Welcome to the Schizophrenia Resource Centre

Welcome, this website is intended for healthcare professionals in EMEA with an interest in the treatment of schizophrenia. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

Stigmatising attitudes towards people with mental disorders: Results from a survey of Japanese high school students

Psychiatry Research, 1, 215, pages 229 - 236

Abstract

The aim of the current study was to assess the stigmatising attitudes of Japanese high school students towards people with depression, social phobia and psychosis/schizophrenia. In 2011, 311 students aged 15–18 years filled out an anonymous self-report questionnaire, which included a case vignette describing either depression, schizophrenia or social phobia and two questionnaires to assess stigmatising attitudes towards people with these disorders. Exploratory Structural Equation Modelling (ESEM) was used to determine the dimensionality and loading pattern of the stigma items in the two scales, to establish dimensions of stigma and to compare levels on these dimensions between genders. Stigmatising attitudes towards people with mental disorders in young Japanese people are substantial. ESEM revealed that the structure of stigmatising attitudes in young Japanese people is comparable in personal and perceived attitude stigma, with each forming distinct dimensions and each comprising ‘weak not sick’ and ‘dangerous/unpredictable’ components. The social distance dimension of stigma was separate from other components. Stigmatising attitudes relating to dangerousness/unpredictability were the lowest for social phobia and highest for schizophrenia. Females had lower stigmatising attitudes than males. These findings echo those of Australian studies and extend them by demonstrating a similar structure of stigma in another cultural group, namely young Japanese people.

Keywords: Stigma, Depression, Social phobia, Schizophrenia, Japanese high school students, Mental health literacy.

1. Introduction

Stigma is often cited as an issue of major concern to those with mental disorders. Stigma involves both attitudes and behaviour, but research has generally focussed on attitudes, which are more enduring than intermittent behaviours and more easily measurable. Stigmatising attitudes towards people with mental disorders are common in adolescents (Hinshaw, 2005 and Walker et al, 2008). Such attitudes may interfere with help seeking and treatment adherence, and may adversely affect quality of life, causing a young person to feel abnormal, socially disconnected and dependent on others (Sirey et al, 2001, Corrigan, 2004, and Meredith et al, 2009). The impact on help seeking is of particular importance, as evidence suggests that, despite adolescence being the period of peak onset of mental disorders, young people with symptoms of mental disorders are less likely seek appropriate help than those in other age groups (Zachrisson et al, 2006 and Rickwood et al, 2007). In addition, stigmatising attitudes may lead to discriminatory behaviours towards others with mental disorders ( Lasalvia et al., 2013 ).

A number of surveys have assessed stigmatising attitudes in adult Japanese populations. They include a survey of 2000 adults aged between 20 and 69 that was conducted in 2003/4. This survey assessed the respondent's personally-held attitudes (personal stigma) and also their beliefs about other people's attitudes (perceived stigma) towards someone with depression or schizophrenia ( Griffiths et al., 2006 ). A more recent survey used the same questionnaire to examine attitudes to schizophrenia in psychiatric nurses ( Hanzawa et al., 2012 ). Other studies include a web-based survey that examined attitudes towards schizophrenia in the general population, psychiatric staff, physicians and psychiatrists ( Hori et al., 2010 ); a study examining stigmatising attitudes towards ‘mental disorders’ in 1211 members of the public ( Tanaka et al., 2004 ); and a study of teachers' attitudes to schizophrenia ( Kurumatani et al., 2004 ). Other studies have compared attitudes in Japan with those in other countries, including a study that compared desire for social distance from a person with schizophrenia in China and Japan ( Haraguchi et al., 2009 ) and a comparison of stigmatising attitudes towards, schizophrenia, depression and obsessive-compulsive disorder in members of the public in Bali and Japan ( Kurihara et al., 2000 ) These studies indicated that health professionals were less likely to have stigmatising attitudes than the general public and that levels of stigmatising attitudes vary across cultures.

Most of these Japanese studies did not assess differences according to age and gender, although Tanaka et al. (2004) found that older age was associated with more stigmatising attitudes. Studies of adolescents in other countries do not point to clear gender-related differences, although they do suggest that stigma declines with increasing age (Dietrich et al, 2006 and Jorm and Wright, 2008). While some studies have found that females have less stigmatising attitudes than males (Griffiths et al, 2008a and Calear et al, 2010), reviews of associations between stigmatising attitudes and gender have not generally found reliable gender differences in stigmatising attitudes (Jorm and Oh, 2009 and Jorm et al, 2012a). However, there have not been any studies of stigmatising attitudes in young Japanese people.

Surveys of stigmatising attitudes in other countries show that these are common among young people, particularly those relating to dangerousness and unpredictability (Calear et al, 2010 and Kasow and Weisskirch, 2010). Stigmatising attitudes tend to vary by disorder and are higher in relation to schizophrenia than to depression and anxiety disorders ( Reavley and Jorm, 2011 ). Studies also suggest that stigma is a multidimensional construct, with previous analyses using data from large surveys of the Australian population showing that stigmatising attitudes, as measured by the Depression Stigma Scale (DSS; Griffiths et al., 2004 ), has two components, relating to beliefs that a person with a mental disorder is ‘weak not sick’ and ‘dangerous or unpredictable’. Scales reflecting these dimensions had different patterns of association with respondent age and gender, and the type of mental disorder portrayed in the vignette. Stigmatising attitudes relating to dangerousness/unpredictability were the lowest in response to social phobia and PTSD and the highest in response to schizophrenia, whereas attitudes relating to ‘weak not sick’ were highest for social phobia (Jorm and Wright, 2008 and Yap et al, 2013). However, further work is needed to establish whether the dimensions are comparable in different cultures.

