australian preschool children [14], were reported frequently by the mothers of Chinese preschoolers. Even though it remains unclear why these symptoms are relatively prevalent among Chinese preschool children, this finding is consistent with the notion that cultural factors are involved in the manifestation of children’s anxiety symptoms [26]. Parents’ over-controlling practices and teachers’ rigorous instructional styles in Chinese societies may be related to the high prevalence of these symptoms. For instance, in China, very specific instructions (e.g., ‘‘Has toys in exactly the right order or position’’) are commonly given at home and in the kindergartens. This interpretation requires further exploration. Consistent with the study by Spence et al. [14] demonstrating age differences in preschoolers’ anxiety symptoms, our data showed that younger children (3- and
123
Child Psychiatry Hum Dev
4-year-olds) were reported to have higher anxiety levels in comparison to older children (5- and 6-year-olds). It is hypothesized that the high level of anxiety symptoms in younger children may be related to their life transition. Most Chinese children enter into kindergarten and start a new period in their life at age 3–4 years. They may not only suffer from the fear of leaving their parents and home but also experience difficulties adjusting to kindergarten. Thus, the elevated levels of anxiety symptoms in younger children may be a response to this transition. Consistent with earlier studies relating to preschoolers [1, 5, 14], no significant gender differences in anxiety scores were found in the present study. In contrast, much of the research with school-age children found that girls report higher anxiety scores than boys [27, 28]. Therefore, it seems that gender differences in anxiety symptoms do not emerge in preschool period, and may become more distinct with increasing age. Comparisons to previous studies revealed that, with exception of social phobia, Chinese preschool children displayed higher anxiety levels than their Australian and Dutch peers in previous studies [14, 15]. These crossnational differences in anxiety levels for younger children are comparable to those reported for older children and adolescents. For instances, Chinese school children reported higher degree of anxiety than their American counterparts [29]. Chinese adolescents reported higher anxiety levels than their Dutch and German peers [30]. These discrepancies may be accounted for by different socialization practices in Chinese and Western cultures. It has been speculated that Chinese socialization practices stress self-control, emotional restraint, and obedience to authority [31, 32], which may contribute to children’s internalizing issues such as anxiety. Western socialization practices, by contrast, encourage individualism and independence [33], which in turn may contribute to the lowering of anxiety problems. From a cultural context, the manifestations of anxiety in Chinese children may be adaptive and may represent ‘‘normal’’ reactions to dominant cultural forces [19]. In this vein, a higher level of anxiety may not necessarily reflect the presence of anxiety pathology in Chinese preschool children. It should be acknowledged that the present study suffers from several limitations. First, only community sample was included in this study, so it is not clear whether our findings can be generalized to clinical populations. Second, the current study only recruited the samples from the Shandong region. It remains unknown whether our results can be applied to preschool children in other Chinese regions. Further research needs to recruit some representative samples from different regions in China to replicate these results. Third, the data of this study were based solely on mother-report. Although parents as informants of young
123
children’s anxiety are satisfactory [14, 15], future studies still need to include direct behavioral observations or anxiety ratings by teachers to confirm this measure. Finally, the relatively low alpha for separation anxiety disorder subscale (a = .55) in the current sample must be acknowledged, although it is consistent with the result (a = .59) in the previous study [15]. It may be due to the low number of items for this subscale (n = 5) and the fact that this subscale includes individual items which intend to assess children’s separation anxiety but are only loosely related to each other. This seems to indicate that the individual items of this subscale may have greater utility than the aggregate [34].
Summary The current study makes significant inroads to understanding anxiety problems in Chinese preschool children. This study found that the structure of anxiety in Chinese preschool children included five factors, and this five-factor structure applied to different age and gender groups in Chinese preschool populations. Inconsistent with prior research [11], obsessive–compulsive disorder had satisfactory psychometrics and high endorsement, and can be viewed as an independent anxiety factor in Chinese preschool children. Compared to other studies with Western samples, anxiety symptoms of Chinese preschool children were found at a high level. More studies are needed to explore cultural factors that account for obsessive–compulsive disorder as an independent factor and the higher rates of anxiety in Chinese young children.
References 1. Egger HL, Angold A (2006) Common emotional and behavioral disorders in preschool children: presentation, nosology, and epidemiology. J Child Psychol Psychiatry 47:313–337 2. Eley TC, Bolton D, O’Connor TG, Perrin S, Smith P, Plomin R (2003) A twin study of anxiety-related behaviors in preschool children. J Child Psychol Psychiatry 44:945–960 3. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE (2005) Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in National Comorbidity Survey Replication. Arch Gen Psychiatry 62:593–602 4. Rapee RM, Schniering CA, Hudson JL (2009) Anxiety disorders during childhood and adolescence: origins and treatment. Annu Rev Clin Psychol 5:311–341 5. Furniss T, Beyer T, Guggenmos J (2006) Prevalence of behavioral and emotional problems among six-years-old preschool children. Soc Psychiatry Psychiatr Epidemiol 41:394–399 6. Bayer JK, Sanson AV, Hemphill SA (2006) Children’s moods, fears, and worries: development of an Early Childhood Parent Questionnaire. J Emot Behav Disord 14:41–49
