Background
The first episode or prodromal symptoms of schizophrenia typically occur in adolescence [
1,
2]. However, stigma towards schizophrenia is one of the most important factors related to the obstruction of its early detection and treatment [
3,
4]. Therefore, it is vital to decrease stigma among parents of adolescents to contribute to the removal of this barrier [
5]. There have been many studies on demographic and social characteristics associated with stigma towards schizophrenia, including educational attainment, occupation, and previous contact with people with schizophrenia [
4,
6,
7]. However, the factors that create stigma among parents of adolescents have rarely been researched. A notable exception is a recent study by Yoshii [
8] that showed that the factors reducing the risk of stigma associated with schizophrenia in Japanese parents were family income, previous contact with people with schizophrenia, and participation in welfare activities for people with mental illness.
In recent decades, various educational programs have been developed aimed at reducing the stigma associated with schizophrenia. For example, a video about schizophrenia was played for 571 students from eight high schools across Canada [
9]. In another example, a video-based education program was implemented among 255 students from three middle schools in Hong Kong [
10]. In both cases, post-program tests found that students displayed significantly less stigma towards schizophrenia than before the program. Nonetheless, there has been no educational program specifically developed for the parents of adolescents, and hardly any studies on the factors associated with the effectiveness of educational programs. Because stigma reduction is beneficial for early detection and treatment for adolescents developing their first episode of schizophrenia, it is vital to expose parents to such programs. This study therefore evaluated the effect of a web-based educational program for parents of adolescents across Japan. Additionally, the factors associated with the effect were explored through multivariate analysis. The results should contribute to the more effective design of future educational programs.
Discussion
In this study, demographic and social factors were used to characterize the changes in levels of stigma towards schizophrenia, which were measured by Link’s Devaluation-Discrimination Scale before and after watching an internet-based educational program. Among parents of Japanese middle and high school students, mean stigma scores before and after watching the educational program on schizophrenia were identical (32.8 ± 4.4), and our analysis comparing the distributions indicated that the pre- and post- program distributions were not significantly different (Wilcoxon signed-rank test P-value = 0.176). The scores of pre- and post-program stigma found in the present study were similar to the stigma score reported by Berge
et al. (32.86 ± 6.22) [
19], suggesting that the result of this study measured by Link’s Devaluation-Discrimination Scale is a typical result with little fluctuation among the general population.
The design of the present study, in terms of both the delineation of the working sample and the choice of the dependent variable, facilitated the detection of new findings. Besides the parents with considerably-changed stigma scores, there were also some whose stigma scores changed slightly. These slight changes could be caused by either the educational program or random error. Consequently, if all these heterogeneous participants were studied simultaneously, it would be difficult to test the true educational effect of the program [
20]. We therefore studied only those parents who showed considerably increased or decreased levels of stigma towards schizophrenia. In addition, this study is different from common multivariate analytical studies that have used stigma level as the dependent variable. Instead, we have explored the risk factors associated with the effectiveness of stigma-change through education. As a result, several little-known factors associated with stigma scores emerged, including employment status. In addition, we were able to highlight other important factors that are easily overlooked, such as differences associated with work status, occupation, and levels of basic knowledge.
In the logistic regression analysis that included only demographic and social factors as independent variables, employment status and occupation were selected as significant factors associated with considerable changes in levels of stigma (comparing the odds of considerable decreases to the odds of considerable increases). To the best of our knowledge, this is the first time that the effect of employment status has been shown on stigma towards schizophrenia. The employment status described as “other” includes those who are self-employed, working in family-owned businesses, and unemployed. The overwhelming majority of this group was unemployed (66.4%, data not shown). Compared with those with “other” employment status, levels of stigma were significantly more likely to decrease among both full- and part-time employees. The degrees of freedom of social contact are relatively high among full- and part-time employees compared with the “other” group. In particular, unemployed and self-employed people have fewer opportunities to be in contact with different kinds of people and ideas. In contrast, full- and part-time employees, such as doctors, salesmen, lawyers, and civil servants, come into contact with a wide range of people in their professional lives. This result for employment status and occupation, then, could be seen as meaning that people with wider social and interpersonal relationships were more likely to be affected by the program in terms of reducing levels of stigma towards schizophrenia. A previous study on social contact and stigma indicated that people may become broader-minded and more accepting of difference through intergroup contact, with differences going beyond age and affinity orientation. It was also reported that requiring greater interpersonal disclosure may help to reduce stigma [
21].
