Factors Associated With Stigmatizationof Persons With Mental Illness

The stigmatization of persons with mental illnesses continues to be a primary deterrent to prevention and treatment efforts. A recent review of research on stigma documented discrimination against persons with mental illnesses in housing, jobs, and social interactions (1). Furthermore, empirical studies published in a special section of Psychiatric Services revealed that greater concern about stigmatization among persons with mental illnesses was associated with lower self-esteem (2), discontinuation of medications (3), and social impairment (4).

Researchers have most often assessed stigma associated with mental illness by surveying the public about attitudes toward "mental patients" or "persons with mental illness," terms that likely evoke images of chronic psychopathology. Consequently, it is unclear whether evidence of stigma is indicative of prejudice toward all mental illness or only its more severe forms. Although some studies have focused on the stigma associated with specific disorders (5,6), most researchers have chosen to blend subtypes into one catchall phrase.

Yet, with increased media coverage of precisely named psychological disorders such as "depression" and "bipolar disorder," it seems unlikely that the average adult construes mental illness in an all-encompassing manner. One purpose of the study reported here was to determine whether different types of mental illnesses are associated with different levels of stigma. Likewise, because diagnostic labels themselves may induce prejudice (7), we used a design that would allow us to distinguish stigma that results from abnormal behavior from stigma that results from the behavior's label.

Research on psychological correlates of stigmatization of mental illness has shown that an important—and surprising—predictor of stigma is greater belief in biological or genetic bases of mental illness (8). Phelan (9) theorized that genetic causes of mental illness are perceived as permanent, which raises discomfort because of the implication that complete recovery is impossible. Conversely, then, optimism about the effectiveness of treatment might be expected to reduce stigma. The final objective of our study was to test whether, in fact, attitudes toward treatment are associated with stigma.

The study had three goals. First, we compared levels of stigma associated with two different types of mental illness, schizophrenia and depression. Second, we explored the impact of psychiatric terminology on stigma. Third, we examined the relationship between beliefs about treatment and stigma, predicting that more optimistic attitudes toward treatment would be associated with lower levels of stigma.

Methods

The study was conducted from October through December of 2002. After receiving approval from our institutional review board, we recruited 116 undergraduate students (68 women and 48 men) from general education classes at a small public liberal arts university in the Southeast. A majority of the students were white (102 persons, or 88 percent), and their mean age was 21.7±5.4 years. After being informed of their rights as research participants, all students volunteered to complete the research survey in class.

The survey contained demographic questions as well as measures of stigma and attitudes toward treatment. Stigma was assessed by asking respondents to read vignettes about two individuals with mental illnesses: "Adrian," who had schizophrenia, and "Toby," who had depression. Following each vignette were six questions, adapted from a social distance scale (10), which measured respondents' comfort with Adrian and Toby in situations such as being a neighbor or dating, ranging from 1, very uncomfortable, to 5, very comfortable. The reliability for each scale was very good (Cronbach's alpha for each=.83).

To test the second hypothesis—that psychiatric terminology is associated with stigma—two forms of each vignette were developed, differing in the language used to describe Adrian and Toby. Form A contained diagnoses ("paranoid schizophrenia" and "major depressive disorder") and medical language (for example, "hallucinations" and "decreased appetite"), whereas form B contained only behaviorally descriptive language (for example, "sees things that are not there" and "rarely feels hungry"). Half the sample received form A, and the other half received form B.

The brief measure of attitudes toward treatment, developed for use in this study, consisted of three items: "There is little that can be done for a person with schizophrenia," "There are good treatments available to help people cope with depression," and "Most treatments for mental illness are painful and ineffective." Participants rated the items on a 6-point scale of agreement ranging from 1, strongly disagree, to 6, strongly agree. Cronbach's alpha was .58.