Child Psychiatry Hum Dev 7. Warren SL, Huston L, Egeland B, Sroufe LA (1997) Child and adolescent anxiety disorders and early attachment. J Am Acad Child Adolesc Psychiatry 36:637–644 8. Cole DA, Peeke LG, Martin JM, Truglio R, Seroczynski AD (1998) A longitudinal look at the relation between depression and anxiety in children and adolescents. J Consult Clin Psychol 66:451–460 9. Kovacs M, Devlin B (1998) Internalizing disorders in childhood. J Child Psychol Psychiatry 39:47–63 10. Zahn-Waxler C, Klimes-Dougan B, Slattery MJ (2000) Internalizing problems of childhood and adolescence: prospects, pitfalls, and progress in understanding the development of anxiety and depression. Dev Psychopathol 12:443–466 11. Edwards SL, Rapee RM, Kennedy SJ, Spence SH (2010) The assessment of anxiety symptoms in preschool-aged children: the revised Preschool Anxiety Scale. J Clin Child Adolesc Psychol 39:400–409 12. Achenbach TM (1991) Manual for the Child Behavior Checklist/ 4–18 and 1991 profile. University of Vermont, Department of Psychiatry, Burlington 13. Goodman R (1997) The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry 38:581–586 14. Spence SH, Rapee R, McDonald C, Ingram M (2001) The structure of anxiety symptoms among preschoolers. Behav Res Ther 39:1293–1316 15. Broeren S, Muris P (2008) Psychometric evaluation of two new parent-rating scales for measuring anxiety symptoms in young Dutch children. J Anxiety Disord 22:949–958 16. Zhao JX, Wang MF (2009) Developmental characteristics of preschoolers’ anxiety. Chin J Clin Psychol 17:723–725 17. Crocetti E, Hale WW, Fermani A, Raaijmakers Q, Meeus W (2009) Psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED) in the general Italian adolescent population: a validation and a comparison between Italian and the Netherlands. J Anxiety Disord 23:824–829 18. Hale WW, Crocetti E, Raaijmakers QAW, Meeus WHJ (2011) A meta-analysis of the cross-cultural psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED). J Child Psychol Psychiatry 52:80–90 19. Liu JH, Cheng H, Leung PWL (2011) The application of the Preschool Child Behavior Checklist and the Caregiver–Teacher report form to Mainland Chinese children: syndrome structure, gender differences, country effects, and inter-informant agreement. J Abnorm Child Psychol 39:251–264 20. Su LY, Li XR, Yang ZW, Luo XR, Wan GB (1997) The application of Child Behavior Checklist and Teachers’ Report Form in epidemiological study of child mental health: choosing an optimal cut-off point. Chin Mental Health J 11:37–40
21. Marsh HW, Hau KT (1996) Assessing goodness of fit: is parsimony always desirable? J Exp Educ 64:364–390 22. Hu L, Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model 6:1–55 23. Cheung GW, Rensvold RB (2002) Evaluating goodness-of-fit indexes for testing measurement invariance. Struct Equ Model 9:233–255 24. Meng XL, Rosenthal R, Rubin DB (1992) Comparing correlated correlation coefficients. Psychol Bull 111:172–175 25. Ishikawa S, Shimotsu S, Ono T, Sasagawa S, Kondo-Ikemura K, Sakano Y, Spence SH (2013) A Parental Report of Children’s Anxiety Symptoms in Japan. Child Psychiatry Hum Dev. doi:10. 1007/s10578-013-0401-y 26. Muris P, Schmidt H, Engelbrecht P, Perold M (2002) DSM-IVdefined anxiety disorder symptoms in South African children. J Am Acad Child Adolesc Psychiatry 41:1360–1368 27. Ishikawa SI, Sato H, Sasagawa S (2009) Anxiety disorder symptoms in Japanese children and adolescents. J Anxiety Disord 23:104–111 28. Su LY, Wang K, Fan F, Su Y, Gao XP (2008) Reliability and validity of the Screen for Child Anxiety Related Emotional Disorders (SCARED) in Chinese children. J Anxiety Disord 22:612–621 29. Weine AM, Phillips JS, Achenbach TM (1995) Behavioral and emotional problems among Chinese and American children: parent and teacher reports for ages 6 to 13. J Abnorm Child Psychol 23:619–639 30. Zhao JX, Xing XP, Wang MF (2012) Psychometric properties of the Spence Children’s Anxiety Scale (SCAS) in Mainland Chinese children and adolescents. J Anxiety Disord 26:728–736 31. Chen X, Rubin K, Li B (1997) Maternal acceptance and social and school adjustment in Chinese children: a four-year longitudinal study. Merrill Palmer Q 43:663–681 32. Xie D, Leong FTL (2008) A cross-cultural study of anxiety among Chinese and Caucasian American university students. J Multicult Couns Dev 36:52–63 33. Essau CA, Leung PWL, Conradt J, Cheng H, Psych D, Wong T (2008) Anxiety symptoms in Chinese and German adolescents: their relationship with early learning experiences, perfectionism, and learning motivation. Depress Anxiety 25:801–810 34. Whiteside SP, Brown AM (2008) Exploring the utility of the Spence Children’s Anxiety Scales parent- and child-report forms in a North American sample. J Anxiety Disord 22:1440–1446
123
ORIGINAL PAPER
Anxiety Disorder Symptoms in Chinese Preschool Children Meifang Wang • Jinxia Zhao
Ó Springer Science+Business Media New York 2014
Abstract The present study investigated anxiety disorder symptoms in Chinese preschool children. A total of 1,854 mothers of children aged 3–6 years completed the 28-item Chinese version of Spence Preschool Anxiety Scale (PAS). Results demonstrated that the structure of anxiety in Chinese preschool children included five factors, and this fivefactor structure applied to different age and gender groups. Inconsistent with the study by Edwards et al. (J Clin Child Adolesc Psychol 39:400–409, 2010) suggesting that obsessive–compulsive disorder did not comprise an independent factor in Australian preschoolers, this symptom can be viewed as an independent factor in Chinese preschoolers. Younger children displayed higher anxiety levels than older children. Anxiety symptoms of Chinese preschoolers were found at a high level compared to other studies. Substantial differences were found with regard to the content of prevalent anxiety symptoms among Chinese and Australian preschool children. The applicability of the five-factor structure of PAS and research implications are discussed. Keywords Anxiety Preschool children Preschool Anxiety Scale China
M. Wang (&) J. Zhao Department of Psychology, Shandong Normal University, No. 88 East Wenhua Road, Jinan 250014, People’s Republic of China e-mail: [email protected]; [email protected] J. Zhao Department of Education, Linyi University, Linyi, People’s Republic of China
