Knowledge,contact and stigma of mental illness: Comparing three stakeholder groups in Hong Kong

Abstract

Background:

It remains unknown what the impacts of multiple dimensions of contact (e.g. level, quantity and quality) are on the stigma of mental illness.

Aims:

To explore the relationship between the multiple dimensions of contact and stigma of mental illness among family members (FM), mental health workers (MHW) and community residents (CR) in Hong Kong.

Methods:

The stigma, contact and knowledge were measured in FM, MHW and CR in Hong Kong. Multiple regression analyses were used.

Results:

MHW (n = 141) had higher knowledge, more contact and lower stigma of mental illness than CR (n = 95) or FM (n = 62). Knowledge and contact quality were significantly associated with lower stigma of mental illness in the three groups. However, contact level and contact quantity were not significantly associated with most stigma components. The contact level was positively associated with stigma of mental illness among FM and CR.

Conclusions:

The results of this study highlight the differences in knowledge, contact and stigma of mental illness among different stakeholder groups. This study suggests that positive contact (e.g. equal, supportive, voluntary and pleasant contact) reduces stigma of mental illness, while negative contact (e.g. unfriendly, unsupportive, unpleasant contact) may increase stigma. The Enhancing Contact Model (ECM) should be tested in future anti-stigma interventions.

Keywords

Contact stigma family members mental health workers community residents

Introduction

Stigma of mental illness is a global public health issue (Henderson & Thornicroft, 2009; Ran et al., 2021; World Health Organization [WHO], 2013). According to social cognitive model of stigma (Corrigan et al., 2003, 2014), which built on attribution theory (Weiner, 1988) and theory of dangerousness (Angermeyer et al., 2004), stigma of mental illness is conceptualized in terms of stereotypes (negative beliefs about persons with mental illness (PMI), e.g. dangerousness, personal responsibility for mental illness), prejudice (negative emotional reactions resulting from endorsement of negative stereotypes, e.g. fear, anger, low pity) and discrimination (negative behaviour acting out prejudice, e.g. avoidance, withholding help, endorsing segregation and coercion) (Corrigan & Bink, 2005; Zhang et al., 2019).

Previous research showed high prevalence of public stigma of mental illness worldwide (Evans-Lacko et al., 2012; Lauber & Rössler, 2007), leading to a wide range of negative consequences for PMI (Clement et al., 2015; Oexle et al., 2017; Thornicroft et al., 2009). Evidence also showed stigmatizing attitudes to be shared by those closely associated with PMI, namely family members (FM) and mental health workers (MHW) (Corrigan & Nieweglowski, 2019; Thornicroft et al., 2016; Van Dorn et al., 2005). The stigma can manifest in forms of expressed emotions, paternalism or avoidance, contributing to poorer prognosis of PMI (Knaak et al., 2017; Moses, 2010; Ran et al., 2018; Zhang et al., 2019).

Existing studies have identified knowledge and contact as two major determinants of stigma (Corrigan et al., 2012). Knowledge is a critical component of mental health literacy that challenges false stigmatizing beliefs and aids management of mental illness (Reavley & Jorm, 2011; Thornicroft et al., 2007; Wei et al., 2015). Among the general public, research found knowledge to have a small-to-medium effect in reducing stigma (Morgan et al., 2018). Among FM and MHW, knowledge is especially crucial in facilitating better understanding and management of PMI’s conditions (Fang et al., 2020; Madianos et al., 2005; Shin, 2004).

