What do the general population know,believe and feel about individuals with autism and schizophrenia: Results from a comparative survey in Denmark

Abstract

Few studies investigate what members of the general population know about individuals with autism. Only one study has previously investigated how beliefs about autism differ from those about other psychiatric disorders. This study surveyed a convenience sample of the general adult population, within the Northern Region of Denmark, about their knowledge, attitudes and beliefs about individuals with autism and schizophrenia. The respondents (N = 440) possessed basic knowledge and were able to differentiate between the two disorders. Schizophrenia was associated with perceived danger (32.8%), while autism was associated with high intelligence (40.1%) and creativity (27.3%). Respondents were more positive towards interacting with individuals with autism (p < 0.001), but desire for social distancing was pronounced for both disorders in more intimate relationships. Significantly, more respondents reported that they would find it difficult and feel ashamed with regard to being diagnosed with schizophrenia (p < 0.001). Conversely, respondents significantly reported being more likely to elaborate on their diagnosis of autism to their colleagues (p < 0.001). Perceived dangerousness decreased the willingness for social interactions, while knowing someone with autism or schizophrenia increased the willingness to socially interact. Misconceptions and stigma must be addressed in order to minimize the social stigma and rejection associated with these disorders and ultimately improve the quality of life and psychological well-being of affected individuals and their families.

Keywords

adults autism spectrum disorders

Introduction

Misconceptions about mental illnesses have been documented in several studies for disorders such as schizophrenia, bipolar disorders and attention-deficit/hyperactivity disorder (ADHD) (Durand-Zaleski et al., 2012; Lebowitz, 2013) and are believed to cause social distancing and discrimination (Link and Phelan, 2014). A recent survey asking practitioners, parents and adults with autism about the topics they believe should be the focus of future autism research, identified a need for improving public awareness about autism. In particular, parents and practitioners called for action against stereotypes and common public perceptions about autism that perpetuate the media and the community. They called for more attention to the more problematic ‘other side of autism’ as a supplement to the frequently portrayed stories about children with exceptional mathematical skills or savants (Pellicano et al., 2014). These testimonies make the study of beliefs and misperceptions about mental illnesses, such as autism, an important area of research. Since mental illness covers a broad range of symptoms and behaviours, it is expected that the beliefs and knowledge will differ from one diagnosis to another.

In this study, we focus on the beliefs and knowledge about autism and schizophrenia within the general population. We chose to compare these two disorders for various reasons. First, both disorders have pervasive impact on the lives of the afflicted individuals. Second, the titles of these disorders are well-known among laypersons due to the exposure in news, popular films and documentaries. Third, the prevalence rates of the two disorders are fairly similar. Studies estimate the lifetime prevalence rates of 0.6% and 0.4% for autism spectrum disorder and schizophrenia, respectively (Elsabbagh et al., 2012; McGrath et al., 2008). Comparing knowledge about autism with, for example, knowledge about ADHD would potentially produce biased results, as the likelihood of knowing something about a disorder with a 5% prevalence rate such as ADHD could be greater than for a condition that affects only 0.6% of the population. Fourth, despite there being great differences between the two disorders on factors such as the age of onset of the illness, symptoms and prognosis, there are; however, a number of similarities between autism and schizophrenia such as social impairment and the aetiology of both disorders being substantially influenced by genetic factors. Comparing perceptions of these two disorders will highlight whether members of the general population can distinguish the two disorders from each other and could potentially reveal beliefs unique to one of the two disorders.

We conducted a literature review on the topic to inform the study and survey design. Only two studies were identified investigating the perceptions of the general public about autism, while 25 studies were identified for schizophrenia. We briefly review the literature below.

Symptoms

There was a general paucity in the studies investigating which symptoms or behaviours the general population believe to be associated with autism. One French survey conducted in 2009 by Durand-Zaleski et al. (2012) found that 7% of their sample believed individuals with autism to be dangerous (Durand-Zaleski et al., 2012). This study did not investigate the respondent’s knowledge about other aspects of behaviours associated with autism, and no studies to our knowledge have investigated this issue. On the other hand, surveys regarding beliefs about symptoms and behaviours associated with schizophrenia are more frequent. Survey studies conducted in Greece, Germany and Buenos Aires between the years 2001 and 2007 demonstrated that 55.9%–67.4% of the survey samples were able to identify auditory hallucinations as a symptom of schizophrenia. However, the same studies reported that their respondents also incorrectly associated a number of symptoms with the disorder, for example, 44.4%–81.3% believed split personality to be a symptom of schizophrenia (Economou et al., 2009; Gaebel et al., 2002; Leiderman et al., 2011). Studies have also investigated the general population’s views regarding perceived danger or violence posed by individuals with schizophrenia. Compared to the finding for autism, the belief about the dangerousness of individuals with schizophrenia was more frequent and studies have reported that 18.2%–73.1% believe that individuals with schizophrenia are dangerous or violent (Bag et al., 2006; Crisp et al., 2000; Durand-Zaleski et al., 2012; Gaebel et al., 2002; Grausgruber et al., 2007; Griffiths et al., 2006). These surveys were conducted in France, Turkey, United Kingdom, Germany, Austria, Australia and Japan during the years 1998–2009.

Knowledge regarding symptoms varied considerably from study to study. The differences potentially reflect variations in mental health literacy from country to country, differences related to when the surveys were conducted and potential differences in methodology. While some studies used a simple questionnaire form only mentioning the mental disorder (e.g. Durand-Zaleski et al., 2012), others used vignettes depicting an individual with a mental disorder, before asking questions about knowledge and attitudes (e.g. Crisp et al., 2000). These differences in methodology may have primed respondents differently and led to variations in responses.