Introduction In the past two decades, research efforts in childhood and adolescence anxiety disorders have greatly increased our understanding of anxiety in older children. However, we know relatively little about the anxiety problems in very young children [1, 2]. Research suggests that anxiety disorders have an early age of onset [3, 4], with prevalence rates around 9 % in preschool populations [1]. This prevalence among young children is comparable to those reported for older children [5]. Meanwhile, preschool anxiety problems have demonstrated stability into midchildhood and adolescence [6, 7], and are associated with future onset of other disorders like depression and conduct disorders [8–10]. These early-onset disorders may result in moderate to marked life impairment for young children if left untreated [4, 11]. All these findings highlight the importance of understanding and screening anxiety symptoms in early childhood. Despite early onset and severity of anxiety disorders, it is rare that young children with anxiety problems receive appropriate care and treatment [1, 5]. One of the main barriers is the lack of reliable and valid instruments for assessing preschool anxiety. Thus, psychometrically sound instruments for identifying anxiety symptoms in early childhood are urgently needed. Currently, several measures such as the Child Behavior Checklist [12], the Strengths and Difficulties Questionnaire [13], and the Children’s Moods, Fears and Worries Questionnaire [6] are often used for assessing anxiety of preschool children. Although these measures possess adequate psychometric properties and provide valuable information on psychopathological symptoms of preschool children, they do not identify the specific anxiety symptoms. Instead, they tend to address more global aspects of internalizing behaviors, such as
123
Child Psychiatry Hum Dev
anxiety, depression, and withdrawal. Previous studies suggest that the patterns of anxiety symptoms can already be sub-typed in the preschool period [1, 14]. Therefore, there seems to be a need for symptom-specific and ageappropriate instruments to assess anxiety problems in preschool children. To the best of our knowledge, the Preschool Anxiety Scale (PAS) is the only measure that specifically assesses multiple anxiety symptoms in preschool-aged children [14]. The PAS is a 28-item parent-report measure that was designed to assess anxiety dimensions specified in the DSM-IV. The factor structure and construct validity of the PAS were examined in a large Australian community sample [14]. Confirmatory factor analysis revealed five factors: separation anxiety disorder, physical injury fears, social phobia, obsessive–compulsive disorder, and generalized anxiety disorder. Construct validity of the PAS was also supported by significant correlations with internalizing factor of the CBCL to a greater extent than with externalizing factor of the CBCL. Notably, several items relating to obsessive–compulsive disorder in the PAS were reported infrequently by Australian parents. Specifically, only 1.84 % of mothers and .76 % of fathers rated these items as ‘‘quite often true’’ or ‘‘very often true’’. Therefore, in a subsequent study, Edwards et al. [11] slightly modified this measure with another Australian sample, resulting in a revised version of the PAS (PAS-R). Psychometric evaluation of the PAS-R indicated that, except for obsessive– compulsive disorder, the total scale and the other subscales displayed satisfactory internal consistency, test–retest reliability (over a period of 12 months) and good discriminate validity. Given the fact that obsessive–compulsive disorder subscale had quite poor psychometrics (mother/father report: a = .46/.41; 12-month reliability = .57/.51) and relatively infrequent endorsement in Australian samples, this subscale was therefore removed from the PAS-R. These findings seem to indicate that the symptom of obsessive–compulsive disorder is not included in the structure of anxiety among Australian preschoolers. However, there is evidence that obsessive–compulsive disorder represented an independent factor among Dutch preschoolers. Specifically, Broeren and Muris [15] investigated the applicability of the original PAS in Dutch children aged 2–6 years. Exploratory factor analysis revealed five factors that are broadly consistent with the structure of the original PAS, and all the five factors had acceptable internal consistency and construct validity. The aforementioned findings demonstrated that the structure of anxiety symptoms in very young children may differ across cultures. To date, the psychometric properties of the PAS have been examined predominately in Australian and Dutch community samples. Given the reported variation in the
123
structure of anxiety symptoms in different countries, the current study intends to examine the structure of anxiety symptoms in Chinese preschool children. A Chinese translation of the PAS has already been used in at least one study [16]. In that study, the items relating to obsessive– compulsive disorders in the original PAS were reported frequently by the mothers of Chinese preschoolers, from which we can infer that obsessive–compulsive disorder may exist as a distinct factor in Chinese preschool children. Based on the above mentioned reasons, this study will explore the structure of anxiety in Chinese preschoolers using the original PAS containing the symptom of obsessive–compulsive disorder. Meanwhile, this study also evaluates the factorial invariance of the structure of anxiety symptoms in different age (3- and 4-year-olds and 5- and 6year-olds) and gender (males and females) groups, since a wide range of evidence seems to indicate that children’s anxiety symptoms vary significantly between these groups [17, 18]. Additionally, considering the differences in socialization practices between Chinese and Western cultures, the anxiety levels of Chinese children may be different from those of Western peers. For example, in a study using CBCL to examine preschoolers’ behavioral problems, Liu, Cheng, and Leung [19] found that Chinese preschool children scored higher on internalizing problems than their American counterparts. Therefore, the other aim of this study is to compare anxiety symptoms of the current Chinese community sample with those found in previous studies of the PAS in other cultures. In summary, the primary purpose of this study was to examine anxiety disorder symptoms of Chinese preschool children in several ways: (1) to examine the factor structure of anxiety symptoms in Chinese preschool children; (2) to investigate age and gender differences and the prevalence of anxiety symptoms in Chinese preschool children; (3) to compare anxiety symptoms of Chinese preschool children to those of Australian/Dutch counterparts.
Methods Participants Participants consisted of 1,854 Chinese young children aged 3–6 years (M = 4.93 years, SD = .95) and their mothers (M = 33.05 years, SD = 2.62). They were recruited from six public kindergartens in the metropolitan areas of Jinan and Linyi, located in Shandong Province, Eastern China. The kindergartens were selected by a clustered random sampling method. A total of 988 boys and 866 girls from 63 classes and their mothers participated in the present study. The majority of mothers had a high school (20 %) or a college/university education (74 %).