Results

The mean score on the stigma scale for the schizophrenia vignette was 17.4±4.3, and the mean score for the depression vignette was 19.87±4.51 (range of possible scores, 6 to 30). This difference was statistically significant (t=6.79, df=115, p<.001), indicating more negative perceptions of schizophrenia than of depression (higher scores indicate greater comfort). Table 1 shows the mean±SD scores for each item on the stigma scale.

Because there was a strong correlation between participants' stigma scores for the two vignettes (r=.60, p<.001), we combined the scores to produce a total stigma score for each participant (range of possible scores, 12 to 60). Total stigma scores were not significantly associated with age, race, or religion, but female respondents (35.52±8.41) had scores that indicated less stigma than did the scores of male respondents (38.5± 7.31, t=2.03, df=114, p<.05).

To examine the impact of psychiatric terminology on stigma, total stigma scores on form A (diagnostic and medical language) were compared with those on form B (behaviorally descriptive language). No significant difference was found.

Finally, to test for an association between attitudes toward treatment and stigma, total stigma scores were correlated with the treatment scale. The two measures were significantly related (r=.25, p<.01), indicating that persons with more positive expectations about treatment were more comfortable with persons with mental illness.

Discussion and conclusions

Our findings indicate that levels of stigma vary across categories of mental illness. Specifically, schizophrenia is more stigmatized than depression. Such results suggest that blanket terms such as "mental patient," which have dominated research on psychiatric stigma, are not useful. In previous research on this issue (6), a clever study manipulation underscored this point: participants correctly identified the diagnoses of individuals described in vignettes who had either depression or schizophrenia but did not necessarily label these same characters as having a "mental illness."

The knowledge that stigmatization of depression may not be as prevalent as commonly assumed may be encouraging news for persons who have this disorder, who are more likely to discontinue medication when they perceive high levels of stigma (3). Public reports of stigmatization of all mental illnesses are not only inaccurate but also potentially harmful, because they are likely to exaggerate the presence of stigma among many subgroups of mental illness, leading patients to believe stigmatization is more widespread than it actually is.

The relatively higher levels of stigmatization of schizophrenia point to the need for targeted intervention efforts. Educational programs to reduce stigma have been successful in the past but do little to improve attitudes about the controllability of mental illness (5). Armed with specific information about which disorders are most stigmatized and why, researchers may be able to design more specific interventions.

This study did not demonstrate an association between psychiatric terminology and stigma scores, which suggests that labels may be less stigmatizing than previously assumed. However, a stronger test of this issue would incorporate longer, more diverse vignettes and an older, more representative sample.

Study participants who had more positive attitudes about treatment demonstrated significantly less stigmatizing attitudes toward mental illness, which suggests that one way to decrease discomfort is to educate people about treatment possibilities. By emphasizing personal agency and improvement, this approach may be more effective than focusing on biologically based interventions, which may be associated with uncontrollability.

The female participants in our study reported significantly greater comfort than did the men toward persons with schizophrenia and depression. Because previous research findings on gender and stigma have been mixed, future studies in this area should attempt to clarify this relationship by varying the gender of the persons described in the vignettes (the names of the individuals in the vignettes we used were intentionally gender neutral) and surveying a gender-balanced sample.

Overall, our findings suggest that reducing the stigmatization of mental illness continues to be an important goal for mental health professionals. Although the generalizability of our findings may be limited by our reliance on student participants, it is likely that college students are in fact both more comfortable with mental illness and more concerned about social desirability than the general population, which means our results would underestimate overall levels of stigma. Future research should assess stigma associated with a wider variety of predictors and disorders. The more precise researchers can be about variables associated with stigma, the more effectively we can pursue our ultimate goal: the elimination of prejudice toward individuals with mental illnesses.

The authors are affiliated with the department of psychology at the University of North Carolina at Asheville. Send correspondence to Dr. Himelein at the Department of Psychology, CPO 1960, University of North Carolina, Asheville, North Carolina 28804-8508 (e-mail, [email protected]).

Enlarge table