The result on occupation type could be interpreted as a “stigma reduction” effect and a “degree of reduction difficulty” in different sectors. According to statistics of the Japanese Ministry of Health, Labour and Welfare [
22], the employment rate of mentally-disabled people is significantly lower than that of physically-disabled people (0.7 vs. 19.3%). Therefore, opportunities for contact with mentally-disabled people in the workplace are likely to be very limited. In addition, 70.7% of employed mentally-disabled people engage in the tertiary industry while only 29.3% engage in other industries. This indicates that those engaged in the primary and secondary industries would have hardly any contact with mentally-disabled people. Thirty-five percent of the tertiary industry was the medical sector, and previous research has reported low stigma among doctors and nurses [
23,
24]. Furthermore, workplaces in the tertiary industry are often offices, while other industries sometimes provide dangerous working conditions, such as a pelagic (tuna) fishing ship, a mine, or a welding factory, all of which often involve staff security issues. As reported in a study on the relationship between stigma and barriers to employment, people in industries using heavy machinery and dangerous equipment have higher levels of stigma against mental illness because of worries about the dangers to and of employees with mental illness [
25]. Therefore, the program about schizophrenia may have given the message to people engaged in these industries, with very little knowledge of or contact with mentally-disordered people, that they would have higher security needs around co-workers with mental illness. There is a possibility that their surprise and hesitation might subconsciously cause increases in stigma levels. It has been previously reported that gender differences were found in public attitudes towards mental disorder and that females showed more emotional concern about mental health problems than did males [
26,
27]. This might indicate that gender difference was one of the reasons why homemakers in this study, who are mostly female (96.9% homemakers were female, data not shown), showed lower levels of stigma towards schizophrenia.
Apart from demographic and social characteristics, other factors surveyed before watching the program were also used as independent variables in the logistic regression analysis. These factors included the pre-program scores on Link’s Devaluation-Discrimination Scale, basic knowledge about schizophrenia, social distance, and the ability to distinguish schizophrenia from other disorders. These measurements of schizophrenia-associated stigma obtained before the educational program were incorporated in a second model to predict the stigma changes caused by the program. Knowledge, pre-program score on Link’s Devaluation-Discrimination Scale, and social distance were included in the final logistical regression model as potentially-significant independent variables. This study demonstrated that the pre-program level of stigma was a significant explanatory variable. According to a previous study on anti-stigma education, lower levels of understanding and a more negative attitude towards mental illness allowed more potential for improvement following the program [
28]. It can therefore be inferred that parents with higher stigma levels before the program are more likely to view schizophrenia more positively after watching the program. This result is shown by the logistic regression presented in Table
3–2, and also by the distribution comparison of pre- and post-program Link’s Devaluation-Discrimination Scale scores shown in Table
2.
In this study, some parents did not change their attitudes towards schizophrenia. There are two possible causes of stigma not changing. One is the increase of stigma synchronized with the improvement of knowledge about schizophrenia. In this study, increases in levels of stigma tended to occur among parents who had more knowledge before watching the program. This is consistent with some previous studies that have indicated that stigma towards mental illness increases with increased knowledge [
29‐
32]. For example, research in 2013 by Loch
et al. reported that the more information people were given about schizophrenia, the more negatively they viewed the illness [
29]. The reason might be that many laypeople do not have the opportunity to connect with the real condition of schizophrenia. Therefore, these “knowledgeable” people actually know little about the daily life of people with schizophrenia after only reading written descriptions, and they consequently retain their original attitudes towards schizophrenia. The second possible reason is that the content of the educational program might make the audience hesitate, become discouraged, or feel hopeless because of the complexity and incomprehensibility of schizophrenia. This “side effect” of schizophrenia literacy has been confirmed by several recent studies [
31‐
33], and might explain why levels of stigma towards schizophrenia sometimes remain unchanged or worsen following an educational program [
34,
35]. Specially, as parents of adolescents confront this complicated illness, associating the symptoms with negative aspects of people with schizophrenia, their fear of the illness and worries about their children may be stimulated.