Child Psychiatry Hum Dev
Procedure The design of this study was approved by the Institutional Review Board at Shandong Normal University. Permissions to administer questionnaires from the kindergarten principals and maternal consents were obtained prior to data collection. Mothers were invited to participate in a meeting at the kindergarten. Approximately 90 % of mothers attended the meeting. During the meeting, they completed the Chinese version of the PAS and a questionnaire on family demographics. Those mothers who were absent from the meeting on the day of testing (e.g., due to illness or working) received a packet containing questionnaires and instructions from their child’s teacher. They were asked to fill it out at home and return the completed form to their child’s teacher. To examine the test–retest reliability of the original PAS in Chinese culture, a subgroup of mothers (n = 217) were retested over a period of one month after the initial screening. To further examine the convergent and divergent validity of the measure, a subgroup of mothers (n = 1,076) were also asked to complete the Children Behavior Checklist [12] at the first screening. Measures Preschool Anxiety Scale (PAS) As mentioned previously, the PAS is a 28-item parentreport measure of anxiety symptoms for preschool children [14]. The PAS consists of five subscales: separation anxiety disorder (five items), physical injury fears (seven items), social phobia (six items), obsessive–compulsive disorder (five items) and generalized anxiety disorder (five items). Mothers were asked to rate the items of each subscale on a five-point scale ranging from 0 (not at all true) to 4 (very often true). The scores of total scale and each subscale can be calculated by adding the responses of the relevant items. The English version of the PAS was translated into Chinese by the authors, and then back-translated to the original language by a bilingual translator. It was assured that the content of the translated Chinese version was similar to the original English version by the translators. A study with Chinese preschool children suggested that the item ‘‘Is nervous of going swimming’’ was most frequently missing [16]. Thus, a new item ‘‘Is nervous of going across the road’’ was added considering the growing traffic problems and parents’ emphases on children’s security in China. Frequency analyses revealed that, in the present study, the item ‘‘Is nervous of going swimming’’ and the new added item ‘‘Is nervous of going across the road’’ had 10.21 and 1.7 % missing values, respectively. Following the criterion employed by Broeren and Muris
[15], a threshold of 5 % missing values was used to remove unsatisfactory items in this study. The item ‘‘Is nervous of going swimming’’ was therefore excluded from the scale and replaced by the new added item ‘‘Is nervous of going across the road’’ in this study. Children Behavior Checklist (CBCL) The CBCL designed for 4–18 year old children [12] was selected as an indicator of convergent and divergent validity of the PAS. It is a 118-item parent-report measure to assess two broad band factors, Internalizing and Externalizing. Mothers were asked to evaluate on a three-point scale (0 = not true, 1 = somewhat true, 2 = very true) regarding whether the behavior is occurring now or has occurred over the past 6 months. The CBCL was validated in China and demonstrated adequate reliability and validity [20]. In the current sample, Cronbach’s alpha coefficients were .92 for the total scale, .86 for Internalizing factor and .87 for Externalizing factor. Data Analysis To examine the factor structure of anxiety symptoms in Chinese preschool children, CFA was performed using Robust Maximum Likelihood estimation in LISREL 8.70. Since in large sample sizes the v2 statistic is likely to be significant, normed fit index (NFI), non-normed fit index (NNFI), comparative fit index (CFI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR) were calculated in the present study to evaluate the fit of the model. Values of NFI, NNFI and CFI greater than .90 indicate an acceptable fit [21]. RMSEA value below .06 and SRMR value below .08 indicate a relatively good fit [22]. To further examine whether the factor structure of anxiety symptoms is invariant across age and gender groups, we conducted two multi-group confirmatory factor analyses. In each multigroup analysis, a model in which all factor loadings were constrained to be equal (constrained model) across age or gender groups was compared to a model in which these factor loadings were free to vary (unconstrained model) across groups. Following the criterion for evaluating factorial invariance across subgroups described by Cheung and Rensvold [23], the differences in goodness-of-fit indices (DGFIs) between constrained and unconstrained models less or equal to .01 are considered reasonably invariant. Additionally, Cronbach’s alpha coefficients were calculated to evaluate the internal consistency of the PAS total and subscales. Pearson product moment correlations were utilized to assess test–retest reliability and convergent and divergent validity. Age and gender differences in anxiety
123
Child Psychiatry Hum Dev
scores were examined using multivariate analyses of variance (MANOVA) and effect size statistics. The prevalence of anxiety symptoms was evaluated through frequency statistics. Missing values were estimated in SPSS using the estimated means (EM) procedure.