Given the lack of data on stigmatising attitudes in Japanese youth, the aim of this study was to carry out a survey assessing high school students' stigmatising attitudes towards those with depression, schizophrenia and social phobia. We also aimed to use the data gathered in the survey to determine the factor structure of stigmatising attitudes in Japanese youth as measured by the DSS ( Griffiths et al., 2004 ) and the Social Distance Scale (SDS: Link et al., 1999 ) and to establish the number of dimensions these items tap. We also sought to compare levels on these dimensions between males and females.

2. Methods

2.1. Participants and procedure

Initially, one author (KY) approached local high schools and met with principals and teachers to explain the purpose of the study and seek the support of the school. Schools that agreed to participate were provided with hardcopy surveys, which were distributed by homeroom teachers of Grades 1 to 2 (ages 15–17 years) during one homeroom period of approximately 40–50 min. Student participants were provided with information regarding the study, and their freedom to decline participation, on the survey paper. Consent to participate was implied by survey completion. Three hundred and eleven surveys were distributed in two schools, during December 2010. All distributed surveys were returned by students to their homeroom teachers, then to the research staff. The study was approved by the University of Fukuoka Ethics Committee.

2.2. Instruments

The survey was based on the Mental Health Literacy Interview developed by Jorm and colleagues (Jorm and Wright, 2007 and Jorm and Wright, 2008), which presents to participants a vignette of a young person with a mental disorder and then asks a range of questions regarding problem recognition, help-seeking intentions, beliefs about treatment and stigmatising attitudes. Vignettes describing young people with depression, psychosis or social phobia have been used and validated in previous research on youth mental health literacy in Australia (Jorm and Wright, 2007, Jorm et al, 2008, and Wright et al, 2011). These vignettes were translated into Japanese by one of the authors (KY) with the assistance of a psychiatrist and checked by a second person. All surveys were completed in Japanese.

In order to limit respondent burden, students were randomly allocated one of three vignettes: depression, psychosis or social phobia, with each package of surveys containing equal numbers of each vignette, randomly distributed throughout the package. The gender of the character described in the vignette was also randomly allocated; the English language character ‘John’ was translated to ‘Shōta-san’ and the English language character ‘Jenny’ was translated to ‘Ai-san’. The character in the vignette was described as being 15 years old. The Japanese vignettes are available from the author on request.

After reading the vignette, students were asked what, if anything, they thought was wrong with the person described. They were also asked about where they would go for help if they had a problem like the person in the vignette, their beliefs about a range of interventions and self-help behaviours, barriers to help seeking (reported in Yoshioka et al. (submitted for publication) ) and also about their stigmatising attitudes, which are the focus of the current paper. Questions about sociodemographic characteristics were also included. These included age, gender and living situation (both parents, mother only, father only, with neither parent).

Stigmatising attitudes were assessed with two sets of statements, one assessing the respondent's personal attitudes towards the person described in the vignette (personal stigma) and the other assessing the respondent's beliefs about other people's attitudes towards the person in the vignette (perceived stigma). The items were adapted to be suitable for young people ( Jorm and Wright, 2008 ) based on a scale for adults (Griffiths et al, 2004 and Griffiths et al, 2006).

2.2.1. Personal stigma scale

The personal stigma items were: (1) people with a problem like (Shota-san/Ai-san)'s could snap out of it if they wanted; (2) a problem like (Shota-san/Ai-san)'s is a sign of personal weakness; (3) (Shota-san/Ai-san)'s problem is not a real medical illness; (4) people with a problem like (Shota-san/Ai-san)'s are dangerous; (5) it is best to avoid people with a problem like (Shota-san/Ai-san)‘s so that you do not develop this problem; (6) people with a problem like (Shota-san/Ai-san)'s are unpredictable; and (7) if I had a problem like (Shota-san/Ai-san)'s I would not tell anyone.

2.2.2. Perceived stigma scale

The perceived stigma items covered the same statements but started with ‘Most other people believe that…’ Ratings of each were made on a 5-point Likert scale ranging from ‘strongly agree’ to ‘strongly disagree’. As these scales have not previously been used in young Japanese people, percentages of people agreeing with individual items are reported here. For these analyses the ‘agree’ and ‘strongly agree’ categories were combined.

2.2.3. Social distance scale

Self-reported willingness to have contact with the person described in the vignette was measured using by a social distance scale suitable for young people ( Jorm and Wright, 2008 ) which was an adaptation of a scale developed by Link et al. for adults ( Link et al., 1999 ). The items were rated according to the person’s willingness to (1) go out with (Shota-san/Ai-san) on the weekend; (2) to invite (Shota-san/Ai-san) around to your house; (3) to go to (Shota-san/Ai-san)‘s house; (4) working closely with (Shota-san/Ai-san) on a project; and (5) to develop a close friendship with (Shota-san/Ai-san). Each item was rated on a 4-point scale ranging from definitely willing to definitely unwilling. For the analyses of agreement with individual items, the ‘definitely unwilling’ and ‘probably unwilling’ categories were combined.

