## 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

# Cross-cultural comparisons of attitudes toward schizophrenia amongst the general population and physicians: A series of web-based surveys in Japan and the United States

Psychiatry Research

### Abstract

Cross-cultural differences in attitudes toward schizophrenia are suggested, while no studies have compared such attitudes between the United States and Japan. In our previous study in Japan ( Hori et al., 2011 ), 197 subjects in the general population and 112 physicians (excluding psychiatrists) enrolled in a web-based survey using an Internet-based questionnaire format. Utilizing the identical web-based survey method in the United States, the present study enrolled 172 subjects in the general population and 45 physicians. Participants' attitudes toward schizophrenia were assessed with the English version of the 18-item questionnaire used in our previous Japanese survey. Using exploratory factor analysis, we identified four factors labeled “social distance,” “belief of dangerousness,” “underestimation of patients’ abilities,” and “skepticism regarding treatment.” The two-way multivariate analysis of covariance on the four factors, with country and occupation as the between-subject factors and with potentially confounding demographic variables as the covariates, revealed that the general population in the US scored significantly lower than the Japanese counterparts on the factors “social distance” and “skepticism regarding treatment” and higher on “underestimation of patients' abilities.” Our results suggest that culture may have an important role in shaping attitudes toward mental illness. Anti-stigma campaigns that target culture-specific biases are considered important.

Keywords: Stigma, Schizophrenia, Cross-cultural comparison, Survey.

### 1. Introduction

Negative attitudes toward mental illness, schizophrenia in particular, are considered the most significant obstacles impeding improvement in the lives of individuals with these conditions and their families ( Kadri and Sartorius, 2005 ). While stigmatizing attitudes toward schizophrenia are prevalent worldwide ( Thornicroft et al., 2009 ), such attitudes can be shaped by cultural and historical forces ( Fabrega, 1991 ). There is marked heterogeneity in the outcomes of schizophrenia across countries (), and one of the reasons for this may be the cross-cultural differences in stigma ( Pescosolido et al., 2008 ). Therefore, understanding the cultural differences in attitudes toward schizophrenia is critical as it could lead to stigma reduction and subsequent improvement in course and outcome.

Studies comparing attitudes toward schizophrenia between Japan and other countries have revealed even more negative attitudes in Japan than in other countries (). Furnham and Murao (2000) compared British and Japanese lay theories of schizophrenia and found that the British believed individuals with schizophrenia to be less dangerous and abnormal than did the Japanese. Kurihara et al. (2000) contrasted public attitudes toward the mentally ill between the general public in Indonesia (Bali) and those in Japan using case vignettes, and observed that the Balinese held more favorable attitudes toward schizophrenia than the Japanese. Kurumatani et al. (2004) compared elementary school teachers' attitudes toward schizophrenia between Japan and Taiwan, showing that Japanese teachers were more likely than their Taiwanese counterparts to predict that parents and neighbors of individuals with schizophrenia would display stigmatizing attitudes toward such individuals. Griffiths et al. (2006) used case vignettes and compared stigma toward depression and schizophrenia between the general public in Australia and that in Japan, demonstrating more negative attitudes in the latter. To our knowledge, however, no reports have been published thus far that compare attitudes toward schizophrenia in Japan and those in the United States – countries that are comparable in terms of economic development as well as psychiatric treatment options. We recently conducted a web-based survey to investigate differences in attitudes toward schizophrenia between the general public and three groups of healthcare professionals (i.e., psychiatric staff, physicians and psychiatrists) among a Japanese population ( Hori et al., 2011 ). The main finding of this study was that attitudes of the general population and of physicians were equally stigmatizing, as compared to those of psychiatric staff and psychiatrists ( Hori et al., 2011 ).

In this context, the present study aimed to investigate the potential differences in attitudes of the general public and of medical specialists between the United States and Japan. To make cross-cultural comparisons possible, we herein apply the same web-based approach to the target population in the United States as the one used in our previous survey conducted in Japan. We hypothesized that negative attitudes of both the general population and medical specialists in Japan would be greater than those of each counterpart in the United States.

### 2. Methods

The present research was programmed into an Internet-based questionnaire format, using a web-based survey tool. A series of web-based surveys were conducted, first in Japan and then in the United States. The survey methods in these two countries were virtually identical. The Internet penetration rate (% population) in Japan and in the US is 80.0% and 78.3%, respectively, according to Internet World Stats ( http://www.internetworldstats.com/top25.htm ).

