Development and validation of a depression scale for Asian adolescents
Abstract
Items covering both core and culture-specific facets of depression were generated based on literature review and clinical experience. They were modified following focus group discussions with depressed adolescents and adolescents in the community. The newly constructed Asian Adolescent Depression Scale (AADS) was administered to a clinical and a community sample of adolescents together with other rating scales. The AADS comprised 4 factors (negative self-evaluation, negative affect, cognitive inefficiency and lack of motivation) and demonstrated sound psychometric properties. Negative socially oriented self-evaluation and cognitive inefficiency were important in Singaporean adolescents’ conceptualization of depression and are likely to be Asian culture-specific dimensions.
Introduction
Depression is recognized as a significant health issue for adolescents due to its relatively high prevalence in the general population and the fact that mood disorders are associated with substantial morbidity and mortality due to suicide (Kashani & McNaul, 1997). Studies on community samples of adolescents show point prevalence rates of between 2.9% and 8% (Kashani et al., 1987; Lewinsohn, et al., 1993a; Birmaher et al., 1996; Rey et al., 2001). Cohort studies also show that rates of depression are increasing in younger people in the West (Ryan et al., 1992; Lewinsohn et al., 1993b; Kovacs & Gatsonis, 1994). The incidence of depression markedly increases during adolescence, with girls outnumbering boys by a ratio of about 2:1 (Angold et al., 2002; Hoffman et al., 2003).
Depression in adolescents can be described in 8 dimensions (DSM-IV, American Psychiatric Association, 1994): (1) affective manifestations like depressed or irritable mood, (2) loss of interest, (3) cognitive dysfunction including inability to think or concentrate, (4) vegetative manifestations like insomnia, loss of appetite or loss of weight, (5) psychomotor manifestations like agitation or retardation, (6) negative cognitions like feelings of worthlessness or inappropriate guilt, (7) existential concerns like recurrent suicidal thoughts and (8) fatigue or lack of energy.
Currently, clinicians and researchers in Singapore use measures of depression for adolescents that were constructed and standardized in the West, such as the Children Depression Inventory (CDI) (Kovacs, 1981), which covers an age range of 7–17, the Centre for Epidemiological Studies Depression Scale (Radloff, 1977) and the Mood and Feelings Questionnaire (Costello & Angold, 1988). Specifically, our issues of concern are: (1) whether Western tools are comprehensive enough to cover the range of symptoms manifested by depressed Asian adolescents, (2) whether the items in Western measures are understood by Asian adolescents with the same intended meaning and (3) whether there are any culture-specific symptoms that are not included in Western instruments.
Depending on the values and belief systems of a culture, the goals to which an individual directs his motivational efforts may differ (Markus & Kitayama, 1991). Thus, depressive symptoms manifested in different cultural contexts may give rise to different phenomenological experiences. Asians live in a collective way of life where others in the social context are seen as an integrative part of the self-construct (Triandis, 1989). Hence, their psychological state is greatly influenced by the social and interpersonal processes around them (Markus & Kitayama, 1991). Perceptions of a lack of parental understanding and peer acceptance, both related to interpersonal factors, were also found to be the strongest predictors of depressed mood in a community sample of Chinese adolescents in Hong Kong (Stewart et al., 1999). In this study, we postulated that the interpersonal meaning of depression, which is absent from most scales constructed in the West, is a potential culturally sensitive dimension to the depressive symptomatology of Asians. Additionally, it has been found that Western populations also tend to report symptoms in affective terms, while Asians tend to report symptoms in somatic terms, when a traditional Western constructed instrument, the Zung Self-Rating Depression Scale (Zung, Richards, & Short, 1965) was used (Chang, 1985). This might be due to the fact that the Zung Self-Rating Depression Scale does not contain an interpersonal dimension and the Asian participants could only choose from the available dimensions. However, this cross-cultural study supported the notion that in different cultural environments, the phenomenological experiences of depression differ.