2.3. Statistical analysis

2.3.1. Exploratory structural equation modelling

Previous exploratory Principal Components Analyses (PCA) which included self and perceived stigma items in the same analyses in an Australian sample have been equivocal regarding the number of factors and loading pattern (Griffiths et al, 2004 and Jorm and Wright, 2008). However, it is clear that the set of items is multidimensional. Jorm and Wright (2008) named two of the factors they found ‘Dangerous/unpredictable’ and ‘Weak not sick’. A limitation of PCA is that cannot accommodate the parallel nature of the personal and perceived stigma items and mixes all items together. In the current study, a novel technique – Exploratory Structural Equation Modelling (ESEM) – was used. This simultaneously fitted separate two-factor exploratory factor models to both the personal and perceived stigma items, but allowed testing of constraining the pattern of loadings of items on these factors. Testing these constraints sought to determine whether the measurement structure of personal and perceived stigma was comparable. The establishment of parallel measurement structures allowed scores on the personal and perceived stigma scales to be meaningfully compared.

For the social distance items, a single confirmatory factor on which all items were permitted to load was specified. This factor was permitted to correlate freely with all exploratory factors. Item responses were treated as ordinal data, with polychoric correlations estimated between items. Model parameters were estimated using a robust Weighted Least Squares Method with Diagonal Weight Matrix (WLSMV) in Mplus 6.11 ( Muthén and Muthén, 1998-2010 ). Models were compared using the Difftest procedure in Mplus.

3. Results

Table 1 shows the sociodemographic characteristics of students. Just over 55% of these were male and just under 45% female. Participants were aged between 15 and 17 years (M=16.1, SD=0.69) with the majority being 16 or 17. Over 85% lived with both parents. The numbers of students who received each vignette were: depression n=99, schizophrenia n=117, and social phobia n=95.

Table 1 Sociodemographic characteristics of students.

Socio-demographic characteristic n %
Gender
  Male 171 55.0
  Female 139 44.7
  Missing 1 0.3
 
Age category
  15 46 14.8
  16 160 51.4
  17 100 32.2
  18 3 1.0
  Missing 1 0.3
 
Living situation
  With both parents 265 85.2
  With mother only 35 11.3
  With father only 2 0.6
  With neither parent 1 0.3
  Refused/Missing 8 2.6

To simplify interpretation, initial consideration of the results involved the proportion of respondents agreeing with individual items (see Table 2 for personal and perceived stigma items and Table 3 for social distance items). This involved identification of the specific stigmatising attitudes that were most common and least common for each vignette, giving an initial profile of how stigma manifests for the various vignettes.

Table 2 Percentage of students who ‘agree’ or ‘strongly agree’ with statements about personal and perceived attitudes to mental disorders.

Statement Depression a Psychosis/schizophrenia b Social Phobia c Difference between vignettes (χ2 (4) and p value)
Personal attitudes
Person could snap out of the problem 25.5 28.2 40.0 5.93, p=0.204
Problem is a sign of personal weakness 24.2 23.1 29.5 2.45, p=0.653
Problem is not a real medical illness 23.2 26.5 lowastlowast 47.4 lowastlowast 16.71, p=0.002
People with this problem are dangerous 12.1 21.4 lowastlowast 5.3 13.71, p=0.008
Avoid people with this problem 5.1 6.8 7.4 1.71, p=0.789
People with this problem are unpredictable 18.2 lowast 30.8 lowastlowastlowast 14.7 21.93, p<0.001
If I had this problem I would not tell anyone 35.4 34.2 32.6 0.31, p=0.989
 
Perceived attitudes
Person could snap out of the problem 37.4 40.2 46.3 1.99, p=0.738
Problem is a sign of personal weakness 43.4 45.3 49.5 0.80, p=0.938
Problem is not a real medical illness 31.3 29.3 42.1 4.89, p=0.299
People with this problem are dangerous 39.4 51.3 33.0 8.13, p=0.087
Avoid people with this problem 31.3 34.2 31.6 0.84, p=0.934
People with this problem are unpredictable 44.4 52.1 36.8 5.23, p=0.265
If I had this problem I would not tell anyone 33.3 35.9 36.8 1.04, p=0.904

a Partition of χ2 comparing depression to psychosis/schizophrenia.

b Partition of χ2 comparing psychosis/schizophrenia to social phobia.

c Partition of χ2 comparing depression to social phobia.

lowast p<0.05.

lowastlowast p<0.01.

lowastlowastlowast p<0.001.

Notes:χ2 tests and χ2 partitions ( Agresti, 2002 ) were used to assess differences in the distribution between vignettes.

Table 3 Percentage of students unwilling to socially interact with each person described in the vignette.