Results Confirmatory Factor Analysis The correlated five-factor model reported by Spence et al. [14] in the original PAS was tested in current Chinese preschool populations. The results demonstrated that all 28 items had loadings in excess of .31 on their hypothesized factors (see Table 1) and that the values for fit indices fell within the acceptable to good ranges (NFI = .93, NNFI = .94, CFI = .94, RMSEA = .057, SRMR = .056), indicating a satisfactory fit of the five-factor model to the present data. The five factors were found to be strongly intercorrelated, with rs ranging from .52 to .87. Multi-group confirmatory factor analyses were conducted to test the factorial invariance of this model across demographic groups. We first tested whether the model is invariant between younger (3- and 4-year-olds) and older (5- and 6- year-olds) children. Table 2 shows that the model in which the factor loadings were constrained to be equal between younger and older children provided a good fit, very similar to that of the model in which the factor loadings were free to vary across groups. The differences in GFIs below .01 suggested factorial invariance across age groups. After combining the data of younger and older children, we then tested the factorial invariance of the model between boys and girls. Results from model comparisons showed that the differences in GFIs were less than .01 between unconstrained and constrained models (see table 2), which supported the factorial invariance of the model across gender groups. Internal Consistency and Test–Retest Reliability Cronbach’s alpha coefficients were calculated to evaluate the internal consistency of the PAS in Chinese preschool children. Coefficient a values were .87 for the total scale, .55 for separation anxiety disorder, .72 for physical injury fears, .75 for social phobia, .68 for obsessive–compulsive disorder, and .70 for generalized anxiety disorder. To examine test–retest reliability of the Chinese PAS, a total of 217 participants (M = 5.11 years, SD = 1.01; 110 boys and 107 girls) were reassessed one month later. Results demonstrated that the Pearson correlation coefficients (r) were .73 for the total scale, .58 for separation anxiety disorder, .71 for physical injury fears, .64 for social
123
Table 1 Factor loadings of the five-factor model after confirmatory factor analysis Item
SAD
6. Is reluctant to go to sleep without you or to sleep away from home
.31
12. Worries that something bad will happen to his/her parents
.56
16. Worries that something bad might happen to him/her (e.g., getting lost or kidnapped), so he/ she won’t be able to see you again
.58
22. Becomes distressed about your leaving him/her at preschool/ school or with a babysitter
.42
25. Has nightmares about being apart from you
.48
PIF
7. Is scared of heights (high places)
.46
10. Is afraid of crowded or closedin places 13. Is scared of thunder storms
.58
17. Is nervous of going across the road
.53
20. Is afraid of insects and/or spiders
.48
SP
OCD
.56
24. Is frightened of dogs
.44
26. Is afraid of the dark
.55
2. Worries that he/she will do something to look stupid in front of other people
.60
5. Is scared to ask an adult for help (e.g., a preschool teacher)
.56
11. Is afraid of meeting or talking to unfamiliar people
.58
15. Is afraid of talking in front of the class (e.g., show and tell)
.59
19. Worries that he/she will do something embarrassing in front of other people
.62
23. Is afraid to go up to group of children and join their activities
.52
3. Keeps checking that he/she has done things right (e.g., that he/ she closed a door, turned off a tap)
.56
9. Washes his/her hands over and over many times each day
.50
18. Has to have things in exactly the right order or position to stop bad things from happening
.54
21. Has bad or silly thoughts or images that keep coming back over and over
.56
27. Has to keep thinking special thoughts (e.g., numbers or words) to stop bad things from happening
.59
GAD
Child Psychiatry Hum Dev
Age and Gender Differences
Table 1 continued Item
SAD
PIF
SP
OCD
GAD
1. Has difficulty stopping him/ herself from worrying
.53
4. Is tense, restless or irritable due to worrying
.66
8. Has trouble sleeping due to worrying
.56
14. Spends a large part of each day worrying about various things
.58
28. Asks for reassurance when it does not seem necessary
.53
SAD separation anxiety disorder, PIF physical injury fears, SP social phobia, OCD obsessive–compulsive disorder, GAD generalized anxiety disorder
A MANOVA on the total score and subscales of the PAS was conducted to examine age and gender differences of anxiety symptoms in preschool children. Based on the Wilks’ lambda criterion, results of the MANOVA indicated that the combined dependent variables were significantly different between younger (3- and 4-year-olds) and older (5- and 6-year-olds) groups [F(5,1846) = 7.81, p \ .001, g2 = .021], but not for gender groups [F(5,1846) = 1.52, p [ .05, g2 = .004] or gender by age interactions [F(5,1846) = .15, p [ .05, g2 = .000]. Specifically, younger children were found to score significantly higher on the total score and all the subscales of the PAS as compared to the older children (see table 3). Prevalence of Anxiety Symptoms
phobia, .63 for obsessive–compulsive disorder, and .59 for generalized anxiety disorder. Convergent and Divergent Validity To further explore the convergent and divergent validity of the Chinese PAS, correlations between the PAS and CBCL subscales were calculated in a subgroup of 1,076 participants. The total score and subscales of the PAS correlated significantly with the CBCL Internalizing score (r = .31 to .59) and with the CBCL Externalizing score (r = .21 to .40). Following the recommendation of Meng et al. [24] for significant correlations, a series of Z tests were conducted to determine whether correlations between PAS scales and Internalizing score (convergent relationship) are significantly stronger than correlations between PAS scales and Externalizing score (divergent relationship). The results indicated that the correlations with the Internalizing score are significantly greater than the correlations with the Externalizing score for all scales of the PAS, Z-scores ranged from 3.48 to 10.91, ps \ .001. Overall, this finding supported the convergent and divergent validity of the PAS.
Following previous studies [14, 25], the most prevalent symptoms of Chinese preschool children were examined by calculating the percentage of mothers who rated each item as either 3 (quite often true) or 4 (very often true) for their children. As shown in Table 4, the most prevalent symptoms in Chinese preschool children were related to separation anxiety (‘‘Is reluctant to go to sleep without you or to sleep away from home’’), physical injury fears (dark, dogs, insects and/or spiders, thunder storms, crowded or closedin places) and social phobia (‘‘Is afraid of talking in front of the class or preschool group’’). Additionally, three items (‘‘Has things in exactly the right position’’, ‘‘Washes his/ her hands over and over many times each day’’, and ‘‘Keeps checking that he/she has done things right’’) relating to obsessive–compulsive disorder were also among the top 10 presenting problems. Comparisons to Previous Studies To investigate whether the means of anxiety symptoms experienced by the current Chinese community sample differed from those found in previous studies of the PAS
Table 2 Tests of factorial invariance across age (younger and older) and gender (boys and girls) groups v2
df
p
NFI
NNFI
CFI
RMSEA
SRMR
Unconstrained model
2,758.94
680
\.001
.922
.933
.940
.057
.058
Constrained model
2,808.56
703
\.001
.920
.934
.939
.057
.058
Model Invariance across age groups
Invariance across gender groups Unconstrained model
2,791.22
680
\.001
.921
.932
.939
.058
.057
Constrained model
2,826.64
703
\.001
.920
.934
.939
.058
.059
NFI normed fit index, NNFI non-normed fit index, CFI non-normed fit index, RMSEA root mean square error of approximation, SRMR standardized root mean square residual
123
123
Discussion
* p \ .05; ** p \ .01; *** p \ .001
SAD separation anxiety disorder, PIF physical injury fears, SP social phobia, OCD obsessive–compulsive disorder, GAD generalized anxiety disorder
1.43 ± 2.19
2.95 ± 2.91 18.84 ± 11.57
with Australian [14] and Dutch community samples [15], a series of t tests were conducted. As presented in Table 3, Chinese preschool children in the current sample were found to score significantly higher on the total scale and subscales of the PAS than Australian [ts(1,853) [ 8.70, ps \ .001] and Dutch samples [ts(1,853) [ 19.43, ps \ .001] in previous studies, with exception of the social phobia subscale on which no significant difference was found between Chinese preschool children and their Dutch peers (t = .21, p [ .05). To compare the most prevalent anxiety symptoms of Chinese preschoolers with those found in prior research of the PAS, Table 4 also presents the results reported by Spence et al. [14] in Australian community samples. The contents of the most prevalent anxiety symptoms in Chinese preschool children and Australian samples showed substantial differences, although they also displayed many similarities. Specifically, three items referring to obsessive–compulsive disorder mentioned above, which were rare in Australian preschoolers, were evidently reported frequently in Chinese preschool children.