Aetiology

Studies have also investigated layperson’s knowledge about the aetiology of autism and schizophrenia. A majority of respondents in the two identified surveys on autism conducted in France in 2009 and the United States in 2010 believed biological mechanisms, such as heritability (33%–44%), to be the primary cause of autism (Durand-Zaleski et al., 2012; Holt and Christensen, 2013). Beliefs that parent–child interactions and stressful life events are also causes of autism existed in 22%–23% (Durand-Zaleski et al., 2012). In surveys on schizophrenia, psychosocial factors were more frequently highlighted as primary causes for development of schizophrenia, compared to the findings for autism. In a series of studies from France, Greece, Buenos Aires, Australia, the United States and Canada conducted in the years 1997–2010, 10%–94.5% believed alcohol and substance abuse, unstable social environments and childhood problems to be the primary causes of schizophrenia (Durand-Zaleski et al., 2012; Economou et al., 2009; Leiderman et al., 2011; Reavley and Jorm, 2014). A recent study from Australia comparing attitudes and knowledge about schizophrenia in 1995, 2003 and again in 2011 found an increasing tendency towards highlighting genetic factors and childhood problems as the causes of schizophrenia. However, beliefs about intrapsychic factors, such as weakness of character, as causes of schizophrenia appeared to decrease (Reavley and Jorm, 2014). Again, variability in the methodology, time periods and locations of the surveys could explain the large discrepancies in findings.

Social distance

A series of studies have also investigated how the general population feel about being in contact with individuals with schizophrenia. However, there are no such studies examining this for autism. These studies were conducted in Turkey, Greece, Australia, Japan, Germany, United Kingdom, Pakistan and Canada during the years 2001–2011. The findings indicate that the preference for social distance increases with the degree of social intimacy, for example, 45.2%–93% expressed unwillingness to marry a person with schizophrenia (Bag et al., 2006; Economou et al., 2009; Griffiths et al., 2006; Reavley and Jorm, 2012), 29.9%–72% were unwilling to enter into any sort of work-related relationship with a person with schizophrenia (Bag et al., 2006; Griffiths et al., 2006; Reavley and Jorm, 2012) and finally, 19.7%–22.8% expressed unwillingness to befriend a person with schizophrenia (Gaebel et al., 2002; Reavley and Jorm, 2012). Again, discrepancies in findings across studies are most likely related to methodological, temporal and contextual characteristics of the studies.

Some studies indicated that gender (Economou et al., 2009; Gaebel et al., 2002) increased knowledge regarding schizophrenia and having social acquaintances with individuals with schizophrenia reduced the need for social distancing (Furnham et al., 2008). On the other hand, beliefs about dangerousness increased the need for social distancing (Grausgruber et al., 2007; Norman et al., 2012). A recent meta-analysis found that bio-genetic beliefs about the causes of schizophrenia showed a weak but significantly negative correlation with layperson’s willingness to socially interact with individuals with schizophrenia (Kvaale et al., 2013).

Aims

It is clear from the literature review that there is a great paucity in studies investigating the general public’s attitudes and knowledge regarding individuals with autism. This is problematic, as individuals with autism might experience and suffer the emotional and social consequences of social stigma just as often as individuals with schizophrenia. Thus, this subject deserves to be addressed and investigated further. Based on the literature review and our experience, our priori hypotheses were that respondents would tend to hold more positive attitudes towards individuals with autism, and we believed that respondents to a greater extent could support biological hypotheses about aetiology of autism, compared to schizophrenia. We hypothesized that the potentially divergent beliefs would have an impact on the willingness of respondents to engage in social interactions with affected individuals.

The aims of the present survey were to (1) investigate layperson’s perceptions of autism and schizophrenia, (2) investigate similarities and differences between the beliefs about the two disorders and (3) investigate how beliefs about the disorder and socio-demographic factors might influence negatively as well as positively on the acceptance of socially interacting with individuals suffering from autism or schizophrenia.

Materials and methods

Sample

The survey was carried out between March and April 2014 among inhabitants of the Northern Region of Denmark. All participants were aged 18 years and older. A non-probability sampling method was used. The respondents were sampled utilizing two methods: 233 complete questionnaires were collected in pen-and-paper format in shopping centres, a library and on the street, and 207 questionnaires were completed and collected utilizing a web-based questionnaire distributed on the social media platform Facebook, only including individuals living in the Northern Region of Denmark. For purposes of analysis, we decided to exclude questionnaires where (1) information on demographic questions was missing on one or more questions or (2) more than three questions had not been answered in the remaining sections. Out of the 256 questionnaires collected in pen-and-paper format, we excluded 23 because of missing information. We believed the omission of responses to be random errors, since no clear pattern emerged in which questions and responses had been omitted. The software used to develop the online questionnaire included validation options, and thus, only 2 out of 209 questionnaires were excluded due to participants not completing the survey.

Recruitment

Individuals were either asked online or on the street about whether or not they wished to participate in an anonymous survey investigating the knowledge about schizophrenia and autism in Northern Jutland. An identical declaration of consent describing the purpose of the study was available as the first page of both the online and pen-and-paper version of the questionnaire. Apart from the technical validation option in the online questionnaire, questionnaires filled out online and in pen and paper were identical. No diagnosis was emphasized over the other, to avoid attracting respondents with a specific interest in either schizophrenia or autism. The pen-and-paper format of the questionnaire was handed out to respondents along with a clipboard and pen. The respondents filled out the questionnaire in the street of the shopping centre where they were recruited, without the interference of surveyors.

Survey

For this study, we decided to develop a new questionnaire. To our knowledge, only one study (Durand-Zaleski et al., 2012) had compared laypeople’s knowledge and beliefs about autism and schizophrenia, but we found that the questionnaire used in the French study was too short to answer our research questions. While the questionnaire used in this study was not validated in terms of measuring test–retest reliability and so on, we did ensure the face validity by testing the questionnaire in a small pilot study among acquaintances not working within psychiatry or psychology. In the pilot study, the questionnaire was filled out and we received oral feedback about the perceptions and misunderstandings of the individual items. This feedback was used to correct individual items. To support the content validity, individual items were inspired from the literature on especially knowledge and beliefs about schizophrenia, the diagnostic criteria for both disorders in International Classification of Diseases, 10th Revision (ICD-10) and published studies about age of onset and aetiology (see below). We prioritized to use a combination of binary response options and Likert scales to engage participants throughout the questionnaire.