Social interaction Depression a Psychosis/schizophrenia b Social phobia c Difference between vignettes (χ2 (2) and p value)
To go out with Shota-san/Ai-san on the weekend 45.9 lowastlowast 64.1 lowast 50.0 7.99, p=0.018
To work on a project with Shota-san/Ai-san 40.8 lowastlowast 59.0 lowastlowast 37.6 11.51, p=0.003
Invite Shota-san/Ai-san around to your house 46.9 lowastlowast 68.7 lowastlowastlowast 41.9 17.37, p<0.001
Go to Shota-san/Ai-san’s house 45.9 57.0 lowast 42.6 4.87, p=0.088
Develop a close friendship with Shota-san/Ai-san 33.7 lowastlowast 53.0 lowastlowastlowast 28.0 15.51, p<0.001

a Partition of χ2 comparing depression to psychosis/schizophrenia.

b Partition of χ2 comparing psychosis/schizophrenia to social phobia.

c Partition of χ2 comparing depression to social phobia.

lowast p<0.05.

lowastlowast p<0.01.

lowastlowastlowast p<0.001.

Notes:χ2 tests and χ2 partitions ( Agresti, 2002 ) were used to assess differences in the distribution between vignettes.

3.1. Personal and perceived stigma

For personal stigma, for the depression and schizophrenia vignettes, avoiding the person was the item least likely to be endorsed, whereas the item relating to not telling anyone was the most likely to be endorsed. For the social phobia vignette, the item relating to dangerousness was least likely to be endorsed, while the item relating to the disorder not being a real medical illness was the most likely to be endorsed.

Analysis of individual items showed that personal stigma was lower than perceived stigma for all disorders. For perceived stigma, for the depression items, the item relating to avoiding the person was least likely to be endorsed, whereas the item relating to unpredictability was most likely to be endorsed. For schizophrenia, the item relating to the problem not being a real illness was least likely to be endorsed, while the item relating to dangerousness was most likely to be endorsed. For social phobia, the item relating to dangerousness was least likely to be endorsed, whereas the item relating to personal weakness was most likely to be endorsed.

For the personal stigma items, belief that the disorder was not a real illness was highest for social phobia and the lowest for schizophrenia (χ2 (4)=16.71, p=0.002), whereas beliefs that a person was dangerous (χ2 (4)=13.71, p=0.008) or unpredictable (χ2 (4)=21.93, p<0.001) were the highest for schizophrenia and the lowest for social phobia. Similar patterns were seen for the perceived stigma items, although the differences were not significantly different.

3.2. Exploratory structural equation models

A series of nested ESEM models were fitted to the survey data. A preliminary analysis indicated that models with single factors each for personal and perceived stigma items would not be an acceptable fit. Thus, the ESEM models allowed for two factors each for personal and perceived stigma items. These were exploratory factors on which all items loaded freely, rotated to an oblique solution using the Geomin algorithm. Residuals of corresponding personal and perceived stigma items were permitted to correlate. The social distance items were modelled as a single confirmatory factor that was permitted to correlate freely with the personal and perceived stigma factors. The first model permitted loadings on personal and perceived stigma factors to differ. The second model constrained loadings on corresponding personal and perceived factors to equality. The third model added constraint of response category thresholds. Outcomes of a final possibility – equality of residual variances for corresponding personal and perceived stigma items – are not reported due to technical problems fitting these models.

χ2 tests of model fit were significant in all cases. This reflects the sensitivity of this index to sample size and to small deviations from model specification. Model 1 provided an excellent fit to the data with all indices exceeding recommended thresholds for acceptable fit ( Yoshioka and Misawa, 2013 ) (root mean square error of approximation (RMSEA)) (90% CI)=0.056 (0.045–0.066), TLI=98.98, CFI=0.99). Constraining personal and perceived factor loadings (Model 2) resulted in relatively minor decrements in fit (RMSEA (90% CI)=0.061 (0.051–0.070), Tucker-Lewis Index (TLI)=0.98, comparative fit index (CFI)=0.98). Constraining item thresholds (Model 3) introduced a notable decrease in model fit (RMSEA (90% CI)=0.071 (0.063–0.080), TLI=0.97, CFI=0.97). Accordingly, it was concluded that Model 2 provided a good, parsimonious fit to the data. Importantly, this model establishes that personal stigma and that perceived in others have parallel measurement structures. Thus, scales derived from this model were the focus of subsequent analyses.

Factor loadings shown in Table 4 indicate that the first factor – hereafter referred to as ‘weak not sick’–comprises items which characterise the problem portrayed in the vignette as a personal weakness under the control of the person rather than as a medical condition. The second factor – hereafter referred to as ‘dangerous/unpredictable’ – was defined by items characterising the vignette character as unpredictable and dangerous. The item indicating a preference to avoid the person also loaded on this factor. The item ‘If I had this problem, I would not tell anyone.’ loaded moderately on the dangerous/unpredictable factor. As Table 4 shows, all social distance items in both surveys loaded substantially and relatively uniformly on the single factor specified for them.

Table 4 Factor loadings.