Singapore is a small island city with a resident population of 3.3 million. As a multiracial and multicultural society, it consists of 3 main ethnic Asian communities: Chinese 76.8%, Malays 13.9%, Indians 7.9% and other races 1.4% (Singapore Department of Statistics, 2001). Children and adolescents in Singapore attend primary school between the ages of 6–12, secondary school between the ages of 13–16, and post-secondary education either at a junior college, polytechnic or technical institute from 17 to 19 years of age (Tan, Gopinathan, & Ho, 1997). Since the independence of Singapore from British colonial rule in 1965, English has been the predominant language used in school and at the workplace. However, Singaporean parents educated before the independence of Singapore may not be well versed in English and may speak various Asian languages like Chinese, Malay and Tamil at home. Hence, Singaporean adolescents may use the English language, but their phenomenological experiences of personal and emotional matters may be affected by the mother tongue-mediated home culture. This mixed linguistic usage has resulted in different ways of expression, including direct translations from mother tongues and the use of colloquialisms in their presentation of psychological symptoms.
During adolescence, the individual is facing the challenge of developing his unique identity through individuation and becoming independent of his family (Erikson, 1959). Chou (2000) also found that depressive symptomatology was associated with 2 aspects of emotional autonomy, i.e. individuation and deidealization of parents. In a collectivist Asian society where the emphasis is on integration rather than individuation, we questioned if identity formation would become a challenge to Asian adolescents and lead to depression. On the other hand, since others are of high importance in the self-construct of collectivist Asians, we speculated that concern over self-other relationships might be an important concomitant of depression for Asian adolescents. Adolescents in Asia, especially in the Confucian Heritage Cultural communities such as Japan, Taiwan, Hong Kong and Singapore, also face tremendous pressure for academic achievement. A study comparing stress and coping in Singaporean and American adolescents found that academic pressure was the most frequently experienced stressor for Singaporean adolescents, in contrast to American adolescents, who were more concerned over romantic relationships and individuation from parents (Koh, Koh, & Chang, 2002). We therefore proposed that concerns over academic competence and the cognitive skills related to academic success or failure might be very important for Singaporean adolescents.
Depression is a condition whose manifestations and phenomenological experiences are intimately formed by the culture and the developmental tasks faced by adolescents (Kleinman & Good, 1985). Presentation of symptoms by the adolescents is further influenced by the linguistic skills of the adolescents and the commonly understood usage of terms in the community. As a common human condition, we hypothesized that depression has certain “core symptoms” that can be found in any cultural community, such as lack of motivation and feeling of sadness. We proposed that surrounding these defining symptoms of depression, different culture or linguistic groups may show different phenomenological experiences and symptom presentations. These are “culturally sensitive” symptoms that may be specific to certain cultures. In the present context, we look for “core symptoms” in the overlap between symptoms reported by Singaporean adolescents and adolescents from other countries, and “culturally sensitive” symptoms in the unique findings of Singaporeans.
The objectives of this study are firstly, to examine the manifestations of depression in Singaporean adolescents, and secondly, to develop and validate a depression scale for Asian adolescents, incorporating the above-postulated interpersonal dimension, as well as items relating to the various developmental tasks of adolescence that might impact on mood and self-esteem.
Method
Item generation
We attempted to identify the phenomenological experiences of depression in Singaporean adolescents by using a combination of inductive and deductive approaches. This entailed a qualitative stage involving clinical observation and focus group discussions, followed by a quantitative stage using standard psychometric techniques to construct and validate a depression scale. This served to identify items of depression that were meaningful to Singaporean adolescents and ensure the psychometric vigor of the resulting scale.