The final survey consisted of three main sections. In the first section, demographic data were collected. In the second section, 24 questions were regarding the knowledge/beliefs about the symptoms/behaviours, level of intelligence, primary aetiology, age of onset and prognosis of autism and schizophrenia. First, respondents were asked to identify behaviours or characteristics from a list that they believed to characterize individuals with autism and schizophrenia. Respondents could tick off as many items as they wished to. We characterized the responses as either in accordance with ICD-10 diagnostic criteria or not. Behaviours in accordance with diagnostic criteria in the ICD-10 for schizophrenia covered both positive and negative symptoms; social withdrawal, delusions, abnormal/locked bodily positions, neglect of hygiene and hallucinations. We categorized social withdrawal as a behaviour in accordance with the ICD-10 classification, while difficulties establishing social relationships was coded as a non-characteristic belief, since individuals with schizophrenia do not necessarily have an inherent social handicap as is seen for autism spectrum disorder. However, respondents might have had difficulties distinguishing the subtle differences of these two statements.

For autism, the following were rated as in accordance with ICD-10 criteria: delayed language development, great knowledge within a circumscribed area, diminished ability to understand own and others feelings, difficulties establishing relationships to others, special routines or habits and special interests. These symptoms thus covered behavioural manifestations of communication difficulties, social impairment and repetitive behaviours.

In addition, we listed three negative items that were not characteristic of either of the disorders: ‘dangerous to self and others’, ‘lacks self-control’ and ‘has a split personality’. Finally, two positive items were as follows: ‘creative’ and ‘caring’. We aimed at balancing positive and negative characteristics in order to avoid priming subjects in either direction. These characteristics were intended to be a starting point for exploring potential characteristics associated with particularly autism, given the lack of studies on this topic (see ‘Introduction’ section).

Next, we asked respondents if they believed individuals with schizophrenia and autism to be less, equally or more intelligent than the average person, and we asked when respondents believed the two disorders to be developed in childhood, adolescence, young adulthood, adulthood or old age. It is generally accepted that autism typically onsets in childhood (Atladottir et al., 2015), and although schizophrenia in some instances has a childhood onset, adolescence and young adulthood are the developmental ages where schizophrenia most frequently has its incidence (Sutterland et al., 2013). On a 5-point Likert scale (agree, partly agree, neutral, partly disagree and disagree), we measured the respondents beliefs about the prognosis of autism and schizophrenia. Finally, we asked participants to rate what they believed to be the primary causes of autism and schizophrenia with options covering biological, environmental and psychological factors.

In the third section, we asked the respondents five questions about their feelings about engaging in social interactions of varying degrees of proximity and intimacy with people with autism and schizophrenia and, furthermore, five questions investigating the stigma/self-stigma people would experience whether either they or family members were diagnosed with autism or schizophrenia. For the social interaction questions, the participants rated their attitudes on a 5-point Likert scale (I would feel very okay/okay, neutral, I would not feel okay/not feel okay at all). In the analyses, we decided to collapse the response into three categories ‘very ok/ok’, ‘neutral’ and ‘not okay/not okay at all’ for ease of interpretation and presentation. For the stigma/self-stigma questions, respondents were asked to answer the questions yes/no. The study used a dependent sample design allowing for comparisons of differences in beliefs, knowledge and attitudes about autism and schizophrenia within each responder. The raw data are available from the authors upon request.

Ethics

Since the survey did not ask participants for sensitive personal information that could potentially identify individual respondents, participation was anonymous, all collected questionnaires contained a declaration of consent and all individuals were above the age of 18 years (the legal age in Denmark). And participants did not receive any incentive for participation. The study did not have to be approved by an ethical committee or approved by the data protection agency.

Statistical analyses

Data were primarily analysed descriptively. In order to detect whether people held different knowledge, beliefs and attitudes for the two disorders, differences were analysed using the McNemar test for paired nominal data and paired t-test and Wilcoxon signed-rank test for paired ordinal data. We calculated the percentage of behaviours recognized by the respondents that were in accordance with the diagnostic criteria of the ICD-10. In addition, we performed multiple linear regression analyses to identify whether any factors were associated with larger percentages of recognized symptoms. To identify whether any factors were predictive of percentage recognized symptoms of autism/schizophrenia and the willingness to socially interact with people with autism or schizophrenia, data were analysed using logistic regression analysis with both enter and stepwise methods. Alpha levels were set at 0.05 and data were analysed using STATA 12th edition and SPSS 22nd edition (IBM Corp, 2013; StataCorp LP, 2011).

Since the final sample was non-representative of the population of the North Region of Denmark on gender, age and educational level and occupational status, data were weighted on gender to approximate population parameters. No weighting was done on age due to small frequencies of participants in older age groups. No census data were available on the educational achievement or occupational status of inhabitants of the Northern Region of Denmark, and thus, no weighting was done on these parameters. To avoid the possible confounding factors of age, educational achievement or occupational status regression, analyses were controlled for these variables.

Results

Sample characteristics

A total of 440 respondents completed the survey. Descriptive information of the sample is available in Table 1. After weighting data, the male–female distributions were equal to the population distributions within the Northern Region of Denmark. The study oversampled participants of younger age, and thus, the median age was 25 years with an interquartile range (IQR) of 22–43. A large proportion of respondents knew someone with either autism (49.9%) or schizophrenia (33.8%). A total of 1% (n = 7) declared to have autism and 0.7% (N = 3) to have schizophrenia. This prevalence is similar to population-based estimates of the prevalence of both disorders.

Table 1.

Sample characteristics.

	N	%
Sex
Male	222	50.4
Female	218	49.6

	Median	IQR


Age	25	22–43


	N	%

Highest education
Primary/secondary school	27	6.1
Vocational education	25	5.8
High school	195	44.2
Short further education (1–2 years)	28	6.3
Medium further education (2–4 years)	108	24.6
Long further education (5+ years)	42	9.5
Other	15	3.4
Employment
Student	216	49.2
Working	117	26.7
Retired	54	12.3
Unemployed	29	6.7
Other	22	5.1
Knows someone with
Schizophrenia	148	33.8
Autism	219	49.9
Respondents with
Schizophrenia	3	0.7
Autism	5	1.0

IQR: interquartile range.