Item Loadings (standard error)
Japanese survey
Weak-not-sick Dangerous/unpredictable Social distance
1. Person could snap out of the problem 0.83 lowastlowastlowast (0.05) −0.01 (0.01)
2. Problem is a sign of personal weakness 0.51 lowastlowastlowast (0.08) 0.32 lowastlowastlowast (0.05)
3. Problem is not a real medical illness 0.53 lowastlowastlowast (0.06) 0.12 lowast (0.05)
4. People with this problem are dangerous 0.07 (0.13) 0.68 lowastlowastlowast (0.05)
5. Best to avoid people with this problem 0.27 lowast (0.12) 0.53 lowastlowastlowast (0.05)
6. People with this problem unpredictable −0.20 (0.16) 0.85 lowastlowastlowast (0.05)
7. If I had this problem I would not tell anyone 0.00 (0.02) 0.31 lowastlowastlowast (0.04)
 
Social distance
1. Go out with on the weekend 0.89 lowastlowastlowast (0.01)
2. Go to house 0.87 lowastlowastlowast (0.02)
3. Invite around to your house 0.87 lowastlowastlowast (0.01)
4. Work on a project with 0.90 lowastlowastlowast (0.01)
5. Develop a close friendship with 0.87 lowastlowastlowast (0.02)

lowast p<0.05.

lowastlowastlowast p<0.001.

lowastlowastp<0.01.

All factors were permitted to correlate freely. The correlation between the two personal stigma factors was moderate (r=0.28, p<0.001), with the comparable association for perceived stigma being larger (r=0.58, p<0.001). There was a moderate associations between the personal and perceived ‘weak not sick’ factors (r=0.33, p<0.001) and a high association between the dangerous/unpredictable factors (r=0.53, p<0.001). The social distance factor was generally highly correlated with the personal dangerous/unpredictable factor (r=0.59, p<0.001). Other correlations were negligible and/or non-significant.

3.2.1. Scale scores

Scale scores were constructed to reflect the structure of the dimensions revealed by modelling. Alternatives to this approach are further structural modelling or the use of factor scores. The former is inappropriate when looking for associations with multiple categorical variables and both approaches limit application by other researchers. In contrast, scale scores are easily calculated and used. Scores reflecting each factor were calculated as the mean of items that loaded substantially (>0.30) on the factor, with higher mean scores indicating more stigmatising attitudes. These scores were compared to factor scores estimated by Mplus. Despite the small number of items comprising the scales, most correlated very highly with the corresponding factor scores and had Cronbach's alpha values in the acceptable to outstanding range (0.68–0.97).

3.2.2. Social distance

Desire for social distance from those with schizophrenia was highest for all social distance items (see Table 3 ). Mean (SD) scores for the SDS were as follows: depression M=2.40 (SD=0.66), schizophrenia M=2.71 (SD=0.75), social phobia M=2.38 (SD=0.76). A one-way ANOVA showed these to be significantly different (p=0.001). Tukey post-hoc comparisons of the three groups indicated that desire for social distance was higher for schizophrenia than depression (p=0.006) and social phobia (p=0.004), while the differences for depression and social phobia were not significant (p=0.981). The Cronbach alpha value for the social distance scale was 0.92.

3.3. Group differences and correlates of the stigma scales

Having established their comparable measurement properties, it is valid to compare distributions and responses on the personal stigma scales to their corresponding perceived stigma scale. Fig. 1 shows the distribution the personal and perceived scales in both surveys while Fig. 2 shows the distribution of the SDS. These show differences in the distribution of ‘weak not sick’ and ‘dangerous/unpredictable’ scores for personal compared to perceived stigma. For the ‘weak not sick’ scale, personal ratings were positively skewed, with more respondents reporting non-stigmatising attitudes. Appraisals of others were less generous, with a more stigmatized modal rating. Differences in the distributions of ‘dangerous/unpredictable’ scales were largely similar. Social Distance had a more symmetric distribution of responses.

gr1

Fig. 1 Distribution of the personal and perceived stigma scale scores (item means).

gr2

Fig. 2 Distribution of the social distance scale scores (item means).

3.4. Differences due to gender and vignette

ANOVA was used to investigate differences due to gender and vignette. Means and standard deviations for each scale are shown in Table 5 . Significant gender differences were found for all scales other than the perceived ‘weak not sick’ scale, with males having higher stigmatising attitudes than females.

Table 5 Stigma scales – Means and standard deviations (in parentheses) by gender.

  Personal stigma Perceived stigma Social distance
Weak-not-Sick Dangerous/unpredictable Weak-not-Sick Dangerous/unpredictable
Gender
Male(N=164–168) 2.77 lowastlowast (0.92) 2.53 lowastlowastlowast (0.90) 3.22(0.75) 3.35 lowastlowast (0.97) 2.67 lowastlowastlowast (0.79)
Female(N=137–138) 2.52(0.62) 2.11(0.70) 3.11(0.66) 3.04(0.93) 2.30(0.60)

lowastlowast p<0.01.

lowastlowastlowast p<0.001.

NB Ns for subgroups have minor variation between scales due to missing responses.

Fig. 3 shows mean scale scores as a function of the disorder portrayed in the vignette described to the respondent. Differences between vignettes were significant only for the personal ‘dangerous/unpredictable’ scale (F (2303)=6.25, p=0.002). This was due to the difference between schizophrenia and social phobia (p=0.001). Other contrasts were not significant.

gr3

Fig. 3 Mean stigma scale scores as a function of vignette type.