Contact may promote personal exchange, enhance empathy and challenge negative stereotypes (Corrigan et al., 2012). Most public stigma research showed contact to be associated with less fear, perceived dangerousness, social distance and coercive attitude (Alexander & Link, 2003; Couture & Penn, 2003; Thornicroft et al., 2016). Meta-analyses found that contact intervention outperformed education and yielded medium-sized stigma reduction (Corrigan et al., 2012; Griffiths et al., 2014). However, a recent review cautioned against overgeneralizing inverse contact-stigma relationship to FM and MHW (Corrigan & Nieweglowski, 2019). For FM in many countries, they often live with PMI and shoulder primary caregiving duties. Their experiences with PMI may contain positive aspects that foster supportive attitudes and/or negative aspects that foster stigmatizing attitudes (Corrigan & Nieweglowski, 2019; Moses, 2010; Van der Sanden et al., 2016). Similarly, for MHW, their interaction with PMI can be positive (e.g. altruistic desire to help, better mental health knowledge, supportive attitudes) (Burks et al., 2012; Del Olmo-Romero et al., 2019; Yuan et al., 2017) and/or negative (e.g. burnout, associative stigma, therapeutic pessimism) (Knaak et al., 2017; Zuardi et al., 2011).

As negative interactions with PMI can potentially contribute to positive contact-stigma relationship, further studies should better explore the under-researched relationship between contact quality (e.g. positive and negative contact) and stigma (Corrigan & Nieweglowski, 2019; Couture & Penn, 2003; Li et al., 2020). Given possible differences in knowledge and contact experiences with PMI among FM, MHW and CR, stigma phenomenon of each group should also be separately examined (Corrigan & Nieweglowski, 2019; Van Dorn et al., 2005).

In Hong Kong, traditional Chinese perspective prevails, regarding mental illness as a threat to social order (Lam et al., 2010; Wong, 2000). Such moralizing views may contribute to more severe public stigma towards mental illness (Chien et al., 2014; Lee et al., 2005). Community residents (CR) have limited mental health knowledge, fear PMI’s propensity to violence and oppose psychiatric rehabilitation facilities near their residence (Chan et al., 2016; Fang et al., 2020; Siu et al., 2012). Moreover, stigmatizing views are commonly shared by FM and MHW in Hong Kong (Chien et al., 2014; Lee et al., 2006). FM are susceptible to Chinese notions of family shame attached to mental illness (Mak & Cheung, 2012), causing them to criticize or maintain secrecy of PMI in their family (Chien et al., 2014; Lee et al., 2005). Similarly, MHW in Hong Kong were found to hold negative attitude and inadequately involve PMI in decision-making (Chien et al., 2014; Lee et al., 2006). Local studies presented mixed findings on how knowledge and contact are linked to stigma in different groups (Chan et al., 2016; Siu et al., 2012; Wan & Wong, 2019).

Evidence shows that positive contact (defined as equal, cooperative, supportive, intimate, voluntary and friendly contact) are essential for stigma reduction (Allport, 1954; Pettigrew et al., 2011). We proposed a multidimensional conceptualization of contact including: (1) contact level (how intimate); (2) contact quantity (how frequent); and (3) contact quality (how positive). The enhancing contact model (ECM) was proposed firstly by M.S. Ran for addressing the role of contact, especially positive contact, on reducing the stigma of mental illness (Li et al., 2020; Fang et al., 2020). However, further studies should be conducted to test the ECM. This study aimed to compare the relationships between multiple dimensions of contact and stigma. We hypothesized that contact quality (e.g. positive or negative) was the most critical influencing factor on stigma level.

By now, no existing study has systematically compared the stigma of mental illness and its influencing factors among FM, MHW and CR in Hong Kong. This is the first study to compare among FM, MHW and CR in Hong Kong on: (1) level of cognitive, affective and behavioural components of stigma and (2) influencing factors of different components of stigma (i.e. knowledge, contact, socio-demographics).

Method

Participants

Our participants included FM, MHW and CR. Eligible FM were required to have at least one relative who met the criteria of fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) for diagnosis of any mental disorder(s). They should be immediate family members who supported PMI. Eligible MHW included social workers, nurses, psychologists, counsellors or other workers (e.g. peer support workers, programme assistants) who provided direct care to PMI in Integrated Community Centres for Mental Wellness (ICCMWs). CR refers to adult community members without diagnosis of mental disorder. FM and MHW were recruited from 16 ICCMWs and two Relative Resource Centres (RRCs) from July 2019 to January 2020 in three regions (Hong Kong island, Kowloon and the New Territories) of Hong Kong. Subjects of FM and MHW might be partly representatives of those in ICCMWs in Hong Kong. CR were recruited across the territory through snowball sampling with the help of ICCMWs.