Beliefs and knowledge about individuals with autism and schizophrenia

The respondents showed recognition of the symptoms/behaviours characteristically for people with autism according to ICD-10 criteria with 54.4%–76.5% of the respondents choosing these items (Table 2). The negative non-characteristic behaviours for people with autism were not frequently selected by respondents, but 27.3% believed people with autism to be creative and 40.1% believed them to be more intelligent than non-affected individuals. Most respondents supported that autism is primarily caused by biological factors such as hereditability (38.3%) or caused by brain damage (42.7%). A total of 90% identified that autism has its onset during childhood. Compared to the respondents’ beliefs about people with schizophrenia, less believed that people with autism require lifelong support and care (64.1%), and less believed that autism can be outgrown (19%). More believed that people with autism, compared to individuals with schizophrenia, could have jobs, although differences were small. Finally, 73%–75% believed that people with autism can lead normal lives and live independently at a rate statistically non-distinguishable from the beliefs about people with schizophrenia (Table 2).

Table 2.

Beliefs/knowledge about people with schizophrenia and autism.

	Autism		Schizophrenia		x²	p
	N	%	N	%	x²	p
Behaviours/symptoms
Withdraws from social life	310	70.4	170	38.7	106.4	<0.001
Delayed speech development	155	35.1	12	2.7	172.7	<0.001
Dangerous to self and others	17	3.9	144	32.8	147.2	<0.001
Exceptional great knowledge about specific topics	336	76.5	20	4.6	403.8	<0.001
Delusions (e.g. believes their being watched)	12	2.7	317	72.1	394.4	<0.001
Impaired ability to understand own and the feelings of other	278	63.2	101	22.9	182.9	<0.001
Lacks self-control	40	9.2	73	16.5	13.3	<0.001
Difficulties establishing relationships to other people	315	71.6	165	37.6	127.7	<0.001
Creative	120	27.3	28	6.3	90.0	<0.001
Split personality	11	2.5	312	70.9	376.7	<0.001
Abnormal, locked body positions over longer periods of time	49	11.0	20	4.6	16.7	<0.001
Caring	30	6.8	4	1.0	27.2	<0.001
Specific routines and habits	317	72.1	69	15.8	300.0	<0.001
Neglect of hygiene	29	6.6	39	9.0	1.9	0.171
Specific interests	240	54.4	24	5.4	273.7	<0.001
Hallucinations (e.g. sees/hears things that are not there)	7	1.7	322	73.4	416.0	<0.001

	N	%	N	%	z	p

Intelligence
More intelligent	173	40.1	9	2.1	−8.1	<0.001
Normal intelligence	173	40.1	256	81.5
Less intelligent	52	12.0	31	7.0

	N	%	N	%	x²	p

Primary aetiology
Hereditary	161	38.3	154	36.7	0.9	0.351
Brain damage (e.g. lack of oxygen during birth)	179	42.7	33	7.9	57.5	<0.001
Unstable upbringing with lack of proper care	17	4.1	87	20.9	72.3	<0.001
Contagion	0	0.0	2	0.4	–	–
Stressful life events (e.g. losing a family member/loosing job, etc.)	10	2.3	75	18.0	75.3	<0.001
Malnutrition (e.g. consumption of an excessive amount of sugar)	3	0.7	1	0.2	–	–
Lack of exercise	0	0.0	0	0.0	–	–
Alcohol/substance abuse during pregnancy	10	2.3	15	3.6	1.1	0.296
Alcohol/substance abuse by the person with the disorder	2	0.4	31	7.5	33.6	<0.001
Lack of will power	2	0.5	4	0.9	–	–
Age of onset
Onsets in childhood (age 0–12 years)	395	90.0	68	15.6	422.5	<0.001
Onsets in adolescence (age 13–20 years)	36	8.2	219	50.0	206.0	<0.001
Onsets in young adulthood (age −30 years)	26	6.0	168	38.4	176.8	<0.001
Onsets in adulthood (age 31–64 years)	7	1.5	37	8.6	37.2	<0.001
Onsets late in life (age 65+ years)	10	2.2	11	2.6	0.1	0.687
Prognosis
Needs lifelong care	282	64.1	333	75.6	14.3	<0.001
Can live a normal life	324	73.5	338	76.9	1.1	0.295
Can live alone	331	75.2	315	71.6	0.8	0.369
Can have a regular job	306	69.4	277	62.9	7.1	0.008
Can outgrow diagnosis	84	19.0	145	32.9	7.3	0.007

The respondents frequently recognized three of the listed symptoms of schizophrenia (social withdrawal, suffers from delusions and hallucinations) with 38.7%–73.3% of the respondents identifying them as characteristics of schizophrenia. However, attributes, not necessarily characteristic of individuals, with schizophrenia were also frequently chosen, for example, 32.8% believed that people with schizophrenia to be dangerous to themselves and others, 37.6% believed that people with schizophrenia have difficulties in establishing social relationships and 70.9% believed that people with schizophrenia to be suffering from the split personality phenomenon.

The majority (81.5%) believed people with schizophrenia to be of normal intelligence (Table 2). A total of 36.7% endorsed that schizophrenia is primarily a hereditary condition, 20.9% believed the primary reason to be the result of an unstable upbringing and 18.0% believed that stressful life events is the primary cause of schizophrenia. Thus, most respondents supported a primarily psychosocial aetiology of schizophrenia. The majority believed schizophrenia to be a disorder with an adolescent/young adult onset. A minority (32.9%) believed that schizophrenia can be outgrown but the majority believed that people with schizophrenia can lead normal lives with independent housing and participation in the work force, but also that most will require lifelong support and care (Table 2).