Through literature search and clinical observation, we identified several categories of depressive symptoms in Singaporean adolescents. We then generated descriptive statements to serve as initial items to be verified and modified by the adolescents themselves. Items from the 8 commonly described dimensions of depression were adopted with modification, and items relating to the developmental tasks of adolescence such as individuation (for example, “I feel confused without my parents’ guidance” and “I feel that I can never get out of my parents’ control”), development of self-concept (for example, “I am confused about what kind of person I am” and “I feel lost most of the time and have no direction in life”) and body image concerns (for example, “I hate the way I look”), as well as the interpersonal dimension (for example, “I feel that I can no longer make my parents happy”, “I feel that I am not wanted”, “I feel that nobody cares for me or loves me”, “I feel that I am not as good as others” and “I feel that my friends do not accept me”) and psychosomatic manifestations (for example, “I often get headaches”, “I often get stomachaches” and “my body aches”), were generated based on the theoretical concept of the dimension, giving rise to an initial pool of 71 items spanning across 13 dimensions. The items were worded in the everyday language used by our adolescents and incorporated into a self-report measure with a 5-point response scale, ranging from (1) “Strongly Disagree” to (5) “Strongly Agree”.
Focus group discussions
Nine clinicians from the Child Guidance Clinic of Singapore, comprising 5 psychiatrists, 2 psychologists, a social worker and an occupational therapist, were asked to assess the face validity of the items and provide feedback on their wording. The Child Guidance Clinic is a psychiatric outpatient clinic for children and adolescents below the age of 19. It is part of the Institute of Mental Health of Singapore and is the main provider of child and adolescent mental health services in this country.
Focus group discussions were then held with 25 clinically depressed adolescents aged 13–19 (9 males, 16 females) attending the Child Guidance Clinic and 75 adolescents aged 13–19 (31 males, 44 females) from the community to derive a closer understanding of their experiences and assess their comprehension of the language used in the scale. This served to check the face validity of the scale, as a first step towards establishing the construct validity of the measure.
Five clinical focus groups and 8 community focus groups, each comprising 5–10 adolescents, were conducted. The community sample was randomly selected from 2 secondary schools, a junior college, a polytechnic, 2 churches and 2 community centres. The wide spectrum of facilities from which the sample was drawn ensured the representativeness of the sample. The adolescents were asked about their understanding of depression and invited to share their experiences about the times they felt depressed. They were also required to identify items from the scale which applied to them, highlight the items they did not understand and amend the wording of any of the items to what they felt was more appropriate. They were then asked to respond to the actual scale in order to gather feedback on its presentation format. All the focus group discussions were conducted by 3 of the investigators, comprising a psychiatrist and 2 psychologists.
Initial item elimination
Based on the data obtained from the focus groups, items that were not endorsed by at least one-third of the clinical sample and one-third of the community sample were eliminated. Forty-seven items were thus eliminated, with 24 items remaining, which spanned across 10 originally proposed dimensions.
All items relating to individuation, body image concerns and psychosomatic manifestations had to be eliminated due to low endorsement. The clinical and community focus groups did not differ in terms of the types of items endorsed.
Validation in the community
Participants. Two hundred and forty-two adolescents aged 13–19 were recruited from an average secondary school (n=162, 81 males, 81 females), an average junior college (n=20, 10 males, 10 females), a polytechnic (n=40, 20 males, 20 females) and a technical institute (n=20, 10 males, 10 females). The proportion of students from the various types of educational institutions was representative of the Singapore population. Consent was obtained from the principals of the institutions and the students were randomly sampled in each institution to give an equal number of each gender. The ethnic composition of the adolescents corresponded to that of the Singapore population, with 76.9% Chinese, 14.9% Malays, 2.9% Indians and 5.4% other races. The mean age of the males was 16.25±2.29 years and the mean age of the females was 16.07±2.51 years. All the students who were sampled agreed to participate in the study.
Materials. The 24 items generated in the previous phase were incorporated into a scale, the Asian Adolescent Depression Scale (AADS), a self-report measure with a 5-point response scale ranging from (1) “Strongly Disagree” to (5) “Strongly Agree”, targeting symptoms in the preceding 2 weeks. The AADS was administered to the adolescents together with the CDI (Kovacs, 1981), a widely used and validated 27-item self-rated measure with sound psychometric properties that assesses depressive symptoms in children and adolescents. Each item on the CDI comprises 3 statements ranging from (1) “absence of symptom” to (3) “definite symptom”.