2.15 ± 2.57 17.28 ± 11.83
1.23 ± 2.08 3.98 ± 3.21 .58
.039 -.02
.65 .19
-.034 4.04 ± 3.24
4.21 ± 2.75 27.62 ± 13.00 6.26* 18.86***
5.54* .109
.115 .201
3.81 ± 3.16 4.16 ± 3.25
4.33 ± 2.92 29.08 ± 13.48
OCD
GAD Total
4.00 ± 2.78 26.44 ± 12.83
3.93 ± 3.19
4.10 ± 2.95 27.87 ± 13.46
4.16 ± 2.85 27.73 ± 13.21
6.69 ± 4.11
5.43 ± 4.22 4.66 ± 3.74
6.51 ± 4.37 8.91 ± 4.91
.07 .013
5.45 ± 3.90
3.01 -.082 9.12 ± 4.99 8.72 ± 4.83
5.47 ± 3.89 11.09**
6.03* .114
.152 5.16 ± 3.75
9.19 ± 4.90
5.75 ± 4.04
PIF
SP
8.63 ± 4.91
5.42 ± 3.92
.45 .031 5.19 ± 2.91 5.28 ± 2.93 36.55*** .28 4.84 ± 2.71 5.65 ± 3.07
Cohen’s d 5–6 years
SAD
Cohen’s d Girls Boys 3–4 years
F
Total by gender Total by age
Table 3 Means and standard deviations of the PAS (mean ± SD) for current and previous community samples [14, 15]
F
5.24 ± 2.92
Current sample
2.73 ± 2.85
Australian sample
2.34 ± 2.41
Dutch sample
Child Psychiatry Hum Dev
Although a large body of research has consistently shown that anxiety disorders have an early onset in childhood [1, 4, 14], little research has focused on the anxiety problems of very young children. The present study contributes to the existing literature by examining the nature of anxiety symptoms in Chinese preschool children. Confirmatory factor analysis indicated that the correlated five-factor model (including separation anxiety disorder, physical injury fears, social phobia, obsessive– compulsive disorder, and generalized anxiety disorder) in the original PAS fit the current data well. Further multigroup analyses supported the factorial invariance of this model across age and gender groups in Chinese preschool populations. These findings demonstrated that the anxiety symptoms of Chinese preschoolers reflected a five-factor structure, and that both younger and older children, and both boys and girls presented the same pattern of anxiety symptoms. Notably, the items representing obsessive– compulsive disorder comprised an independent factor in Chinese preschool children, which is inconsistent with the study by Edwards and colleagues [11] suggesting that obsessive–compulsive disorder did not comprise an independent factor in Australian young children. The reliability and validity of the Chinese PAS provided empirical support for the five-factor structure of anxiety (particularly for obsessive–compulsive disorder as an independent factor of anxiety) in Chinese preschool populations. Specifically, the current study found that, except
Child Psychiatry Hum Dev Table 4 The percentage of children receiving mother ratings of 3 or 4 (quite often true or very often true) for current and previous Australian samples [14] Item
Table 4 continued Item
Current sample (%)
Australian sample (%)
14. Spends a large part of each day worrying about various things
1.6
.4
Current sample (%)
Australian sample (%)
6. Is reluctant to go to sleep without you or to sleep away from home
44
12.7
21. Has bad or silly thoughts or images that keep coming back over and over
1.1
1.2
26. Is afraid of the dark
26
17.8
1.2
20.7
17.8
25. Has nightmares about being apart from you
.9
24. Is frightened of dogs 20. Is afraid of insects and/or spiders
16.8
10.6
18. Has to have things in exactly the right order or position to stop bad things from happening
14.3
1.6
9. Washes his/her hands over and over many times each day
14.1
1.4
13. Is scared of thunder storms
12.7
13.2
10. Is afraid of crowded or closed-in places
12.2
1.8
15. Is afraid of talking in front of the class (preschool group) e.g., show and tell
11.6
5.9
3. Keeps checking that he/she has done things right (e.g., that he/she closed a door, turned off a tap)
9.9
4.0
7. Is scared of heights (high places)
9.0
4.0
11. Is afraid of meeting or talking to unfamiliar people
8.8
9.8
5. Is scared to ask an adult for help (e.g., a preschool or school teacher)
7.1
2.4
4. Is tense, restless or irritable due to worrying
6.7
1.2
28. Asks for reassurance when it doesn’t seem necessary
6.1
2.4
19. Worries that he/she will do something embarrassing in front of other people
5.6
1.0
17. Is nervous of going across the road
5.3
–
22. Becomes distressed about your leaving him/her at preschool/school or with a babysitter
5.1
4.5
1. Has difficulty stopping him/herself from worrying
5.1
3.9
12. Worries that something bad will happen to his/her parents
4.7
2.8
16. Worries that something bad might happen to him/her (e.g., getting lost or kidnapped), so he/she won’t be able to see you again
4.6
1.4
2. Worries that he/she will do something to look stupid in front of other people 23. Is afraid to go up to group of children and join their activities
4.4
4.3
3.3
7.7
27. Has to keep thinking special thoughts (e.g., numbers) to stop bad things from happening
2.0
1.0
8. Has trouble sleeping due to worrying
1.9
1.6
for separation anxiety subscale having an a of .55, the PAS total scale and subscales displayed satisfactory levels of internal consistency reliability, test–retest reliability and convergent and divergent validity. These results supported the applicability of the five-factor structure of PAS in Chinese preschool populations. Inconsistent with the study by Edwards and colleagues suggesting that the obsessive– compulsive disorder subscale had quite poor psychometrics [11], this subscale had acceptable internal consistency and test–retest reliability in the present study, which suggested that obsessive–compulsive disorder was an important section of the five-factor structure of anxiety in very young Chinese children. The prevalence of anxiety symptoms provided further support for obsessive–compulsive disorder as an independent factor in the structure of anxiety for Chinese preschoolers. In specific, comparisons of the prevalent PAS items (rated as ‘‘quite often true’’ or ‘‘very often true’’) indicated that there were