A higher percentage of recognition of behaviours in accordance with ICD-10 criteria was identified for autism (mean (M) = 62.1, standard deviation (SD) = 25.6) compared to schizophrenia (mean = 39.4, SD = 21.3) (t = −18.5, degree of freedom (df) = 329, p < 0.001). As the descriptive analyses were identified, more respondents knew individuals with autism than individuals affected by schizophrenia. Therefore, we ran post-hoc paired t-tests for individuals who knew or did not know individuals with autism separately. These analyses identified that irrespective of knowing a person affected with autism, more respondents identified a higher percentage of symptoms of autism compared to schizophrenia (does not know someone with autism – M = 58.0, SD = 25.9 vs M = 38.7, SD = 20.5, t = −10.8, df = 218, p ⩽ 0.001; knows someone with autism – M = 66.3, SD = 24.7 vs M = 40.2, SD = 22.2, t = −15.9, df = 218, p < 0.001). For the schizophrenia-related items, being male decreased the percentage recognized symptoms (B = −7.6, 95% confidence interval (CI) = −11.8 to −3.5). Having further education (B = 7.4, 95% CI = 2.6−12.2), knowing someone with schizophrenia (B = 6.1, 95% CI = 2.0–10.3) and believing individuals with schizophrenia to be dangerous to self or others was associated with a higher percent identification of symptoms consistent with ICD-10 criteria (B = 7.3, 95% CI = 3.3–11.4). For the autism-related items, increasing age (B = −0.4, 95% CI = −0.6 to −0.2) and male gender (B = −7.7, 95% CI = −12.5 to −2.9) was negatively associated with the percentage identified symptoms of autism. Having further education (B = 6.9, 95% CI = 1.4–12.5) and knowing someone with autism (B = 8.2, 95% CI = 3.7–12.8) was associated with a higher percent identification of symptoms concordant with ICD-10 criteria for autism spectrum disorder.

Desire for social distance

Figure 1 depicts the respondent’s feelings towards various social interaction situations. Having a boss or being romantically involved with a person with autism/schizophrenia were the categories that the respondents felt less happy with. Although the observations were similar for the two disorders, a significantly greater proportion of individuals rated that they would feel very ok/ok about interacting with people with autism compared to interacting with people with schizophrenia (conversation: t = 5.9, df = 439, p < 0.001; neighbour: t = 12.1, df = 439, p < 0.001; colleague: t = 10.3, df = 437, p < 0.001; friendship: t = 13.3, df = 438, p < 0.001; boss: t = 8.5, df = 436, p < 0.001; romantic relationship: t = 5.6, df = 436, p < 0.001).

Figure 1.

Attitudes towards being in close to distant social interactions with individuals with schizophrenia/autism.

What predicts willingness to socially interact with people with autism and schizophrenia?

Females were in all domains more willing to engage in social interactions with a person with autism, except for having a boss or being romantically involved with a person with autism, where no effect of gender was identified. Knowing a person with autism increased the willingness for the respondents to being positive towards having friends and colleagues with autism. In order to investigate how the gender of the respondent was related to willingness to socially interact with individuals with autism, a series of post-hoc stepwise logistic regression analyses were run. The results indicated that before entering gender as a covariate in the model, knowing someone with autism predicted positive attitudes, but including gender in the model made knowing someone with autism a non-significant predictor in the ‘conversation’ and ‘neighbour’ items, suggesting that more females know people with autism. Post-hoc chi-square analyses supported this hypothesis (x² = 8.2, p < 0.001). Age, educational attainment, occupation and beliefs about the aetiology of autism were non-significantly related to willingness to socially interact. As the questions about willingness to socially interact were believed to be especially sensitive towards expectancy bias, we included mode of response in the model (pen-and-paper format vs online). For all social outcomes, this variable was non-significant (Table 3).

Table 3.

Factors associated with willingness to socially interact with people with autism or schizophrenia.

		A conversation	Neighbour	Colleague	Friendship	Boss	Romantic relationship
		OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Schizophrenia	Gender (ref: male)	2.0 (1.2–3.3)	ns	ns	ns	ns	ns
	Age	ns	ns	ns	1.0 (1.0–1.0)	ns	ns
	Educational attainment (ref: no further education)	ns	ns	ns	ns	ns	ns
	Occupation (ref: being a student)	ns	ns	ns	ns	ns	ns
	Knows someone with SZ (ref: does not know any with SZ)	2.3 (1.3–4.1)	2.4 (1.6–3.7)	2.1 (1.4–3.2)	1.7 (1.1–2.6)	2.1 (1.3–3.5)	2.4 (1.2–4.7)
	Biological aetiology (ref: primarily biological)	ns	ns	ns	ns	ns	ns
	Psychosocial aetiology (ref: primarily psychosocial)	ns	ns	ns	ns	ns	ns
	Believes individuals with SZ are dangerous (ref: does not believe SZ to be dangerous)	ns	ns	ns	0.6 (0.4–0.9)	0.4 (0.2–0.8)	0.3 (0.1–0.7)
	Survey filled out online (ref: filled out in pen and paper)	ns	ns	ns	ns	ns	ns

		A conversation	Neighbour	Colleague	Friendship	Boss	Romantic relationship
		OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Autism	Gender (ref: male)	3.1 (1.8–5.6)	2.3 (1.5–3.7)	1.9 (1.2–2.9)	2.4 (1.5–3.7)	ns	ns
	Age	ns	ns	ns	ns	ns	ns
	Educational attainment (ref: no further education)	ns	ns	ns	ns	ns	ns
	Occupation (ref: being a student)	ns	ns	ns	ns	ns	ns
	Knows someone with ASD (ref: does not know any with ASD)	ns	ns	1.9 (1.2–3.0)	2.0 (1.2–3.0)	ns	ns
	Biological aetiology (ref: primarily biological)	ns	ns	ns	ns	ns	ns
	Psychosocial aetiology (ref: primarily psychosocial)	ns	ns	ns	ns	ns	ns
	Survey filled out online (ref: filled out in pen and paper)	ns	ns	ns	ns	ns	ns

OR: odds ratio; 95% CI: 95% confidence interval; ref: reference group; ns: non-significant at 0.05; SZ: schizophrenia; ASD: autism spectrum disorder.