Measures

Surveys for FM, MHW and CR included measures of stigma, knowledge and contact. Demographic information was also collected, including age, gender, education, income, years of work experience (MHW) and years of residence in the current community (CR). The questionnaires were written in Chinese.

Stigma of mental illness

The nine-item Attribution Questionnaire (AQ-9) (Corrigan et al., 2014) was used to assess respondents’ stigma toward PMI described in a vignette. The questionnaire was validated as a brief version of AQ-27 to measure nine stigma components, that is, cognition: (1) responsibility; (2) dangerousness; (3) anger; (4) pity; (5) fear; (6) help; (7) avoidance; (8) coercion; and (9) segregation. Items are rated on a nine-point Likert scale (1 = not at all; 9 = very much), with higher score indicating higher endorsement of the factor. The internal consistency is satisfactory (Cronbach’s alpha = .73) (Corrigan et al., 2014).

Knowledge

The 12-item Mental Health Knowledge Schedule (MAKS) (Evans-Lacko et al., 2010) was used to assess stigma-related mental health knowledge. Items 1 to 6 assessed knowledge in (1) help-seeking, (2) recognition, (3) support, (4) employment, (5) treatment and (6) recovery. Items 7 to 12 assessed knowledge concerning diagnosis of mental health conditions. Items are rated on five-point Likert Scale (1 = totally disagree; 5 = totally agree). All items are added to form a total score (ranging from 12 to 60), with higher score indicating more knowledge. The option of ‘Don’t know’ was coded as ‘3’. The internal consistency of items 1 to 6 of the scale was moderate (0.65) (Evans-Lacko et al., 2010). The use of MAKS among Chinese population has been validated (Li et al., 2014).

Contact level

The level-of-contact report (CLR) (Link et al., 1987) was used in this study to assess the intimacy level of contacts with PMI. The report lists 12 situations of varying levels of intimacy, ranging from lowest intimacy (‘I have never observed a person that I was aware had a serious mental illness’; score = 1) to highest intimacy (‘I have a mental illness’; score = 12). Participants were asked to check the situations that they have experienced. The index of familiarity was the score of the most intimate situation checked by the participant, with higher score indicating more intimate contact. The rank order has been validated with an inter-rater reliability of 0.83 (Corrigan et al., 2001).

Contact quantity

The five-item Contact Quantity Scale (CQTS) (Islam & Hewstone, 1993) was used in this study to measure the quantity of contact with PMI across different settings: (1) at workplace, (2) as neighbor, (3) as close friends, (4) informal talks and (5) home visits. Items are rated on a seven-point Likert scale (1 = not at all; 7 = very often). All items are summed to form a total score (ranging from 5 to 35), with higher score indicating more contact with PMI. The internal consistency was satisfactory in our sample (Cronbach’s alpha = .76).

Contact quality

The six-item Contact Quality Scale (CQLS) (Islam & Hewstone, 1993) was used in this study to measure the quality of contact with PMI. Participants were asked to rate their contact with PMI on six dimensions: (1) unequal-equal, (2) involuntary-voluntary, (3) superficial-intimate, (4) unpleasant-pleasant, (5) competitive-cooperative and (6) negative-positive. All items are rated on a seven-point Likert scale and summed to form a total score (ranging from 6 to 42), with higher score indicating higher-quality contact. The internal consistency was high in our sample (Cronbach’s alpha = .89).

Data collection

Ethical approval was obtained from the Human Research Ethics Committee of the University of Hong Kong (EA1811006). After full explanation of the study, participants signed consent forms to indicate their voluntary participation. FM and MHW were surveyed in ICCMWs whereas CR completed questionnaires at a quiet site of their own choosing. Questionnaire surveys were anonymous and self-administered. Trained research assistants were available to answer inquiries concerning questionnaire items.