#### 2.1. Participants and procedures of the web-based survey in Japan

As detailed in our previous paper ( Hori et al., 2011 ), a Japanese version of the questionnaire was distributed to panel registrants of an online research panel service provided by Yahoo! Japan Research ( http://research.yahoo.co.jp ). Yahoo is one of the most popular internet service companies in Japan and in the US. The research panel consisted of a wide range of subpanels based on their occupations (manufacturers, farmers, construction workers, physicians, and so forth). Thus, using this subpanel and by asking specific questions, we were able to identify our target participant groups. For example, psychiatrists were identified by requesting participants in the “physicians” subpanel to specify the medical specialty to which they belonged. All the subpanels except for those of medical personnel were classified here as the general population. An invitation email entitled “Lifestyle survey” was sent to panel registrants on 30.05.2009. Participants were given a couple of days to return the questionnaire. On 1.06.2009, the internet research company provided data for the randomly selected 450 subjects. The data were sent to us in the format of a Microsoft Excel file without identifying information (e.g., name, birth date or email address). Of the 450 subjects, one subject who was enrolled as a psychiatrist was excluded because this person demonstrated subpar knowledge pertaining to schizophrenia. Three subjects enrolled as the general population and one additional subject enrolled as a psychiatrist were also excluded because they answered “I disagree” to all of the 18 items on the questionnaire of attitudes toward schizophrenia. Consequently, analyses were performed for 445 participants; 197 subjects in the general population, 100 psychiatric staff other than psychiatrists, 112 physicians other than psychiatrists and 36 psychiatrists. In the present cross-cultural study, data for 197 subjects in the general population and 112 physicians were compared with data for the US counterparts.

With respect to ethical issues, the website has been run by conforming to the personal information protection law as well as the ethical standards, and the panel registrants had voluntarily agreed to participate in web-based surveys at the time of their service registration. There was no possibility for leakage of personal information because the data have been anonymized from the outset. These ethical issues also applied to the US survey.

#### 2.2. Participants and procedures of the web-based survey in the United States

For the survey in the US, an English version of the questionnaire was distributed to panel registrants of an online research panel service provided by an overseas department of the Internet research service company in Japan (Yahoo Value Insight Corporation [now renamed to ‘Macromill, INC']; http://www.macromill.com ). We identified our target participant groups using the same method that was applied in the Japanese survey described above. An invitation email entitled “Lifestyle survey” was sent to panel registrants from 22.06.2009 to 30.06.2009. This email contained information about informed consent, point rewards and a hyperlink to the online survey. On 30.06.2009, the company provided data for the randomly selected 178 subjects in the general population and 47 physicians other than psychiatrists. At the outset of the present US survey, we had attempted to recruit approximately the same number of participants for each of the four panels as those in the Japanese survey. However, responses were obtained for only 19 psychiatric staff (other than psychiatrists) and one psychiatrist in the present US survey, which was due to the small size as well as low response rate for these panels in the US online research service. The sample sizes for these two groups had insufficient statistical power to detect a significant, if any, difference and were excluded from the analysis. Of these subjects, six subjects who were enrolled as the general population and two subjects enrolled as physicians were removed from our analysis as they answered either “I agree” or “I disagree” to all of the 18 items on the questionnaire of attitudes toward schizophrenia. Consequently, 172 subjects in the general population (male/female: 66/106) and 45 physicians (male/female: 20/25) were included in the analyses.

#### 2.3. Questionnaires

As in the Japanese survey ( Hori et al., 2011 ), the questionnaire used in the US survey was composed of three sections, namely demographic information, knowledge about schizophrenia, and the 18-item questionnaire on attitudes toward schizophrenia. Translation from Japanese to English was performed by two researchers (M.R. and H.H.), and the final version of the questionnaire was edited by staff of the Internet research company who had adequate English and Japanese reading and writing comprehension as well as sufficient experience with questionnaire surveys.

Demographic information of the potential participants included: gender, age, years of education, occupation/qualifications (e.g., manufacturers, farmers, physicians, nurses, and pharmacologists, and so on), specialty in medicine when relevant (e.g., Cardiology, Ophthalmology, Neurology, Psychiatry, etc.), self-reported location of living (i.e., Urban, Suburb, or Rural), household annual income (i.e., (1) Up to $20,000, (2)$20,000–39,999, (3) $40,000–59,999, (4)$60,000–79,999, (5) $80,000–99,999, (6)$100,000 +), experience of mental illness via family member or close friend (i.e., “Do you have family or close friends with a past history of psychiatric illness?”), experience of schizophrenia via family member or close friend when relevant (i.e., “Does that include individuals with schizophrenia?”), years of psychiatric education in medical school and/or in undergraduate/graduate studies, and number of books on schizophrenia he/she has read. To exclude those who have past or present psychiatric illnesses, we also asked the following question: “Have you ever been prescribed psychiatric medications, such as anxiolytics, hypnotics, antidepressants, antipsychotics, and/or anticonvulsants?” Those who answered “yes” to this question were excluded from the present study because persons who might have experienced mental illness are likely to hold different attitudes toward schizophrenia. Thus, we could not estimate the difference in prevalence of people who have experienced mental illness (or have been prescribed psychotropics) within the two countries.

Participants' knowledge about schizophrenia was surveyed by two questions, each with five choices (one correct choice and the other four wrong choices): “What is the typical age of schizophrenia onset? (1) Childhood, (2) Adolescence to early adulthood (correct answer), (3) Late adulthood, (4) Middle age, (5) ‘I don't know'” and “What is the characteristic symptom of schizophrenia? (1) Panic attack, (2) Visual hallucination, (3) Auditory hallucination (correct answer), (4) Obsessive-compulsive behavior, (5) ‘I don't know.'” Participants were instructed to answer these questions without referring to any educational materials. A question asking the prevalence rate of schizophrenia, which had been included in the Japanese survey, was excluded this time because it is unlikely that the general population is familiar with study results concerning prevalence and incidence rates of schizophrenia.