Age and gender were generally not predictive of the respondents’ willingness to engage in social interactions with people with schizophrenia, and the sporadic significant associations are interpreted as chance findings. However, across all social interaction domains knowing a person with schizophrenia predicted more positive feelings. For the more intimate social interactions (friendship, boss and romantic relationship), believing a person with schizophrenia to be dangerous significantly decreased positive attitude towards interaction. Educational attainment, occupational status and beliefs about the aetiology of schizophrenia were non-significant (Table 3). As for the findings for autism, the mode of data collection had no statistically significant impact on the respondent’s tendency towards expressing willingness to socially interact with individuals with the diagnosis.

Stigma and self-stigma

Finally, the respondents were asked whether they would find it difficult or shameful if they or a family member received a diagnosis of autism or schizophrenia, and if they were to receive a diagnosis themselves for either of the conditions, whether or not they would disclose it to family members or colleagues (Figure 2). The results suggest a greater degree of potential self-stigma and unwillingness to elaborate on the disorder of schizophrenia to colleagues compared to autism (‘I would find it difficult’: x² = 80.1, p < 0.001; ‘I would feel ashamed’: x² = 89.6, p < 0.001; ‘I would feel ashamed family’: x² = 0.0, p = 1.00; ‘I would tell family members’ x² = 1.0, p = 0.345 and ‘I would tell colleagues’ x² = 333.0, p < 0.001).

Figure 2.

Stigma and self-stigma in respondents.

Discussion

This study is one of very few investigations what members of the general population know, believe and feel about autism. The literature on schizophrenia is more extensive, but this study is only the second directly comparing individual beliefs about autism and schizophrenia (Durand-Zaleski et al., 2012).

Beliefs and knowledge about persons with autism and schizophrenia

In this study, the respondents showed recognition of the symptoms/behaviours associated with autism and schizophrenia. As the comparisons identified, knowledge about behaviours that are characteristic of autism compared to schizophrenia according to ICD-10 criteria was greater. Our findings suggest that while respondents could identify behaviours from all three-core symptom domains of autism, for schizophrenia, respondents were more familiar with positive symptoms than negative symptoms. With respect to the causes of autism, most respondents supported a biological aetiology such as the disorder being caused by hereditary factors (38%) and brain damage (42.7%), and the majority were dismissive of parental factors as causes of autism. Compared to the findings for autism, more respondents believed psychosocial factors to be among the primary causes of schizophrenia. For example, 18%–21% believed unstable upbringings and stressful life events to be the primary cause of schizophrenia. However, 37% of the respondents believed heritability to be an essential factor. It is not clear from our analyses how beliefs about aetiology might potentially affect individuals with autism or schizophrenia. We did not replicate the negative association between desire for social distancing and a primarily biological understanding of the causes of schizophrenia, as was found in the meta-analysis by Kvaale et al. (2013). However, the finding by Kvaale et al. (2013) was not very robust (r = −0.07, 95% CI = 0.01–0.13), and they observed great heterogeneity (I² = 79.9). It is possible that the association is relevant in some countries or socio-demographic groups while not in others, or that variation in findings was due to methodological differences in the included studies. Although it cannot be empirically documented, we believe that at least for individuals with autism and their relatives, it is positive that the term ‘refrigerator parents’ coined from the interpretations of Leo Kanner’s work from 1943 (Kanner, 1968), at least in our sample seems to have been abandoned.

Furthermore, our results indicate that the respondents were moderately optimistic regarding the prognosis of both autism and schizophrenia (can live a normal life, can live alone and can have a regular job). A total of 74 and 77% believed that individuals with autism and schizophrenia are able to live normal lives. At the same time, 64 and 76% believed that individuals with autism/schizophrenia require lifelong treatment. How to interpret these findings is unclear, but it could be suggested that respondents acknowledge the potential lifelong needs of afflicted individuals but still believe that despite the disabilities, a normal to near normal life is possible for the majority. To some extent, the respondents’ perceptions seemed positively biased, since longitudinal studies of individuals with autism and schizophrenia document a high frequency of poor outcomes for both autism and schizophrenia populations during adulthood (Clemmensen et al., 2012; Magiati et al., 2014). As mentioned previously, parents and practitioners highlight the potential negative effects of excessively positive public perceptions of autism, as many individuals with autism as well as their families indeed deal with daily struggles and challenges, which requires public recognition and support (Pellicano et al., 2014).

Our study also identified some relatively common stereotypes about individuals with autism and schizophrenia. In our study, 71% of the respondents believed that people with schizophrenia experience a split personality. This finding is in line with the findings from previous studies (Economou et al., 2009; Gaebel et al., 2002; Leiderman et al., 2011). These perceptions might arise from portrayals of individuals with schizophrenia in films and books. Furthermore, 33% of the respondents in this study believed people with schizophrenia to be dangerous to themselves and others. These findings are in line with or lower than previous studies (Bag et al., 2006; Crisp et al., 2000; Durand-Zaleski et al., 2012; Gaebel et al., 2002; Grausgruber et al., 2007; Griffiths et al., 2006). This characteristic was only endorsed in 3.9% of respondents when they were asked about autism and are in line with the findings by Durand-Zaleski et al. (2012) who found that individuals with autism were significantly less likely to be dangerous than individuals with schizophrenia. How do these differences in perceptions come about? Although an infrequent event, the Danish news and media tend to highlight the diagnosis whenever violence or homicide is committed by an individual with schizophrenia, which can lead to distortion of the beliefs of people with no or limited personal experience with people with schizophrenia. Our analyses suggest that assumptions about dangerousness are harmful in the sense that it, not surprisingly, creates a desire for social distance to individuals with schizophrenia while this stigma seems less relevant for individuals with autism and their relatives. Media and mental health information campaigns should try to diminish this frequent belief about individuals with schizophrenia in our society. We found it puzzling that a higher percent recognition of the symptoms of schizophrenia was significantly related to beliefs about individuals with schizophrenia being dangerous to self and others. It is possible that awareness about primarily the positive symptoms of schizophrenia tend to manifest into a generalized stereotype of individuals with schizophrenia as being more unstable than might actually be the case. In addition, we found it noteworthy that the respondents very infrequently (1% and 7%) believed people with autism or schizophrenia to be caring individuals, despite their difficulties in the domains of social interaction.