Data analysis

All survey data were analyzed using the SPSS version 24.0. Descriptive statistics and Pearson’s correlation coefficients were used for the basic analysis of key variables. One-way ANOVA and Tukey’s HSD post-hoc tests were run to compare stigma, knowledge and contact among MHW, FM and CR. For cases of heteroscedasticity, Welch’s F and Games-Howell post-hoc test results were reported. Simultaneous multiple regression analyses were used to examine knowledge, contact and sociodemographic variables as influencing factors for AQ-9 components among FM, MHW and CR respectively. Missing values were handled by imputation of central tendency values of available cases. All analyses were two-tailed with significance level set at p < .05.

Results

Demographic characteristics

This study included a total of 366 participants (FM: n = 62; MHW: n = 141; CR: n = 163). MHW sample consisted of 54 males (38.3%) and 87 females (61.7%). Their age ranged from 22 to 65 years (M = 33.8; SD = 8.82) with an average working experience of 7.1 years (SD = 6.54). FM sample consisted of 17 males (27.4%) and 45 females (72.6%). Their age ranged from 24 to 73 years (M = 57.7, SD = 11.5). Majority of FM participants received either primary (24%) or secondary education (63%) and had monthly income below 10,000 HKD (75.8%). CR sample consisted of 19 males (11.7%) and 144 females (88.3%). Their age ranged from 18 to 86 years (M = 48.7, SD = 18.5). CR participants had lived in their current community for an average of 20.6 years (SD = 13.1). Sociodemographic characteristics of three groups are summarized in Table 1.

Table 1.

Sociodemographic characteristics of MHW, FM and CR (N = 366).

	MHW (n = 141)	FM (n = 62)	CR (n = 163)
	n (%)	n (%)	n (%)
Gender
Male	54 (38.3)	17 (27.4)	19 (11.7)
Female	87 (61.7)	45 (72.6)	144 (88.3)
Age (Years)
18–40	112 (79.4)	8 (12.9)	61 (37.4)
41–64	28 (19.9)	34 (54.8)	58 (35.6)
⩾65	1 (0.7)	20 (32.3)	44 (27.0)
Education
⩽Primary school	1 (0.7)	15 (24.2)	40 (24.5)
Middle school	1 (0.7)	20 (32.3)	30 (18.4)
High school	20 (14.2)	19 (30.6)	36 (22.1)
⩾College and above	119 (84.4)	8 (12.9)	57 (35.0)
Income (HKD)
$0–$9,999	5 (3.5)	47 (75.8)	122 (74.8)
$10,000–$19,999	38 (27.0)	8 (12.9)	20 (12.3)
$20,000–$29,999	41 (29.1)	3 (4.8)	13 (8.0)
⩾ $30,000	57 (40.4)	4 (6.5)	8 (4.9)
Work experience (Years)
⩽5	72 (51.1)
6–10	39 (27.7)
11–15	12 (8.5)
>15	18 (12.8)
Residence in current community (Years)
⩽10			46 (28.2)
11–20			30 (18.4)
21–30			50 (30.7)
>30			37 (22.7)

AQ-9 scores

Comparing against the midpoint score of 4.5, MHW scored high in help (M = 6.36, SD = 1.75), pity (M = 5.62, SD = 1.86) and fear (M = 4.96, SD = 2.46). FM scored high in pity (M = 6.53, SD = 2.47), coercion (M = 6.40, SD = 2.48), help (M = 5.50, SD = 2.47), dangerousness (M = 5.02, SD = 2.12), fear (M = 4.87, SD = 2.24) and segregation (M = 4.87, SD = 2.53). CR scored high in pity (M = 5.85, SD = 2.08), coercion (M = 5.61, SD = 2.38), help (M = 4.90, SD = 2.04), dangerousness (M = 4.73, SD = 1.79) and fear (M = 4.60, SD = 2.13). Both FM and CR shared similar patterns of AQ-9 component scores.

Between-group comparison in knowledge, contact and stigma

Table 2 shows the knowledge, contact and stigma among MHW, FM and CR.

Table 2.

Knowledge, contact and stigma among MHW, FM and CR.