To evaluate attitudes toward schizophrenia, an 18-item questionnaire was administered. This questionnaire was based on the 13-item questionnaire developed by Uçok et al. (2006) . First, the 13-item questionnaire of Uçok et al. (2006) was translated into Japanese by two researchers (H.H. and H.K.). Then this Japanese version of the questionnaire was back-translated into English by another Japanese researcher. The back-translated English version of the questionnaire was sent to and approved by the original author (Prof. Uçok). Second, we added five items to the original 13-item questionnaire by referring to several prior studies on attitudes (), yielding the 18-item questionnaire that was used in the Japanese survey. Third, the newly added five items were translated into English by two researchers (M.R. and H.H.). Finally, the 18-item questionnaire, consisting of the 13 original items of Uçok et al. (2006) and the new five items, was used in the present survey in the United States. The full content of the 18-item questionnaire is presented in Table 2 . Since these 18 items included those that address either or both mental health literacy and stigma, we grouped the 18 items into these three categories: mental health literacy questions, stigma questions, and questions that touch upon both. We considered items #9, 11, 12, 15, 16, 17 and 18 as the mental health literacy questions, items #2, 5 and 6 as the stigma questions, and items #1, 3, 4, 7, 8, 10, 13 and 14 as the mixed questions ( Table 2 ). Participants were asked to answer each question with either “I agree” or “I disagree.” For items #1, 10, 11, 12, 15, 16, 17 and 18, those items answered with “I agree” were scored 1 and those answered with “I disagree” were scored 2. For the remaining items (#2, 3, 4, 5, 6, 7, 8, 9, 13 and 14), the scoring was reversed, i.e., “I agree,” 2 and “I disagree,” 1. Thus, for all items, the higher score indicated negative attitudes and/or lack of mental health literacy ( Table 2 ).

Table 1 Demographic characteristics and knowledge about schizophrenia of the four groups, stratified by occupation and country.

Characteristic JP_general population (n=197) JP_physicians (n=112) US_general population (n=172) US_physicians (n=45) Analysis
Statistics P
Gender, % female 45.7 10.7 61.6 55.6 χ2(3)=74.6 <0.001
Age, years: mean±S.D. 39.2±11.2 44.2±8.2 45.0±13.6 40.6±11.8 F(3,522)=9.2 <0.001
Education, years: mean±S.D. 14.5±2.4 19.4±2.4 15.3±2.8 21.4±2.2 F(3,522)=162.8 <0.001
Self-reported location of living, % urban/% suburb/ % rural 32.0/46.7/21.3 41.1/40.2/18.8 23.8/48.8/27.3 26.7/51.1/22.2 χ2(6)=10.8 0.09
Household annual income, rank: median (25–75 percentiles) 3.0 (3.0–5.0) 6.0 (6.0–6.0) 3.0 (2.0–4.0) 6.0 (4.0–6.0) Kruskal–Wallis χ2(3)=184.9 <0.001
Experience of mental illness via family member or close friend, % positive 21.1 15.1 30.2 35.6 χ2(3)=12.3 0.006
Experience of schizophrenia via family member or close friend, % positive 4.8 4.7 12.8 24.4 χ2(3)=21.2 <0.001
Psychiatric training, years: mean±S.D. 0.2±1.2 1.6±2.3 0.6±1.7 1.5±2.8 F(3,522)=17.6 <0.001
Number of books on schizophrenia: mean±S.D. 0.2±0.9 2.0±2.9 0.6±1.4 1.6±2.9 F(3,522)=28.7 <0.001
Knowledge about schizophrenia_Onset age, % correct 41.6 73.2 43.6 73.3 χ2(3)=41.7 <0.001
Knowledge about schizophrenia–Characteristic symptom, % correct 11.7 58.9 31.4 60.0 χ2(3)=90.5 <0.001
Number (0/1/2) of correct answers, % 54/39/7 17/34/49 46/33/21 18/31/51 χ2(6)=98.7 <0.001

Table 2 Percentage of subjects who answered “I agree” to each of the 18 items concerning attitudes toward schizophrenia, stratified by group.