4. Discussion

This survey, the first to have examined stigmatising attitudes towards mental disorders in young Japanese people, showed that stigma, particularly perceived stigma was substantial. Males were also more likely to have stigmatising attitudes than females. In addition, ESEM methods revealed personally held stigmatising attitudes and those perceived in others formed distinct dimensions, and also that personal and perceived stigma were not unitary dimensions, but each comprised two components, namely a ‘weak not sick’ component and a ‘dangerous/unpredictable’ component. These findings echo those of Griffiths et al. (2008b) and Jorm and Wright (2008) and extend them by demonstrating a similar structure in another cultural group, namely young Japanese people. We also replicated Jorm and Wright's finding that the social distance dimension of stigma is separate from other components of stigma, supporting the appropriateness of using existing social distance scales in other cultural groups. These findings have implications for the interpretation of findings from studies examining the predictors of stigma ( Griffiths et al., 2008a ), as well as for the design, evaluation, and dissemination of interventions to reduce stigma.

These results may be compared to those of other studies of stigmatising attitudes in Japanese populations, including the 2003/4 study of 2000 adults ( Griffiths et al., 2006 ). Levels of personal stigma appear to be lower in Japanese adolescents than in adults, whereas levels of perceived stigma were generally similar or somewhat lower in adolescents than in adults. Desire for social distance was generally lower in adolescents than in adults, particularly for the depression vignette.

Comparison with surveys of stigma in young Australians ( Reavley and Jorm, 2011 ) suggests that while personal beliefs are similar on some items, young Japanese people are less likely to believe that someone with depression is unpredictable, but are more likely to believe that depression is not a real illness and are also more likely to endorse not telling anyone if they had the problem. A similar pattern was seen for schizophrenia, with young Japanese people also having a lower belief in dangerousness. For social phobia, young Japanese people were also more likely to believe that the person could snap out of it. In the current study, mean (SD) personal stigma scores for males were 2.77 (0.92) and 2.53 (0.90) for the weak-not-sick and dangerous/unpredictable scale scores respectively, while for young Australians the comparable scores were 1.94 (0.79) and 2.40 (0.70) ( Yap et al., in press ). For females, the equivalent Japanese figures were 2.52 (0.62) and 2.11 (0.70), while for young Australians they were 1.67 (0.71) and 2.22 (0.65). Overall, it appears that young Japanese people are more likely than young Australians to believe that mental disorders are due to weakness rather than sickness, while beliefs in dangerousness and unpredictability are similar. They are also less inclined to disclose mental disorders and more likely to report a desire for social distance, with mean (SD) scores in Japanese males (2.67 (0.79)) and females (2.30 (0.60)) higher than scores in young Australians (males: 1.68 (0.56) and females: 1.60 (0.55)). It is possible that the differences arise because the greater exposure of young Australians to mental health education such as that provided by organisations such as beyondblue: (www.beyondlbue.org.au) or MindMatters (www.mindmatters.edu.au ) has led to lower stigmatising attitudes. However, it is also possible that Japanese students are less likely to respond in a socially desirable way. However, comparison with the Australian survey is limited by the differences in age range between the two studies (15–17 in the Japanese survey and 15–25 in the Australian survey). These results support the need for good quality mental health education aimed at young people that focuses on mental disorders as ‘real illnesses’ and addresses the barriers to disclosure.

In young Japanese people, the item concerning reluctance to disclose one's mental health difficulties loaded moderately on the dangerous/unpredictable scale, suggesting that this reluctance may be associated with beliefs about the dangerousness or unpredictability of people with the problem. The item concerning avoiding others with the problem also loaded on the dangerous/unpredictable factor. Similar analysis in young Australians suggests that reluctance to disclose and a desire to avoid the person are less likely to be associated with dangerousness and may point to the need for specific education to address this issue in young Japanese people (Jorm and Wright, 2008 and Yoshioka and Misawa, 2013).

Analysis of scores by vignette showed that attitudes on the ‘weak not sick’ scale, both personally held and those perceived in others, did not vary significantly on account of the mental disorder portrayed in the vignette, although there was a trend for social phobia to be scored more highly on this scale. However, scores on the personal dangerous/unpredictable scale were higher for schizophrenia than for social phobia. These findings are consistent with Jorm and Wright's (2008) finding involving social phobia. This is unsurprising given that anxiety disorders are characterised by withdrawal-like behaviour and hence are less likely to be associated with potential harm to others. These findings, may be influenced by media portrayals of the associations between violence and psychosis (Hocking, 2003 and Berry et al, 2007), given that personal encounters with a violent person with schizophrenia are likely to be rare ( Jorm and Reavley, 2013 ).

Examination of gender differences revealed generally lower stigmatising attitudes in females. While some other studies have shown lower stigmatising attitudes in females, including a study of predictors of depression stigma in Australian adolescents ( Calear et al., 2010 ),, reviews of associations between stigmatising attitudes and gender have not generally found reliable gender differences (Jorm and Oh, 2009 and Jorm et al, 2012a).

Limitations of the study include its moderate sample size and the recruitment from only two schools in one area of Japan. This may have implications for the representativeness of the sample. In terms of the structural models and derived scales, there were only a small number of items on each scale. While psychometrically robust, additional questions might reveal a more complex structure underlying stigma. In addition, other important aspects and potential consequences of stigma (e.g. self-stigma, discriminatory behaviour and experienced stigma) were not assessed in the survey. Responses to the stigma items may be affected by the social desirability of responses, however the comparability of personal and perceived stigma, particularly on the dangerous/unpredictable scale provides some rebuttal to this factor driving responses.