Validation with a clinical sample
Participants. Two hundred and thirty-eight adolescents aged 13–19 (124 males, 114 females) attending the Child Guidance Clinic were recruited from March 2001 to October 2001. We sought to recruit a roughly equal number of depressed and non-depressed adolescents based on the psychiatrist's diagnosis in the case records. Adolescents with an IQ of less than 70 or who were acutely psychotic were excluded from the study. Consent was obtained from the adolescents and their parents. A total of 300 adolescents were approached and 238 agreed to participate in the study. There was a good mix of males and females and the ethnic composition comprised 82.8% Chinese, 11.8% Malays, 4.6% Indians and 0.8% other races. The mean age of the males was 15.42±2.03 years and the mean age of the females was 15.64±1.79 years.
Materials. The adolescents were asked to complete the AADS and the CDI. The Kiddie Schedule of Affective Disorders and Schizophrenia-Present and Lifetime Version (K-SADS-PL) (Kaufman et al., 1996), a semi-structured diagnostic interview designed to assess current and past episodes of psychopathology in children and adolescents according to DSM-IV criteria (American Psychiatric Association, 1994), was also administered by 5 of the investigators, comprising 2 psychiatrists and 3 psychologists. The interviewers were blind to the AADS and CDI scores, and joint assessments were conducted for the first 20 cases in order to establish an acceptable inter-rater reliability. There was 95% agreement with regard to the presence or absence of individual symptoms of depression. A psychiatrist reviewed the interview data and the diagnoses obtained. The adolescents fell into 2 groups: (1) currently diagnosed with depression according DSM-IV criteria on the K-SADS-PL and (2) not diagnosed with depression. For those diagnosed with depression, the clinicians were asked to rate the degree of depression on a 5-point scale ranging from (1) “Mild Depression” to (5) “Severe Depression.”
Results
Development of the factor structure of the AADS
Exploratory Factor Analysis (EFA) was used to identify the factor structure of the AADS. Subsequent identification and cross-validation of the final structure was conducted with Confirmatory Factor Analysis (CFA).
The community sample (n=242) was split into 2 groups, controlling for age, sex and race. The first group, the development group (n=122), was used to develop the factor structure, and the second group, the validation group (n=120), was used for cross-validation. The clinical sample (n=238) was also split into 2 groups, controlling for age, sex, race and the presence of a diagnosis of depression. The first group (n=98) was used to develop the factor structure and the second group (n=140) was used for cross-validation.
Exploratory factor analysis
There were 113 depressed adolescents in the clinical sample. Sixty-six had depression as the primary diagnosis and 47 had depression as a comorbid diagnosis. EFA was done on 5 different groups: (1) Group 1—the developmental sub-sample of the community sample (n=122), (2) Group 2—the developmental sub-sample of the clinical sample (n=98), (3) Group 3—a combined community and clinical sample (n=220), (4) Group 4—clinical subjects diagnosed with depression (n=113) and (5) Group 5—clinical subjects with depression as the primary diagnosis (n=66). The Kaiser–Meyer–Olkin Measure of Sampling Adequacy and the Bartlett's Test of Sphericity were conducted and indices showed that it was appropriate to conduct factor analysis (see Appendix).
Using Principal Component Analysis, the factors were extracted on the basis of eigenvalue>1, as well as the solution from the scree plots, which was used as a guide to select the number of factors to be extracted. As we hypothesized that depressive symptomatology is made of correlated factors, Promax rotation was used. The factor structures of the 5 groups were compared. Group 5, comprising the clinical subjects with depression as the primary diagnosis, was found to have a factor structure that best fitted the theoretical underpinning of the AADS.
Based on the results generated from the data of Group 5, the factor loadings were assessed and the items falling into each factor were identified. The identified factors were evaluated for content relevance and conceptual coherence. Items with a factor loading of 0.3 or below and items that were not conceptually congruent with the other items within the same factor were eliminated. Two items were discarded because they had low loadings on all the factors and the remaining 22 items were distributed across 4 factors: Negative Self-Evaluation (6 items), Negative Affect (7 items), Lack of Motivation (5 items) and Cognitive Inefficiency (4 items).