JP_general population (n=197) JP_physicians (n=112) US_general population (n=172) US_physicians (n=45) US general vs. US physician JP general vs. US general JP physician vs. US physician
χ 2 P χ 2 P χ 2 P
1. Patients with schizophrenia can work (B) 54.3 82.1 83.7 91.1 1.5 0.21 36.5 <0.001 2.0 0.16
2. Would oppose if one of his/her relatives would like to marry someone who has schizophrenia (S) 71.1 76.8 46.5 68.9 7.1 0.008 23.0 <0.001 1.1 0.30
3. Schizophrenia patients can be recognized by his/her appearance (B) 16.8 24.1 8.7 11.1 0.2 0.62 5.2 0.02 3.3 0.07
4. Schizophrenia patients are dangerous (B) 38.1 30.4 44.2 31.1 2.5 0.11 1.4 0.23 0.01 0.93
5. Would not like to have a neighbor with schizophrenia (S) 41.6 41.1 40.1 53.3 2.5 0.11 0.1 0.77 2.0 0.16
6. Schizophrenia patients are not trustworthy (S) 38.1 36.6 30.2 31.1 0.01 0.91 2.5 0.11 0.4 0.51
7. Schizophrenia patients could harm children (B) 55.8 58.0 58.1 60.0 0.05 0.82 0.2 0.66 0.1 0.82
8. Schizophrenia patients should be kept in hospitals (B) 26.4 17.0 18.6 8.9 2.4 0.12 3.2 0.07 1.7 0.20
9. Family members of people with schizophrenia should help with all aspects of care (L) 38.6 28.6 75.6 71.1 0.4 0.54 51.0 <0.001 24.1 <0.001
10. (Suppose you were a psychiatrist) I don't worry about examining a person who is diagnosed with schizophrenia (B) 48.2 64.3 62.2 75.6 2.8 0.09 7.2 0.007 1.9 0.17
11. Would a patient with schizophrenia be treated in the appropriate department of the general hospital (L) 66.0 61.6 70.9 80.0 1.5 0.22 1.0 0.31 4.9 0.03
12. Schizophrenia can be treated (L) 80.7 81.3 91.3 95.6 0.9 0.34 8.3 0.004 5.3 0.02
13. Patients with schizophrenia cannot comprehend their illness (B) 34.0 30.4 39.5 28.9 1.7 0.19 1.2 0.27 0.03 0.86
14. Patients with schizophrenia cannot comprehend nor apply suggested treatment (B) 11.2 10.7 22.1 22.2 0.00 0.99 8.0 0.005 3.5 0.06
15. Schizophrenia has the chance of recovery (L) 89.8 83.9 59.3 64.4 0.4 0.53 46.5 <0.001 7.2 0.007
16. It is important to always inform a person with schizophrenia of their diagnosis (L) 63.5 66.1 83.1 84.4 0.04 0.83 17.9 <0.001 5.3 0.02
17. Patients with schizophrenia often benefit from pharmacologic intervention (i.e. antipsychotic medications) (L) 54.3 90.2 89.5 91.1 0.1 0.76 55.0 <0.001 0.03 0.86
18. Patients with schizophrenia often benefit from psychotherapy (L) 76.1 69.6 86.6 86.7 0.00 0.99 6.6 0.010 4.9 0.03

These five items were newly added to the original questionnaire of Uçok et al. (2006) .Underlined figures represent significant differences.

L: mental health literacy question, S: stigma question, B: mixed question

#### 2.4. Statistical analyses

Averages are reported as means±standard deviation (S.D.) for continuous variables and as medians (25–75 percentiles) for ordinal variables. For categorical variables, data are reported as percentages. Means, medians, and categorical variables were compared using the analysis of variance (ANOVA), Kruskal–Wallis test or Mann–Whitney U test, and the χ2 test, respectively. Correlations between continuous variables were calculated using Pearson's correlation coefficient. Spearman's rank order correlation was used to examine correlations between the number of correct answers concerning knowledge about schizophrenia and the factor scores of attitudes.

Considering that there were as many as 18 items in the questionnaire on attitudes, main analyses were conducted for smaller number of underlying factors in order to avoid the problem of multiple testing as well as to simplify the interpretation of results. For this purpose, an exploratory factor analysis was conducted using the principal axis factoring method with oblique (promax) rotation. The choice of the number of factors was based on theoretical meaningfulness as well as on the Kaiser criterion where variables with eigenvalues equal to or greater than one are extracted as factors. Items with factor loadings≥0.40 were deemed meaningful and assigned to the given factor, with only the highest factor loading for each item being considered. We labeled each factor based on what we believed best characterized the group of items that loaded on a particular factor. While the present study included general populations and physicians both in Japan and the US, we did not pool the sample across the specialty groups or ethnicities in the factor analysis because literature review indicates that such subgroups are not similar with respect to their attitudes toward schizophrenia (). Therefore, we decided to conduct the factor analysis separately in the Japanese general population and in the US general population. Then, by closely comparing the resultant factors between the two groups, we identified factors that were (almost) commonly seen across the groups and main analyses were conducted with these factors. Subsequently, raw scores of the items were averaged within each factor, which yielded a mean score of each factor for each subject. Since the raw score of each item (i.e., “1” or “2”) carried no meaning except that the lower score (i.e., “1”) and higher score (i.e., “2”) indicated positive and negative attitudes, respectively, the mean scores were normalized to the z-score using data of the combined general population groups of Japanese and Americans.

These normalized factor scores were analyzed using the two-way multivariate analysis of covariance (MANCOVA), with country (i.e., the United States vs. Japan) and occupation (i.e., the general population vs. physicians) as the between-subject factors and with potentially confounding demographic characteristics as the covariates. Of the basic demographic characteristics (i.e., age, gender, education and location of living), those variables which were significantly different (including trend level difference) between the four participant groups were considered as potential confounders. Statistical significance was set at two-tailed P<0.05. Analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 21.0 (SPSS Japan, Tokyo).