Contrary to the findings for individuals with schizophrenia, some additional positive stereotypes for autism were identified by our respondents. A total of 27.3% believed people with autism to be creative, versus only 6.3% for schizophrenia and 40.1% believed them to be more intelligent than the average human being. Above average intelligence was only listed as characteristically for individuals with schizophrenia by 2.1% of respondents. No similar studies of autism exist, which makes it difficult to validate whether our results indicate a general population belief. However, parental testimonies suggest that these positive stereotypes regarding autism might be frequent within other populations (Pellicano et al., 2014). Again, exposure to films and media might explain the relatively high frequencies of the observed attributions. Depictions of real or fictional individuals with, for example, savant syndrome, who possess exceptional knowledge or abilities such as being able to draw with great detail and perfection such as Stephen Wiltshire, might have formed the basis of these beliefs. How these attributions can impact on individuals with autism directly is unknown. On one hand, these perceptions could influence laypersons to be positive and interested in people with autism. On the other hand, since high intelligence quotient (IQ) and exceptional creativity are not characteristic for the majority of people with autism, this bias in perception could cause laypersons to expect too much of individuals with autism. This could ultimately be harmful for the self-esteem of individuals with autism who already struggle with being accepted or already have feelings of low self-worth. Our findings could support the belief that relatives of afflicted individuals experience challenges associated with caring for some with autism, which are not sufficiently recognized (Pellicano et al., 2014).

Stigma of self and others

A large part of the respondents stated that they would feel ashamed and find it more difficult to be diagnosed with schizophrenia than to be diagnosed with autism. This could be a consequence of the negative stereotypes associated with schizophrenia. Additionally, more respondents would tell their colleagues if they were diagnosed with autism compared to schizophrenia, which could be an example of the more positive stereotype associated with autism. We do appreciate that these questions may lack face validity compared to the first questions interrogating the respondent’s knowledge about the concepts of autism and schizophrenia. It is of course difficult to imagine and respond validly to questions about how one would feel about getting a diagnosis of either autism or schizophrenia as it is highly hypothetical. As the majority of our respondents knew that autism is a disorder of childhood, the likelihood of receiving an autism diagnosis might not have seemed realistic to the respondents, and thus, they may have been less likely to express negative feelings about such a scenario. More appropriate ways of investigating this issue in future studies needs to be considered.

Social distancing and approach

Continuing the discussion above, the respondents in this study tended to feel more comfortable about interacting with people diagnosed with autism, compared to people diagnosed with schizophrenia. Our findings indicate that the strongest desire for distance was expressed when the respondents were asked how they would feel about being in a romantic relationship or having a boss diagnosed with either disorder. Other studies have also found similar tendencies in area of schizophrenia research (Economou et al., 2009; Gaebel et al., 2002; Grausgruber et al., 2007; Griffiths et al., 2006; Reavley and Jorm, 2012). Since no studies to our knowledge have investigated this topic for individuals with autism, the results need replication. However, the extent to which respondents felt insecure about socially interacting with individuals with both schizophrenia and autism requires attention due to the harmful social and psychological consequences of social rejection.

In this study, females showed a higher degree of willingness to interact with people diagnosed with autism. This finding differs from two American studies of college students that found that males tended to express more positive attitudes towards two vignettes about a case with autism (Matthews et al., 2015; Nevill and White, 2011). In the study by Nevill and White (2011), males more frequently reported having a relative with autism, and it might be that the gender differences were confounded by the effect of knowing someone with autism. In line with this hypothesis, our results suggest that the gender difference in our study could potentially be explained by the fact that more females knew somebody with autism, and thus, that gender is only an indirect factor. The tendency towards females knowing more people with autism than males could be the result of occupation and educational background, as more females than males work in social services and health care settings where the prevalence of autism is higher than within the general population.

An additional analysis concerned whether the mode of data collection had an impact on the respondents answers. This setting, compared to the online version of the questionnaire, could have impacted the respondent’s answers towards expressing more willingness to socially interact with individuals with autism or schizophrenia, in order to appear more socially and ethically correct. However, at least for these questions, no statistically significant differences emerged as a function of collection procedure (Table 3).

Limitations

Although this study possesses strengths with regard to the relatively large sample size, the contribution it makes to the limited literature on autism and the dependent design, limitations apply. First, while we were able to adjust the analyses on gender, due to a small sample size and a series of unknown population distribution parameters, weighting on these other factors was not possible, and thus, data could potentially be biased. However, results of the regression analyses deemed both age, occupational and educational status non-significant as predictors for the desire for social approach/distancing. Second, our list of potential characteristics or potential causes of the disorders that laypersons could associate with autism or schizophrenia was not exhaustive. However, we had to prioritize among items to ensure participant engagement. Third, the questionnaire used in the present survey has not been validated in terms of example test–retest reliability. Future studies should evaluate the questionnaire and develop the properties of both content and form. Fourth, one might argue that forcing individuals into answering a range of the questions as ‘yes/no’, reduces the ability for us to investigate complexity of the beliefs and attitudes towards individuals with autism/schizophrenia in our sample. However, we chose this option in the items related to knowledge (section ‘Materials and methods’) to force respondents into deciding among options and to make the questionnaire more engaging using a variation in response options. Fifth, it is a potential risk that the respondents in this study, to a great extent, self-selected so that, for example, relatives or professionals working with individuals with autism or schizophrenia found it more relevant and interesting to participate in the study than individuals without any personal attachment. However, the degree of bias is unknown since we were neither able to detect the rate of response versus non-response because of the modes of data collection nor identify what potential differences there exist between responders and non-responders. In retrospect, it would have been informative to be aware exactly how well the respondents were acquainted with individuals with the diagnoses. Future analyses should investigate this issue further. In addition, it would have been interesting to investigate whether certain occupational positions could elaborate on the relationship between the female gender and social distance as discussed previously. Finally, it should be mentioned that our results may not generalize to settings outside of Denmark, as our review detected great between-study variations from country to country.