Confirmatory factor analysis
The factor structure of the clinical sample was tested through a series of model fitting analyses, item trimming, model respecification and testing of alternative models using CFA with LISREL 8.5 (Joreskog & Sorbom, 1993). Model fit was assessed using the χ2 statistic and a number of goodness of fit indices.
The 4-factor model with 22 items was tested but did not fit the data. Based on the modification indices, 2 more items were eliminated. The resulting 4-factor model with 20 items was found to have the best fit: χ2=256.32, p<0.05, Adjusted Goodness of Fit Index (AGFI)=0.79, Non-Normed Fit Index (NNFI)=0.92, Comparative Fit Index (CFI)=0.93, Root Mean Square Error of Approximation (RMSEA)=0.068 and Standardized Root Mean Square Residual (SRMSR)=0.061. The 4 factors were: Negative Self-Evaluation (7 items), Negative Affect (5 items), Cognitive Inefficiency (4 items) and Lack of Motivation (4 items) (see Table 1).
| Factor 1: Negative Self-Evaluation |
| 1. I feel that I am not as good as others |
| 2. I feel that I am not wanted |
| 3. I feel hopeless |
| 4. Nothing works out right for me |
| 5. I am confused about what kind of person I am |
| 6. I do not get satisfaction from what I do |
| 7. I feel that I have no control over what happens |
| Factor 2: Negative Affect |
| 1. I have thought about dying |
| 2. I often feel like crying |
| 3. I feel sad most of the time |
| 4. My heart feels heavy |
| 5. I am more bad tempered than before |
| Factor 3: Cognitive Inefficiency |
| 1. I take a long time to decide on things |
| 2. I take a long time to get things done |
| 3. I cannot think well |
| 4. I cannot concentrate on my studies as much as I used to |
| Factor 4: Lack of Motivation |
| 1. I feel that I have no energy to do things most of the time |
| 2. I feel tired most of the time |
| 3. I do not feel like doing anything |
| 4. I do not like going out with friends or meeting people |
The factor structure of the clinical sample was then tested on the community sample and was found to have a reasonably good fit: χ2=408.01, p<0.05, AGFI=0.82, NNFI=0.85, CFI=0.87, RMSEA=0.077 and SRMSR=0.066.
Validation of the AADS
Reliability. Internal reliability measured with Cronbach-α for the AADS was high for both clinical and community samples at 0.94 and 0.91, respectively.
Convergent validity. For the clinical subjects diagnosed with depression, their mean score on the AADS was 3.68±0.54 on a scale of 1–5 and their mean score on the CDI was 0.91±0.30 on a scale of 0 to 2. Their AADS scores correlated with their CDI scores with r=0.74 (p<0.01), providing evidence for the convergent validity of the scale as a measure of depression. Their mean degree of depression rated by the clinicians was 2.24±1.01 on a scale of 1–5.
The 4 factors of the AADS were moderately correlated, with intercorrelations in the 0.60–0.74 range for the clinical sample, and in the 0.57–0.70 range for the community sample. The factors correlated in the 0.83–0.92 range with the total AADS for the clinical sample, and in the 0.79–0.91 range for the community sample, suggesting that there is a single higher order factor of depression accounting for the relationships of the primary factors.
Discriminant validity. The AADS mean scores of the clinical participants diagnosed with depression were significantly higher than those of participants not diagnosed with depression (p<0.01) (see Table 2) and those of participants in the community (p<0.01) (see Table 3).