### 3. Results

Demographic characteristics and knowledge about schizophrenia, stratified by group, are shown in Table 1 . Since the basic demographics, including gender, age, education and location of living, were significantly different (or different at a trend-level) between groups, these four variables were controlled for in the MANCOVA examining differences in attitudes between the four participant groups.

#### 3.1. Factor analysis

As for the Japanese survey, results of the factor analysis were described in our previous study ( Hori et al., 2011 ). For reference purposes, the results are provided herein as Supplementary Table S1 .

The solution of factor analysis in the US general population is shown in Table 3 . Kaiser–Meyer–Olkin measure was 0.72, indicating high sampling adequacy for the factor analysis. Bartlett's test of sphericity, which tests whether the correlation matrix is an identity matrix, was significant (χ2=626.8, d.f.=136, P<0.001), indicating appropriateness of the factor model. The principal axis factoring method yielded six factors with eigenvalues greater than 1.0, explaining 62.1% of the cumulative variance. These six factors were retained according to the Kaiser criterion ( Table 3 ). Correlations among the six factors were not small enough (with the highest correlation being 0.62) to rationalize the orthogonal rotation, which assumes that each factor is totally independent of the other factors; thus, the oblique method used proved to be appropriate.

Table 3 Promax-rotated pattern matrix for the US general population (n=172)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6
1. Patients with schizophrenia can work       0.55
2. Would oppose if one of his/her relatives would like to marry someone who has schizophrenia         0.57
3. Schizophrenia patients can be recognized by his/her appearance     0.48
4. Schizophrenia patients are dangerous   0.74
5. Would not like to have a neighbor with schizophrenia       0.39 0.62
6. Schizophrenia patients are not trustworthy       0.32
7. Schizophrenia patients could harm children   0.85
8. Schizophrenia patients should be kept in hospitals       0.76
9. Family members of people with schizophrenia should help with all aspects of care
11. Would a patient with schizophrenia be treated in the appropriate department of the general hospital           0.32
12. Schizophrenia can be treated
13. Patients with schizophrenia cannot comprehend their illness     0.52
14. Patients with schizophrenia cannot comprehend nor apply suggested treatment     0.62
15. Schizophrenia has the chance of recovery           0.64
16. It is important to always inform a person with schizophrenia of their diagnosis 0.61
17. Patients with schizophrenia often benefit from pharmacologic intervention (i.e. antipsychotic medications) 0.78
18. Patients with schizophrenia often benefit from psychotherapy 0.75

Extraction method: Principal axis factoring method. Rotation method: Promax with Kaiser normalization.

Only factor loadings≥0.3 or≤−0.3 are shown. Factor loadings≥0.4 or≤−0.4 are retained for factor contribution (indicated with underline).

By comparing these two factor analytic results, we considered the following four factors as relatively common ones; “social distance” as the first factor (consisting of items #2 and #5), “belief of dangerousness” as the second factor (items #4 and #7), “underestimation of patients' abilities” as the third factor (items #13 and #14) and “skepticism regarding treatment” as the fourth factor (items #17 and #18). Consequently, the main analyses were conducted with these four factors.

#### 3.2. Knowledge about schizophrenia

As seen in Table 1 , the number of correct answers concerning schizophrenia knowledge was significantly different between the four participant groups. Compared to the Japanese general population, the US general population showed a significantly higher number of correct answers (Mann–Whitney U=14,625.0, P=0.012), while the Japanese physicians and the US physicians did not significantly differ in the number of correct answers (Mann–Whitney U=2488.5, P=0.89). Correlation between knowledge and attitudes (as indexed by the four factors) was calculated within the combined general population group, which revealed a significant negative correlation of the number of correct answers with the “skepticism regarding treatment” factor (Spearman's ρ=−0.15, P=0.004) while no significant correlations were found for the other three factors (all P>0.05).

#### 3.3. Effects of age and gender on attitudes toward schizophrenia

To investigate the possible effects of age and gender on attitudes toward schizophrenia, relationships of attitudes with age and gender were examined in each of the four participant groups. Age was significantly correlated positively with the “social distance” factor in the Japanese general population (r=0.19, P=0.008) and negatively with the “skepticism regarding treatment” factor in the US general population (r=−0.17, P=0.029), while no significant correlation was observed for the other 14 combinations (all P>0.2). Concerning gender differences, males in the Japanese general population showed significantly higher scores on the “underestimation of patients' abilities” factor than their female counterparts (t=2.78, P=0.006), while no significant differences were found for the other 15 combinations (all P>0.2).

#### 3.4. Comparisons of attitudes toward schizophrenia between groups

Results of the 18-item questionnaire on attitudes toward schizophrenia, stratified by occupation and country, are shown in Table 2 . The comparison between the general population vs. physicians in Japan was described in our previous report ( Hori et al., 2011 ). Fig. 1 shows the z-score on the four factors contrasting the four participant groups.