In conclusion, stigmatising attitudes towards mental disorders in young Japanese people are substantial, with males more likely to have stigmatising attitudes than females. ESEM methods revealed that personally held stigmatising attitudes and those perceived in others have a similar structure, each comprising two components, namely ‘weak not sick’ and ‘dangerous/unpredictable’. These findings echo those of Australian studies and extend them by demonstrating a similar structure in another cultural group, namely young Japanese people.

Acknowledgements

Funding for the study was provided by the Japanese Society for the Promotion of Science. Reavley and Jorm received salary support from the National Health and Medical Research Council.

References

  • Agresti, 2002 A. Agresti. in: Categorical Data Analysis (2nd Ed.).. (Wiley, New York, 2002)
  • Berry et al., 2007 J.G. Berry, K. Pidd, A.M. Roche, J.E. Harrison. Prevalence and patterns of alcohol use in the Australian workforce: findings from the 2001 National Drug Strategy Household Survey. Addiction. 2007;102:1399-1410
  • Calear et al., 2010 A.L. Calear, K.M. Griffiths, H. Christensen. Personal and perceived depression stigma in Australian adolescents: magnitude and predictors. Journal of Affective Disorders. 2010;129:104-108
  • Corrigan, 2004 P. Corrigan. How stigma interferes with mental health care. American Psychologist. 2004;59:614-625
  • Dietrich et al., 2006 S. Dietrich, D. Heider, H. Matschinger, M.C. Angermeyer. Influence of newspaper reporting on adolescents' attitudes toward people with mental illness. Social Psychiatry and Psychiatric Epidemiology. 2006;41:318-322
  • Griffiths et al., 2008a K.M. Griffiths, H. Christensen, A.F. Jorm. Predictors of depression stigma. BMC Psychiatry. 2008;8:25
  • Griffiths et al., 2008b K.M. Griffiths, H. Christensen, A.F. Jorm. Predictors of depression stigma. BMC Psychiatry. 2008;8
  • Griffiths et al., 2004 K.M. Griffiths, H. Christensen, A.F. Jorm, K. Evans, C. Groves. Effect of web-based depression literacy and cognitive-behavioural therapy interventions on stigmatising attitudes to depression: randomised controlled trial. British Journal of Psychiatry. 2004;185:342-349
  • Griffiths et al., 2006 K.M. Griffiths, Y. Nakane, H. Christensen, K. Yoshioka, A.F. Jorm, H. Nakane. Stigma in response to mental disorders: a comparison of Australia and Japan. BMC Psychiatry. 2006;6:21
  • Hanzawa et al., 2012 S. Hanzawa, A. Nosaki, K. Yatabe, Y. Nagai, G. Tanaka, H. Nakane, Y. Nakane. Study of understanding the internalized stigma of schizophrenia in psychiatric nurses in Japan. Psychiatry and Clinical Neurosciences. 2012;66:113-120
  • Haraguchi et al., 2009 K. Haraguchi, M. Maeda, Y.X. Mei, N. Uchimura. Stigma associated with schizophrenia: cultural comparison of social distance in Japan and China. Psychiatry and Clinical Neurosciences. 2009;63:153-160
  • Hinshaw, 2005 S.P. Hinshaw. The stigmatization of mental illness in children and parents: developmental issues, family concerns, and research needs. Journal of Child Psychology and Psychiatry and Allied Disciplines. 2005;46:714-734
  • Hocking, 2003 B. Hocking. Reducing mental illness stigma and discrimination - everybody's business. Medical Journal of Australia. 2003;178(Suppl.):S47-48
  • Hori et al., 2010 H. Hori, M. Richards, Y. Kawamoto, H. Kunugi. Attitudes toward schizophrenia in the general population, psychiatric staff, physicians, and psychiatrists: a web-based survey in Japan. Psychiatry Research. 2010;186:183-189
  • Jorm et al., 2008 A.F. Jorm, A.J. Morgan, A. Wright. A comparison of clinician, youth, and parent beliefs about helpfulness of interventions for early psychosis. Psychiatric Services. 2008;59:1115-1120
  • Jorm and Oh, 2009 A.F. Jorm, E. Oh. Desire for social distance from people with mental disorders. Australian and New Zealand Journal of Psychiatry. 2009;43:183-200
  • Jorm et al., 2012a A.F. Jorm, N.J. Reavley, A.N. Ross. Belief in the dangerousness of people with mental disorders: A review. Australian and New Zealand Journal of Psychiatry. 2012;46:1029-1045
  • Jorm et al., 2013 Jorm, A.F., Reavley N.J., 2013. Public belief that mentally ill people are violent: Is the USA exporting stigma to the rest of the world? Australian and New Zealand Journal of Psychiatry. Oct 22 (epub ahead of print)
  • Jorm and Wright, 2007 A.F. Jorm, A. Wright. Beliefs of young people and their parents about the effectiveness of interventions for mental disorders. Australian and New Zealand Journal of Psychiatry. 2007;41:656-666