| Depressed | Non-depressed | t-test | p-value | |
|---|---|---|---|---|
| Negative Self-Evaluation | 3.69±0.73 | 2.51±0.76 | 12.3 | <0.01 |
| Negative Affect | 3.67±0.69 | 2.29±0.75 | 14.8 | <0.01 |
| Cognitive Inefficiency | 3.75±0.74 | 2.74±0.85 | 9.8 | <0.01 |
| Lack of Motivation | 3.57±0.80 | 2.41±0.80 | 11.1 | <0.01 |
| Total AADS | 3.68±0.54 | 2.48±0.66 | 15.3 | <0.01 |
| Depressed | Community | t-test | p-value | |
|---|---|---|---|---|
| Negative Self-Evaluation | 3.69±0.73 | 2.67±0.80 | 11.7 | <0.01 |
| Negative Affect | 3.67±0.69 | 2.43±0.82 | 14.3 | <0.01 |
| Cognitive Inefficiency | 3.75±0.74 | 2.92±0.79 | 9.6 | <0.01 |
| Lack of Motivation | 3.57±0.80 | 2.59±0.80 | 10.7 | <0.01 |
| Total AADS | 3.68±0.54 | 2.65±0.69 | 14.7 | <0.01 |
Characteristics of the depressed adolescents in the clinical sample
There were 113 adolescents diagnosed with depression in the clinical sample, 44.2% male and 55.8% female. The mean age of the males was 15.70±1.90 years and the mean age of the females was 15.89±1.92 years. Ninety-five per cent of the adolescents were diagnosed to have a Major Depressive Disorder, while 5% were diagnosed to have Dysthymia.
Age and sex differences among the depressed adolescents. The depressed female adolescents (n=63) scored higher on the factors of Negative Self-Evaluation and Negative Affect, while the depressed male adolescents (n=50) scored higher on Cognitive Inefficiency and Lack of Motivation. However, these differences were not statistically significant. The older depressed adolescents aged 17–19 (n=46) scored higher than the younger depressed adolescents aged 13–16 (n=67) on all 4 factors, as well as the AADS mean score. However, these differences were statistically significant only for the factor of Lack of Motivation and the AADS mean score (p<0.05).
Discussion
The AADS comprised 4 dimensions: Negative Self-Evaluation, Negative Affect, Cognitive Inefficiency and Lack of Motivation. The dimensions of Negative Affect and Lack of Motivation are similar to those in Western populations (Kovacs, 1981; Ryan et al., 1987). The items in the Negative Self-Evaluation dimension encompassed negative socially oriented self-evaluation, negative cognition and identity confusion. Negative socially oriented self-evaluation figured relatively prominently in the AADS, providing support for the interpersonal dimension in Singaporean adolescents’ concept of depression. Cognitive Inefficiency also stood out in a dimension on its own. In Asian countries, school adjustment, future and career have been identified as adolescents’ top concerns, mirroring the societal and cultural values of the Asian society (Isralowitz & Ong, 1990). These concerns may be intensified when one is depressed, particularly in the presence of impaired concentration and declining academic performance. Our findings are in keeping with those of Greenberger et al. (2000), who found that the quality of family relationships and grades in school had significantly stronger associations with depressive symptomatology among Chinese youth than among youth in the United States. Asian youth may be more concerned about how they are evaluated by their families and peers, and about their academic performance, and this may be reflected in their manifestations of depression. Hence, negative socially oriented self-evaluation and cognitive inefficiency could be possible culture-sensitive dimensions.
Negative body image and separation–individuation conflicts, both linked to depression in Western adolescents (Frank et al., 1997; Siegel et al., 1999), were not important in our adolescents’ concept of depression. Hence, the association of a negative body image with depression could be more of a contemporary Western phenomenon. Furthermore, as Asians come from a collectivistic society with a high level of interdependence (Triandis, 1989), our adolescents may experience less separation–individuation conflicts compared to their Western counterparts.