The two-way MANCOVA on the four factors, with country and occupation as the between-subject factors and with gender, age, education and location of living as the covariates, revealed a significant country-by-occupation interaction on the combination of the four factors [F(4,515)=3.2, Pillai's trace=0.024, P=0.013, partial η2=0.024]. When the interaction was examined separately for each factor, significant interactions were found for the “social distance” [F(1,518)=5.2, P=0.022, partial η2=0.010] and “skepticism regarding treatment” [F(1,518)=4.4, P=0.037, partial η2=0.008] factors. This MANCOVA revealed a significant main effect of country [F(4,515)=6.7, Pillai's trace=0.050, P<0.001, partial η2=0.050], but not occupation [F(4,515)=1.5, Pillai's trace=0.011, P=0.21, partial η2=0.011], on the combination of the four factors. Based on the general consensus that 0.01, 0.06 and 0.14 represent small, moderate and large partial η2, respectively ( Cohen, 1988 ), the effect of country was considered moderate. When the main effect of country was examined separately for each factor, a significant main effect was observed for the “underestimation of patients' abilities” [F(1,518)=5.7, P=0.018, partial η2=0.011] and “skepticism regarding treatment” [F(1,518)=15.6, P<0.001, partial η2=0.029] factors; compared to the Japanese groups, the US groups were more likely to underestimate the abilities of individuals with schizophrenia and less likely to hold skepticism regarding treatment. No significant main effect of occupation was found for any of the four factors (all P>0.1). This MANCOVA further revealed that, when the two countries were compared within each occupational group, the US general population showed significantly greater tendency for the “underestimation of patients' abilities” factor (estimated mean difference controlling for the four confounders: 0.28, 95% confidence interval: 0.07 to 0.49, P=0.009) and less tendency for the “social distance” (estimated mean difference: −0.38, 95% confidence interval: −0.59 to −0.17, P<0.001) and “skepticism regarding treatment” factors (estimated mean difference: −0.59, 95% confidence interval: −0.78 to −0.40, P<0.001) compared to the Japanese counterparts, whereas no significant between-country differences were observed in any of the four factors for the physician groups (all P>0.1). Similarly, when the two occupation groups were compared within each country, physicians showed significantly less tendency for the “skepticism regarding treatment” factors compared to the general population (estimated mean difference: −0.37, 95% confidence interval: −0.63 to −0.10, P=0.007) in the Japanese groups, while physicians showed significantly greater tendency for the “social distance” factor compared to the general population (estimated mean difference: 0.44, 95% confidence interval: 0.06 to 0.82, P=0.022) in the US groups.

### 4. Discussion

The main findings of this study can be summarized as follows: (1) Country, but not occupation, had a significant effect on the overall attitudes toward schizophrenia, (2) Compared to the Japanese general population, their US counterparts showed less social distance and treatment skepticism but more tendency for underestimation of patients' abilities, and (3) Physicians' attitudes were not significantly different between Japan and the US.

#### 4.1. Separable factors underlying attitudes toward schizophrenia

By considering the commonalities between the factor analytic results in the Japanese general population and those in the US counterparts, we identified “social distance,” “belief of dangerousness,” “underestimation of patients' abilities,” and “skepticism regarding treatment” as being the four underlying components of the 18-item questionnaire. It is noteworthy that both of the two items belonging to the first factor consisted of stigma questions, those belonging to the second and third factors comprised mixed questions, and those belonging to the fourth factor were composed of mental health literacy questions ( Table 2 ). Thus, the factor analysis proved to be useful in sorting out a number of questions pertaining to attitudes toward schizophrenia.

Similar factor analysis techniques have been employed in previous cross-cultural studies of attitudes toward schizophrenia. Furnham and colleagues have successfully used this approach to identify underlying factors in their questionnaires consisting of a number of items (). Their identified factors included those similar to the present ones, such as “dangerousness of people with schizophrenia,” “abnormality of schizophrenia,” “respect to people with schizophrenia” and “treatment for people with schizophrenia” ( Furnham and Chan, 2004 ). Kurumatani et al. (2004) also applied an exploratory factor analysis to their nine-item questions about how participants rate attitudes toward the vignette, which revealed “stigma” and “support” as the two underlying factors. These studies have demonstrated that factor analysis facilitates comparisons of attitudes between different cultural groups.

#### 4.2. Comparisons of attitudes toward schizophrenia between Americans and the Japanese

This is the first study to compare attitudes toward schizophrenia between people in the US and those in Japan, while there exist studies that contrast such attitudes of the Japanese with those of the British ( Furnham and Murao, 2000 ), the Balinese ( Kurihara et al., 2000 ), the Taiwanese ( Kurumatani et al., 2004 ), the Australian ( Griffiths et al., 2006 ) and the Chinese ( Haraguchi et al., 2009 ). These previous studies have generally argued that attitudes toward schizophrenia are more unfavorable in Japan than in other countries. The present finding of greater social distance in the Japanese public as compared to that in the US public may appear to confirm our hypothesis. Mental health service in Japan still relies relatively heavily on long-term hospitalized care, with data showing that the number of psychiatric beds per person being the highest rate in the world ( Oshima et al., 2003 ). This may result in more negative attitudes by leading to less contact to the mentally ill, given that contact with an individual with a mental illness is shown to decrease stigma ( Boyd et al., 2010 ); Conversely, as has been suggested (), there is also a possibility that the marked institutionalization in Japan might, in part, be a result of such negative attitudes by the public. It should be noted, however, that social distance should be considered in the context of the distance in general within a given society. It is commonly said that the social distance among Japanese people in general is greater than that among people in the US (). Assuming this, the present finding of greater social distance exhibited towards people with schizophrenia in Japan cannot be immediately interpreted as more stigmatizing than that of the US. Supporting this, similar extent of belief of dangerousness was observed across groups. Further investigations that take into account the social distance within a particular country are required to disentangle the negative attitudes toward schizophrenia.