  • Jorm and Wright, 2008 A.F. Jorm, A. Wright. Influences on young people's stigmatising attitudes towards peers with mental disorders: national survey of young Australians and their parents. British Journal of Psychiatry. 2008;192:144-149
  • Kasow and Weisskirch, 2010 Z.M. Kasow, R.S. Weisskirch. Differences in attributions of mental illness and social distance for portrayals of four mental disorders. Psychological Reports. 2010;107:547-552
  • Kurihara et al., 2000 T. Kurihara, M. Kato, S. Sakamoto, R. Reverger, T. Kitamura. Public attitudes towards the mentally ill: a cross-cultural study between Bali and Tokyo. Psychiatry and Clinical Neurosciences. 2000;54:547-552
  • Kurumatani et al., 2004 T. Kurumatani, K. Ukawa, Y. Kawaguchi, S. Miyata, M. Suzuki, H. Ide, W. Seki, E. Chikamori, H.G. Hwu, S.C. Liao, G.D. Edwards, N. Shinfuku, M. Uemoto. Teachers' knowledge, beliefs and attitudes concerning schizophrenia- a cross-cultural approach in Japan and Taiwan. Social Psychiatry and Psychiatric Epidemiology. 2004;39:402-409
  • Lasalvia et al., 2013 A. Lasalvia, S. Zoppei, T. Van Bortel, C. Bonetto, D. Cristofalo, K. Wahlbeck, S.V. Bacle, C. Van Audenhove, J. van Weeghel, B. Reneses, A. Germanavicius, M. Economou, M. Lanfredi, S. Ando, N. Sartorius, J.J. Lopez-Ibor, G. Thornicroft. Global pattern of experienced and anticipated discrimination reported by people with major depressive disorder: a cross-sectional survey. Lancet. 2013;381:55-62
  • Link et al., 1999 B.G. Link, J.C. Phelan, M. Bresnahan, A. Stueve, B.A. Pescosolido. Public conceptions of mental illness: labels, causes, dangerousness, and social distance. American Journal of Public Health. 1999;89:1328-1333
  • Meredith et al., 2009 L.S. Meredith, B.D. Stein, S.M. Paddock, L.H. Jaycox, V.P. Quinn, A. Chandra, A. Burnam. Perceived barriers to treatment for adolescent depression. Medical Care. 2009;47:677-685
  • Muthén and Muthén, 1998–2010 L.K. Muthén, B.O. Muthén. Mplus User's Guide, Los Angeles, CA. (, 1998–2010)
  • Reavley and Jorm, 2011 N.J. Reavley, A.F. Jorm. Young people's stigmatising attitudes towards people with mental disorders: findings from an Australian national survey. Australian and New Zealand Journal of Psychiatry. 2011;45:1033-1039
  • Rickwood et al., 2007 D.J. Rickwood, F.P. Deane, C.J. Wilson. When and how do young people seek professional help for mental health problems?. The Medical Journal of Australia. 2007;187:S35-39
  • Sirey et al., 2001 J.A. Sirey, M.L. Bruce, G.S. Alexopoulos, D.A. Perlick, P. Raue, S.J. Friedman, B.S. Meyers. Perceived stigma as a predictor of treatment discontinuation in young and older outpatients with depression. American Journal of Psychiatry. 2001;158:479-481
  • Tanaka et al., 2004 G. Tanaka, H. Inadomi, Y. Kikuchi, Y. Ohta. Evaluating stigma against mental disorder and related factors. Psychiatry and Clinical Neurosciences. 2004;58:558-566
  • Walker et al., 2008 J.S. Walker, D. Coleman, J. Lee, P.N. Squire, B.J. Friesen. Children’s stigmatization of childhood depression and ADHD: magnitude and demographic variation in a national sample. Journal of the American Academy of Child and Adolescent Psychiatry. 2008;47:912-920
  • Wright et al., 2011 A. Wright, A.F. Jorm, A.J. Mackinnon. Labeling of mental disorders and stigma in young people. Social Science and Medicine. 2011;73:498-506
  • Yap et al., 2013 M.B.H. Yap, A.J. Mackinnon, N.J. Reavley, A.F. Jorm. The measurement properties of stigmatising attitudes towards mental disorders: results from two community surveys. International Journal of Methods in Psychiatric Research. 2013; (in press)
  • Yoshioka and Misawa, 2013 K. Yoshioka, R. Misawa. Causal attribution of depression and the social distance caused by the stigma. The Japanese Journal of Health Psychology. 2013;25 93–103
  • Yoshioka et al., 2013 K. Yoshioka, N.J. Reavley, L. Hart, A.F. Jorm. Recognition of mental disorders and beliefs about treatment: results from a mental health literacy survey of Japanese high school students. 2013; (submitted for publication)
  • Zachrisson et al., 2006 H.D. Zachrisson, K. Rodje, A. Mykletun. Utilization of health services in relation to mental health problems in adolescents: a population based survey. BMC Public Health. 2006;6:34

Footnotes

a University of Fukuoka, Fukuoka, Japan

b Melbourne School of Population and Global Health, University of Melbourne, Melbourne, Australia

c Centre for Youth Mental Health, University of Melbourne, Melbourne, Australia

lowast Correspondence to: University of Fukuoka, 8-19-1 Nanakuma, Jonan-ku, Fukuoka City 814-0180, Japan. Tel.: +81 92 871 6631; fax: +81 92 871 6654.