On comparing our data to that of another study examining the manifestations of depression in a community sample of 474 Singaporean children aged 6–12 (Koh, Chang, Fung, & Kee, 2001), we found that the symptoms of depression were the similar in both children and adolescents in Singapore, but the prominence of characteristic symptoms varied. Psychosomatic manifestations such as headache, stomachache, body aches and feeling sick were common among the children but all these symptoms were eliminated after focus group discussions with our adolescents. Existential concerns, as well as negative cognitions like negative expectations about the future, catastrophizing and overgeneralizing, were prominent among the adolescents but not among the children. This could be due to relative cognitive immaturity in children and consequently, poorer planning and execution skills (Ryan et al., 1987). These findings are consistent with those of international studies, which have shown a decline in somatic complaints and an increase in hopelessness with age (Carlson & Kashani, 1988). The dimensions of Negative Affect, Lack of Motivation and Cognitive Inefficiency were common in both children and adolescents. However, the adolescents exhibited a wider range of affective and cognitive symptoms, likely due to their better skills of expression. The dimension of negative socially oriented self-evaluation was more prominent among the children, possibly because adolescents, having more experiences in an English medium education system, could be more removed from the Asian home background and the collective emphasis of the Asian family.
With regard to the depressed adolescents in the clinical sample, the older adolescents scored higher than the younger adolescents on all 4 dimensions of the AADS, consistent with the findings of Cooper and Goodyer (1993), where scores on a self-report questionnaire of mood disturbance were found to increase with age. The females scored higher on the dimensions of Negative Self-Evaluation and Negative Affect, while the males scored higher on Cognitive Inefficiency and Lack of Motivation, but these differences were not statistically significant. Nonetheless, these differences mirror the femininity-expressive and masculinity-agentic dichotomy commonly found in adolescent males and females (Spence & Helmreic, 1978). Studies also have shown that girls are more likely to complain of depressed mood (Compas et al., 1997), and have lower self-worth and self-esteem (Avison & McAlpine, 1992; McCauley et al., 1993). Males, on the other hand, have been found to score higher on the dimensions of ineffectiveness and anhedonia on the CDI (Kovacs, 1981). However, elsewhere in research literature, findings related to gender differences have been inconsistent.
Our proposed cut off score on the AADS for a diagnosis of depression in Singaporean adolescents is 4 (AADS mean score for the community sample+2 standard deviations=2.65+2(0.69)=4). However, it would be useful for us to validate the AADS on a Western control sample to ensure that it is effective in discriminating between Western and Asian populations. We would also need to validate it among other Asian populations to see if our results are generalizable to Asians of different cultures living in different countries. Another limitation of this study is the fact that our study population included adolescents aged 13–19, but our comparison tool, the CDI, has only been validated up till the age of 17. Nonetheless, the adolescents’ scores on the AADS correlated fairly well with their scores on the CDI (r=0.74, p<0.01).
Conclusion
Preliminary data supports negative socially oriented self-evaluation and cognitive inefficiency as an essential part of Singaporean adolescents’ expression of depression. These symptoms are absent in most popular Western measures of depression, but reflect the cultural emphasis of Singapore. We propose that these might be “culture-specific” symptoms of Asian collectivist cultures. Dimensions such as negative affect, existential concerns, lack of motivation and negative cognition, symptoms commonly found in Western measures of depression, were also found to be integral parts of their depressive experiences. We propose that these might be “core symptoms” of depression found in any culture. With both core and culture-specific symptoms in the AADS, the accuracy of assessing depression in Asian adolescents could be enhanced.
Acknowledgements
This study was funded by an institutional block grant from the National Medical Research Council of Singapore. We would like to thank the staff of the Child Guidance Clinic of the Institute of Mental Health for their assistance and support. We are also grateful to the principals, teachers and students of the schools involved for participating in the study.
Appendix
See Table 4.
| Kaiser–Meyer–Olkin measure of sampling adequacy | Bartlett's test of sphericity | |||
|---|---|---|---|---|
| χ2 | df | p | ||
| Developmental sub-sample of clinical sample (n=98) | 0.91 | 1517.48 | 276 | <0.01 |
| Developmental sub-sample of community sample (n=122) | 0.88 | 1444.52 | 276 | <0.01 |
| Combined sample (n=98+122=220) | 0.93 | 2886.47 | 276 | <0.01 |
| Clinical subjects diagnosed with depression (n=113) | 0.72 | 936.66 | 276 | <0.01 |
| Clinical subjects with depression as a primary diagnosis (n=66) | 0.61 | 652.77 | 276 | <0.01 |