The finding that US groups were less likely to hold skepticism toward treatment may reflect differences between chronic treatment approaches to schizophrenia between the two countries. While both countries apply similar diagnostic criteria and have access to similar psychotropic medications, long-term treatment strategies differ between the countries in that the US focuses more on community-based approaches. Given this, the US groups may have increased exposure to the rehabilitation process of treatment, seeing more long-term potential for community integration, while the Japanese tend to remain status quo and understand that long-term psychiatric hospitalization is the typical course with limited progression toward functional independence. On the other hand, US groups may need to underestimate the abilities of individuals with schizophrenia compared to the Japanese groups because such community integration often requires lifetime assistance. Viewed from another angle, it is shown that the Japanese have a tendency to feel that schizophrenia is not so much an organic brain condition, but more a reflection of weak character ( Tateyama et al., 1998 ). In this light, the Japanese groups may not necessarily underestimate the ability of patients per se, but they may resent their putative fragility, which may make them skeptical of treatment for a condition they do not view as real.

With respect to knowledge about schizophrenia, the present study revealed that the Japanese public had significantly less knowledge, as indexed by the number of correct answers, than the US public. This lack of knowledge about schizophrenia was reflected in the finding that approximately half of the Japanese public thought of antipsychotic medication as useful compared to approximately 90% of the US public (please see the item #17 in Table 2 ). Although it is obvious that a wide range of factors determine the public's attitude toward mental illness, negative attitudes toward schizophrenia may be due, at least in part, to their lack of knowledge about it, and the present findings may have confirmed this assumption. Indeed, scores of the “skepticism regarding treatment” factor, which were significantly correlated with less knowledge in the combined general population group, significantly differed between the Japanese vs. US public. That said, caution should be exercised before drawing the simple conclusion that stigmatizing attitudes can be reduced by education, because it is also possible that both more knowledge and less negative attitudes might stem from a common underlying factor such as his/her interest in mental illness. In addition, the notion that educating people about mental disorders could automatically lead to improvement of their attitudes toward people with mental illness has been challenged ().

#### 4.4. Conclusion

In summary, the present findings indicate that, while stigmatizing attitudes toward schizophrenia are prevalent in both Japan and the US, some important aspects of attitudes may be different between the general population in these two countries. Less social distance toward individuals with schizophrenia, more bias toward underestimation of the abilities of those individuals, and less skepticism regarding treatment observed among the US populations as compared to the Japanese counterparts may all be understood within the context of the cultural background of each country. The relatively small difference in attitudes between different occupations within each country also indicates that culture can play a major role in determining attitudes toward mental illness. These results emphasize the importance of culture in shaping understanding and acceptance of mental illness, demonstrating country-specific biases based on respective psychiatric practices. Anti-stigma campaigns, therefore, should be targeted toward culture-specific biases in order to have the greatest overall impact. More research on cross-cultural variations is required to fully understand the nature of negative attitudes toward schizophrenia.

### Acknowledgments

This study was funded by the Fulbright Foundation (M.R.), Grant-in-Aid for Young Scientists from the Japan Society for the Promotion of Science (JSPS) (H.H.), Health and Labor Sciences Research Grants (Research on Psychiatric and Neurological Diseases and Mental Health) (H.K.), Grant from Japan Foundation for Neuroscience and Mental Health (H.K.), and Grant-in-Aid for Scientific Research from the JSPS (H.K.).