The above perspectives confirm that simple knowledge-imparting programs are not always effective in decreasing stigma. Similar results were also found in Hong Kong and the United States [
10,
36]. Alternatively, several studies have demonstrated that interpersonal contact with stigmatized people is a recognized effective strategy for reducing public stigma, indicating that the combination of knowledge and contact should be an effective method to intensify the educational effect of programs seeking exclusively to impart knowledge [
35,
37].
There are several possible methods of reducing levels of stigma. In line with the above discussion about the association between employment status and stigma changes and the importance of supplementing knowledge communication with contact for parents of middle and high school students, two types of anti-stigma interventions could be conducted in the future. One involves increasing contact in daily life with people with schizophrenia. Most laypeople lack opportunities to absorb knowledge and make personal contact with people suffering from schizophrenia, especially if their contact opportunities are limited by their employment status. Suitable ways of encouraging this kind of contact might be lectures by psychiatrists and people with or recovered from schizophrenia in community centers or parents’ meetings at school. The second type of anti-stigma intervention involves incorporating contact with people with schizophrenia within the educational program itself. It is possible to include interview videos of people who suffered from schizophrenia as adolescents and their parents, who could talk about the experiences of their actual daily lives and effective treatment methods. These positive examples could make parents hopeful regarding the treatment of schizophrenia and aware of the importance of early detection and treatment. Furthermore, improving the law and insurance systems in primary and secondary industries could make people more willing and relaxed regarding the possibility of working with people with schizophrenia, thus reducing the stigma among this population [
38‐
40].
This study also has several limitations, and some of the conclusions drawn here will lead to further debate. First, because the invitation to take part in the survey and the viewing of the educational program were conducted over the internet, parents who frequently used the internet were more likely to be included in the sample. Therefore, the sample in this study may not be representative of the wider population because of the nature of internet-based surveys, which makes the occurrence of non-response inevitable. Of the 5,000 candidates for participation in this study, 2,310 parents did not respond. For these parents, we were unable to obtain agreement to participate or any data regarding their characteristics, and it is possible there are some differences between respondents and non-respondents in terms of characteristics and attitudes. Second, regional cultural and psychological factors were not considered as explanatory variables of the effects. These would be a valuable consideration future studies. Third, the result of this study, with almost equal numbers of participants in the stigma-increased (n = 550) and stigma-decreased (n = 508) groups, underlines the fact that the theory of how educational efforts can affect stigma changes is in need of further development. However, the primary objective of this study was to explore the factors related to the characterization of considerably decreasing or increasing stigma using appropriate statistical analyses, and the results suggest that our educational program resulted in considerable decreases in stigma for participants with specific characteristics. This finding has encouraged us to conduct further studies. Concretely, we need to evaluate the reproducibility of the findings on the stigma-decreasing effect of our educational program for the Japanese parents of adolescents with certain characteristics. Additionally, it is necessary for us to investigate the cause of increasing stigma for the other parents in detail and to consider whether there are alternative strategies to prevent the considerable increases in stigma observed among some parents in this study. In particular, it is essential that future research give more deliberate consideration to strategies for anti-stigma education for parents with a variety of characteristics.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YL was the principal investigator and was responsible for the study concept and design, was involved in the data management, carried out the statistical analyses, and drafted the manuscript. MW was involved in drafting the manuscript. HY conducted the data collection and was involved in the data management. KA supervised the first author and was involved in interpreting the data and drafting the manuscript. All authors revised the text critically for important intellectual content and read and approved the final manuscript.