substantial differences regarding the contents of the prevalent anxiety symptoms among Chinese and Australian preschoolers. As mentioned above, three items (‘‘Has things in exactly the right position’’, ‘‘Washes his/her hands over and over many times each day’’, and ‘‘Keeps checking that he/she has done things right’’) representing obsessive–compulsive disorder, which were rare in Australian preschool children [14], were reported frequently by the mothers of Chinese preschoolers. Even though it remains unclear why these symptoms are relatively prevalent among Chinese preschool children, this finding is consistent with the notion that cultural factors are involved in the manifestation of children’s anxiety symptoms [26]. Parents’ over-controlling practices and teachers’ rigorous instructional styles in Chinese societies may be related to the high prevalence of these symptoms. For instance, in China, very specific instructions (e.g., ‘‘Has toys in exactly the right order or position’’) are commonly given at home and in the kindergartens. This interpretation requires further exploration. Consistent with the study by Spence et al. [14] demonstrating age differences in preschoolers’ anxiety symptoms, our data showed that younger children (3- and
123
Child Psychiatry Hum Dev
4-year-olds) were reported to have higher anxiety levels in comparison to older children (5- and 6-year-olds). It is hypothesized that the high level of anxiety symptoms in younger children may be related to their life transition. Most Chinese children enter into kindergarten and start a new period in their life at age 3–4 years. They may not only suffer from the fear of leaving their parents and home but also experience difficulties adjusting to kindergarten. Thus, the elevated levels of anxiety symptoms in younger children may be a response to this transition. Consistent with earlier studies relating to preschoolers [1, 5, 14], no significant gender differences in anxiety scores were found in the present study. In contrast, much of the research with school-age children found that girls report higher anxiety scores than boys [27, 28]. Therefore, it seems that gender differences in anxiety symptoms do not emerge in preschool period, and may become more distinct with increasing age. Comparisons to previous studies revealed that, with exception of social phobia, Chinese preschool children displayed higher anxiety levels than their Australian and Dutch peers in previous studies [14, 15]. These crossnational differences in anxiety levels for younger children are comparable to those reported for older children and adolescents. For instances, Chinese school children reported higher degree of anxiety than their American counterparts [29]. Chinese adolescents reported higher anxiety levels than their Dutch and German peers [30]. These discrepancies may be accounted for by different socialization practices in Chinese and Western cultures. It has been speculated that Chinese socialization practices stress self-control, emotional restraint, and obedience to authority [31, 32], which may contribute to children’s internalizing issues such as anxiety. Western socialization practices, by contrast, encourage individualism and independence [33], which in turn may contribute to the lowering of anxiety problems. From a cultural context, the manifestations of anxiety in Chinese children may be adaptive and may represent ‘‘normal’’ reactions to dominant cultural forces [19]. In this vein, a higher level of anxiety may not necessarily reflect the presence of anxiety pathology in Chinese preschool children. It should be acknowledged that the present study suffers from several limitations. First, only community sample was included in this study, so it is not clear whether our findings can be generalized to clinical populations. Second, the current study only recruited the samples from the Shandong region. It remains unknown whether our results can be applied to preschool children in other Chinese regions. Further research needs to recruit some representative samples from different regions in China to replicate these results. Third, the data of this study were based solely on mother-report. Although parents as informants of young
123
children’s anxiety are satisfactory [14, 15], future studies still need to include direct behavioral observations or anxiety ratings by teachers to confirm this measure. Finally, the relatively low alpha for separation anxiety disorder subscale (a = .55) in the current sample must be acknowledged, although it is consistent with the result (a = .59) in the previous study [15]. It may be due to the low number of items for this subscale (n = 5) and the fact that this subscale includes individual items which intend to assess children’s separation anxiety but are only loosely related to each other. This seems to indicate that the individual items of this subscale may have greater utility than the aggregate [34].
Summary The current study makes significant inroads to understanding anxiety problems in Chinese preschool children. This study found that the structure of anxiety in Chinese preschool children included five factors, and this five-factor structure applied to different age and gender groups in Chinese preschool populations. Inconsistent with prior research [11], obsessive–compulsive disorder had satisfactory psychometrics and high endorsement, and can be viewed as an independent anxiety factor in Chinese preschool children. Compared to other studies with Western samples, anxiety symptoms of Chinese preschool children were found at a high level. More studies are needed to explore cultural factors that account for obsessive–compulsive disorder as an independent factor and the higher rates of anxiety in Chinese young children.