### Appendix A. Supporting information

Supplementary data

### References

• M.C. Angermeyer, A. Holzinger, H. Matschinger. Mental health literacy and attitude towards people with mental illness: a trend analysis based on population surveys in the eastern part of Germany. European Psychiatry. 2009;24:225-232 Crossref
• J.E. Boyd, E.P. Katz, B.G. Link, J.C. Phelan. The relationship of multiple aspects of stigma and personal contact with someone hospitalized for mental illness, in a nationally representative sample. Social Psychiatry and Psychiatric Epidemiology. 2010;45:1063-1070 Crossref
• J. Cohen. Statistical Power Analysis for the Behavioral Sciences. second ed. (Lawrence Erlbaum Associates, Hillsdale, 1988)
• H. Fabrega Jr. The culture and history of psychiatric stigma in early modern and modern Western societies: a review of recent literature. Comprehensive Psychiatry. 1991;32:97-119 Crossref
• A. Furnham, E. Chan. Lay theories of schizophrenia. A cross-cultural comparison of British and Hong Kong Chinese attitudes, attributions and beliefs. Social Psychiatry and Psychiatric Epidemiology. 2004;39:543-552
• A. Furnham, M. Murao. A cross-cultural comparison of British and Japanese lay theories of schizophrenia. International Journal of Social Psychiatry. 2000;46:4-20 Crossref
• A. Furnham, L. Wong. A cross-cultural comparison of British and Chinese beliefs about the causes, behaviour manifestations and treatment of schizophrenia. Psychiatry Research. 2007;151:123-138 Crossref
• 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 Crossref
• 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 Crossref
• K. Hopper, J. Wanderling. Revisiting the developed versus developing country distinction in course and outcome in schizophrenia: results from ISoS, the WHO collaborative followup project. International Study of Schizophrenia. Schizophrenia Bulletin. 2000;26:835-846 Crossref
• 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. 2011;186:183-189 Crossref
• B. Hübner-Liebermann, H. Spiessl, K. Iwai, C. Cording. Treatment of schizophrenia: implications derived from an intercultural hospital comparison between Germany and Japan. International Journal of Social Psychiatry. 2005;51:83-96
• N. Kadri, N. Sartorius. The global fight against the stigma of schizophrenia. PLoS Medicine. 2005;2:e136 Crossref
• 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 Crossref
• N. Kuroda, S. Sun, C.K. Lin, N. Morita, H. Kashiwase, F. Yang, Y. Nakatani. Attitudes toward taking medication among outpatients with schizophrenia: cross-national comparison between Tokyo and Beijing. Environmental Health and Preventive Medicine. 2008;13:288-295 Crossref
• 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 Crossref
• T.S. Lebra. Mother and child in Japanese socialization: a Japan–US comparison. P. Greenfield, R. Cocking (Eds.) Cross-cultural Roots of Minority Child Development (Lawrence Erlbaum Associates, Hillsdale, 1994) 259-274
• A.A. Loch, M.P. Hengartner, F.B. Guarniero, F.L. Lawson, Y.P. Wang, W.F. Gattaz, W. Rössler. The more information, the more negative stigma towards schizophrenia: Brazilian general population and psychiatrists compared. Psychiatry Research. 2013;205:185-191 Crossref
• H.R. Markus, S. Kitayama. Culture and the self: implications for cognition, emotion, and motivation. Psychological Review. 1991;98:224-253 Crossref
• I. Oshima, Y. Mino, Y. Inomata. Institutionalisation and schizophrenia in Japan: Social environments and negative symptoms. Nationwide survey of in-patients.. British Journal of Psychiatry. 2003;183:50-56 Crossref
• B.A. Pescosolido, S. Olafsdottir, J.K. Martin, J.S. Long. Cross-cultural aspects of the stigma of mental illness. J. Arboleda-Florez, N. Sartorius (Eds.) Understanding the Stigma of Mental Illness: Theory and Interventions (John Wiley & Sons, Ltd, West Sussex, 2008) 19-35 Crossref
• G.W. Ruhnke, S.R. Wilson, T. Akamatsu, T. Kinoue, Y. Takashima, M.K. Goldstein, B.A. Koenig, J.C. Hornberger, T.A. Raffin. Ethical decision making and patient autonomy: a comparison of physicians and patients in Japan and the United States. Chest. 2000;118:1172-1182 Crossref
• N. Sartorius, W. Gulbinat, G. Harrison, E. Laska, C. Siegel. Long-term follow-up of schizophrenia in 16 countries. A description of the International Study of Schizophrenia conducted by the World Health Organization. Social Psychiatry and Psychiatric Epidemiology. 1996;31:249-258
• B. Schulze. Stigma and mental health professionals: a review of the evidence on an intricate relationship. International Review of Psychiatry. 2007;19:137-155 Crossref
• M. Tateyama, M. Asai, M. Hashimoto, M. Bartels, S. Kasper. Transcultural study of schizophrenic delusions. Tokyo versus Vienna and Tübingen (Germany). Psychopathology. 1998;31:59-68 Crossref
• G. Thornicroft, E. Brohan., D. Rose, N. Sartorius, M. Leese, INDIGO Study Group. Global pattern of experienced and anticipated discrimination against people with schizophrenia: a cross-sectional survey. Lancet. 2009;373:408-415 Crossref
• A. Uçok, H. Soygür, C. Atakli, K. Kuşcu, N. Sartorius, Z.C. Duman, A. Polat, S. Erkoç. The impact of antistigma education on the attitudes of general practitioners regarding schizophrenia. Psychiatry and Clinical Neurosciences. 2006;60:439-443

### Footnotes

a Department of Mental Disorder Research, National Institute of Neuroscience, National Center of Neurology and Psychiatry, 4-1-1, Ogawahigashi, Kodaira, Tokyo, 187-8502, Japan

b Semel Institute for Neuroscience and Human Behavior, University of California at Los Angeles (UCLA), Los Angeles, CA 90024, USA

c Fulbright Foundation, New York, NY 10025, USA

d Association for the Improvement of Mental Health Programmes, Geneva, Switzerland

Corresponding author. Tel: +81 42 341 2711; fax: +81 42 346 1744.