Conclusion

Our findings indicate that the beliefs about people with schizophrenia and autism are well differentiated among members of the general population. Significantly, more symptoms of autism were recognized than for schizophrenia, more positive stereotypes were associated with autism and negative stereotypes were more frequently associated with schizophrenia than autism. Beliefs about the prognosis of autism and schizophrenia were generally optimistic but the exact interpretation is ambiguous. These stereotypes might have affected the stigma associated with the disorder among laypeople as our results suggest that respondents would feel less negative towards receiving a diagnosis of autism compared to schizophrenia. Generally, our respondents were more positive towards socially interacting with individuals with autism compared to schizophrenia. Our results suggest that negative stereotypes cause social distancing, while the meetings between individuals with autism or schizophrenia and the general public reduce the likelihood for individuals with the disorders to experience social isolation and rejection.

Footnotes

Declaration of conflicting interests

None of the authors have conflicts of interest to declare.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Atladottir

Gyllenberg

Langridge

. (2015) The increasing prevalence of reported diagnoses of childhood psychiatric disorders: a descriptive multinational comparison. European Child & Adolescent Psychiatry 24(2): 173–183.

Bag

Yilmaz

Kirpinar

(2006) Factors influencing social distance from people with schizophrenia. International Journal of Clinical Practice 60(3): 289–294.

Clemmensen

Vernal

Steinhausen

(2012) A systematic review of the long-term outcome of early onset schizophrenia. BMC Psychiatry 12: 150.

Crisp

Gelder

Rix

. (2000) Stigmatisation of people with mental illnesses. The British Journal of Psychiatry: The Journal of Mental Science 177: 4–7.

Durand-Zaleski

Scott

Rouillon

. (2012) A first national survey of knowledge, attitudes and behaviours towards schizophrenia, bipolar disorders and autism in France. BMC Psychiatry 12: 128.

Economou

Richardson

Gramandani

. (2009) Knowledge about schizophrenia and attitudes towards people with schizophrenia in Greece. The International Journal of Social Psychiatry 55(4): 361–371.

Elsabbagh

Divan

Koh

. (2012) Global prevalence of autism and other pervasive developmental disorders. Autism Research: Official Journal of the International Society for Autism Research 5(3): 160–179.

Furnham

Raja

Khan

(2008) A cross-cultural comparison of British and Pakistani medical students’ understanding of schizophrenia. Psychiatry Research 159(3): 308–319.

Gaebel

Baumann

Witte

. (2002) Public attitudes towards people with mental illness in six German cities: results of a public survey under special consideration of schizophrenia. European Archives of Psychiatry and Clinical Neuroscience 252(6): 278–287.

10.

Grausgruber

Meise

Katschnig

. (2007) Patterns of social distance towards people suffering from schizophrenia in Austria: a comparison between the general public, relatives and mental health staff. Acta Psychiatrica Scandinavica 115(4): 310–319.

11.

Griffiths

Nakane

Christensen

. (2006) Stigma in response to mental disorders: a comparison of Australia and Japan. BMC Psychiatry 6: 21.

12.

Holt

Christensen

(2013) Utahns’ understanding of autism spectrum disorder. Disability and Health Journal 6(1): 52–62.

13.

IBM Corp (2013) IBM SPSS Statistics for Windows, version 22.0. Armonk, NY: IBM Corp.

14.

Kanner

(1968) Autistic disturbances of affective contact. Acta Paedopsychiatrica 35(4): 100–136.

15.

Kvaale

Gottdiener

Haslam

(2013) Biogenetic explanations and stigma: a meta-analytic review of associations among laypeople. Social Science & Medicine 96: 95–103.

16.

Lebowitz

(2013) Stigmatization of ADHD: a developmental review. Journal of Attention Disorders. Epub ahead of print 13 February. DOI: 10.1177/1087054712475211.

17.

Leiderman

Vazquez

Berizzo

. (2011) Public knowledge, beliefs and attitudes towards patients with schizophrenia: Buenos Aires. Social Psychiatry and Psychiatric Epidemiology 46(4): 281–290.

18.

Link

Phelan

(2014) Stigma power. Social Science & Medicine 103: 24–32.

19.

McGrath

Saha

Chant

. (2008) Schizophrenia: a concise overview of incidence, prevalence, and mortality. Epidemiologic Reviews 30: 67–76.

20.

Magiati

Tay

Howlin

(2014) Cognitive, language, social and behavioural outcomes in adults with autism spectrum disorders: a systematic review of longitudinal follow-up studies in adulthood. Clinical Psychology Review 34(1): 73–86.

21.

Matthews

Goldberg

(2015) College students’ perceptions of peers with autism spectrum disorder. Journal of Autism and Developmental Disorders 45(1): 90–99.

22.

Nevill

White

(2011) College students’ openness toward autism spectrum disorders: improving peer acceptance. Journal of Autism and Developmental Disorders 41(12): 1619–1628.

23.

Norman

Windell

Manchanda

(2012) Examining differences in the stigma of depression and schizophrenia. The International Journal of Social Psychiatry 58(1): 69–78.

24.

Pellicano

Dinsmore

Charman

(2014) What should autism research focus upon? Community views and priorities from the United Kingdom. Autism: The International Journal of Research and Practice 18(7): 756–770.

25.

Reavley

Jorm

(2012) Stigmatising attitudes towards people with mental disorders: changes in Australia over 8 years. Psychiatry Research 197(3): 302–306.

26.

Reavley

Jorm

(2014) The Australian public’s beliefs about the causes of schizophrenia: associated factors and change over 16 years. Psychiatry Research 220(1–2): 609–614.

27.

StataCorp LP (2011) Stata Statistical Software: Release 12. College Station, TX: StataCorp LP.

28.

Sutterland

Dieleman

Storosum

. (2013) Annual incidence rate of schizophrenia and schizophrenia spectrum disorders in a longitudinal population-based cohort study. Social Psychiatry and Psychiatric Epidemiology 48(9): 1357–1365.