References 1. Egger HL, Angold A (2006) Common emotional and behavioral disorders in preschool children: presentation, nosology, and epidemiology. J Child Psychol Psychiatry 47:313–337 2. Eley TC, Bolton D, O’Connor TG, Perrin S, Smith P, Plomin R (2003) A twin study of anxiety-related behaviors in preschool children. J Child Psychol Psychiatry 44:945–960 3. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE (2005) Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in National Comorbidity Survey Replication. Arch Gen Psychiatry 62:593–602 4. Rapee RM, Schniering CA, Hudson JL (2009) Anxiety disorders during childhood and adolescence: origins and treatment. Annu Rev Clin Psychol 5:311–341 5. Furniss T, Beyer T, Guggenmos J (2006) Prevalence of behavioral and emotional problems among six-years-old preschool children. Soc Psychiatry Psychiatr Epidemiol 41:394–399 6. Bayer JK, Sanson AV, Hemphill SA (2006) Children’s moods, fears, and worries: development of an Early Childhood Parent Questionnaire. J Emot Behav Disord 14:41–49
Child Psychiatry Hum Dev 7. Warren SL, Huston L, Egeland B, Sroufe LA (1997) Child and adolescent anxiety disorders and early attachment. J Am Acad Child Adolesc Psychiatry 36:637–644 8. Cole DA, Peeke LG, Martin JM, Truglio R, Seroczynski AD (1998) A longitudinal look at the relation between depression and anxiety in children and adolescents. J Consult Clin Psychol 66:451–460 9. Kovacs M, Devlin B (1998) Internalizing disorders in childhood. J Child Psychol Psychiatry 39:47–63 10. Zahn-Waxler C, Klimes-Dougan B, Slattery MJ (2000) Internalizing problems of childhood and adolescence: prospects, pitfalls, and progress in understanding the development of anxiety and depression. Dev Psychopathol 12:443–466 11. Edwards SL, Rapee RM, Kennedy SJ, Spence SH (2010) The assessment of anxiety symptoms in preschool-aged children: the revised Preschool Anxiety Scale. J Clin Child Adolesc Psychol 39:400–409 12. Achenbach TM (1991) Manual for the Child Behavior Checklist/ 4–18 and 1991 profile. University of Vermont, Department of Psychiatry, Burlington 13. Goodman R (1997) The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry 38:581–586 14. Spence SH, Rapee R, McDonald C, Ingram M (2001) The structure of anxiety symptoms among preschoolers. Behav Res Ther 39:1293–1316 15. Broeren S, Muris P (2008) Psychometric evaluation of two new parent-rating scales for measuring anxiety symptoms in young Dutch children. J Anxiety Disord 22:949–958 16. Zhao JX, Wang MF (2009) Developmental characteristics of preschoolers’ anxiety. Chin J Clin Psychol 17:723–725 17. Crocetti E, Hale WW, Fermani A, Raaijmakers Q, Meeus W (2009) Psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED) in the general Italian adolescent population: a validation and a comparison between Italian and the Netherlands. J Anxiety Disord 23:824–829 18. Hale WW, Crocetti E, Raaijmakers QAW, Meeus WHJ (2011) A meta-analysis of the cross-cultural psychometric properties of the Screen for Child Anxiety Related Emotional Disorders (SCARED). J Child Psychol Psychiatry 52:80–90 19. Liu JH, Cheng H, Leung PWL (2011) The application of the Preschool Child Behavior Checklist and the Caregiver–Teacher report form to Mainland Chinese children: syndrome structure, gender differences, country effects, and inter-informant agreement. J Abnorm Child Psychol 39:251–264 20. Su LY, Li XR, Yang ZW, Luo XR, Wan GB (1997) The application of Child Behavior Checklist and Teachers’ Report Form in epidemiological study of child mental health: choosing an optimal cut-off point. Chin Mental Health J 11:37–40
21. Marsh HW, Hau KT (1996) Assessing goodness of fit: is parsimony always desirable? J Exp Educ 64:364–390 22. Hu L, Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model 6:1–55 23. Cheung GW, Rensvold RB (2002) Evaluating goodness-of-fit indexes for testing measurement invariance. Struct Equ Model 9:233–255 24. Meng XL, Rosenthal R, Rubin DB (1992) Comparing correlated correlation coefficients. Psychol Bull 111:172–175 25. Ishikawa S, Shimotsu S, Ono T, Sasagawa S, Kondo-Ikemura K, Sakano Y, Spence SH (2013) A Parental Report of Children’s Anxiety Symptoms in Japan. Child Psychiatry Hum Dev. doi:10. 1007/s10578-013-0401-y 26. Muris P, Schmidt H, Engelbrecht P, Perold M (2002) DSM-IVdefined anxiety disorder symptoms in South African children. J Am Acad Child Adolesc Psychiatry 41:1360–1368 27. Ishikawa SI, Sato H, Sasagawa S (2009) Anxiety disorder symptoms in Japanese children and adolescents. J Anxiety Disord 23:104–111 28. Su LY, Wang K, Fan F, Su Y, Gao XP (2008) Reliability and validity of the Screen for Child Anxiety Related Emotional Disorders (SCARED) in Chinese children. J Anxiety Disord 22:612–621 29. Weine AM, Phillips JS, Achenbach TM (1995) Behavioral and emotional problems among Chinese and American children: parent and teacher reports for ages 6 to 13. J Abnorm Child Psychol 23:619–639 30. Zhao JX, Xing XP, Wang MF (2012) Psychometric properties of the Spence Children’s Anxiety Scale (SCAS) in Mainland Chinese children and adolescents. J Anxiety Disord 26:728–736 31. Chen X, Rubin K, Li B (1997) Maternal acceptance and social and school adjustment in Chinese children: a four-year longitudinal study. Merrill Palmer Q 43:663–681 32. Xie D, Leong FTL (2008) A cross-cultural study of anxiety among Chinese and Caucasian American university students. J Multicult Couns Dev 36:52–63 33. Essau CA, Leung PWL, Conradt J, Cheng H, Psych D, Wong T (2008) Anxiety symptoms in Chinese and German adolescents: their relationship with early learning experiences, perfectionism, and learning motivation. Depress Anxiety 25:801–810 34. Whiteside SP, Brown AM (2008) Exploring the utility of the Spence Children’s Anxiety Scales parent- and child-report forms in a North American sample. J Anxiety Disord 22:1440–1446
123