Measure	Group	M (SD)	F	p	Post-hoc comparison^a	p
Knowledge	MHW	46.6 (5.12)	5.59	.004**	MHW > FM	.053
	FM	44.9 (4.97)			MHW > CR	.005**
	CR	44.8 (4.55)			FM > CR	.999
Contact level^b	MHW	8.70 (1.81)	132.7	<.001***	MHW < FM	<.001***
	FM	10.5 (1.57)			MHW > CR	<.001***
	CR	4.90 (3.60)			FM > CR	<.001***
Contact quantity	MHW	21.1 (4.89)	207.0	<.001***	MHW > FM	<.001***
	FM	13.5 (4.99)			MHW > CR	<.001***
	CR	8.96 (5.50)			FM > CR	<.001***
Contact quality^{b c}	MHW	31.1 (4.36)	9.88	<.001***	MHW > FM	.219
	FM	29.6 (6.51)			MHW > CR	<.001***
	CR	27.0 (8.51)			FM > CR	.082
Responsibility^b	MHW	1.97 (1.31)	19.4	<.001***	MHW < FM	.003**
	FM	3.02 (2.29)			MHW < CR	<.001***
	CR	3.10 (2.02)			FM < CR	.967
Dangerousness	MHW	3.66 (1.69)	18.0	<.001***	MHW < FM	<.001***
	FM	5.02 (2.12)			MHW < CR	<.001***
	CR	4.73 (1.79)			FM > CR	.540
Pity^b	MHW	5.62 (1.86)	3.35	.037*	MHW < FM	.030*
	FM	6.53 (2.47)			MHW < CR	.570
	CR	5.85 (2.08)			FM > CR	.138
Anger^b	MHW	2.23 (1.49)	6.89	.001**	MHW < FM	.008**
	FM	3.11 (2.06)			MHW < CR	.011*
	CR	2.79 (1.87)			FM > CR	.520
Fear^b	MHW	4.96 (2.46)	1.01	.367	MHW > FM	.962
	FM	4.87 (2.24)			MHW > CR	.361
	CR	4.60 (2.13)			FM > CR	.694
Segregation^b	MHW	2.96 (1.73)	24.6	<.001***	MHW < FM	<.001***
	FM	4.87 (2.53)			MHW < CR	<.001***
	CR	4.26 (2.20)			FM > CR	.216
Coercion^b	MHW	4.05 (1.99)	31.2	<.001***	MHW < FM	<.001***
	FM	6.40 (2.48)			MHW < CR	<.001***
	CR	5.61 (2.38)			FM > CR	.083
Avoidance^b	MHW	2.64 (1.56)	10.7	<.001***	MHW < FM	.107
	FM	3.19 (1.87)			MHW < CR	<.001***
	CR	3.53 (1.82)			FM < CR	.440
Help^b	MHW	6.36 (1.75)	22.7	<.001***	MHW > FM	.039*
	FM	5.50 (2.47)			MHW > CR	<.001***
	CR	4.90 (2.04)			FM > CR	.204

Note. One-way ANOVA comparison.

p < .05. **p < .01. ***p < .001.

N = 366; MHW (n = 141), FM (n = 62) and CR (n = 163).

Tukey HSD and Games-Howell post-hoc tests were used for cases of homogeneity or heterogeneity respectively.

Levene’s test of homogeneity of variances was significant (p < .05). Welch’s F and Games-Howell post-hoc tests were reported.

Participants who scored 5 on Contact Quantity (minimum score indicating no contact with PMI across contexts) were excluded from the measure of Contact Quality. The resultant sample sizes are as follows: MHW (n = 141), FM (n = 62) and CR (n = 95).

Knowledge

The between-group differences of knowledge were significant (F = 5.59, p = .004). MHW had significantly higher knowledge than CR (p = .005) and marginally significantly higher knowledge than FM (p = .053). Difference between FM and CR in knowledge was not significant.

Contact

The between-group differences were significant in contact level (F = 132.7, p < .001), contact quantity (F = 207.0, p < .001) and contact quality (F = 9.88, p < .001). FM had higher contact level than MHW and CR (p < .001) and MHW had higher contact level than CR (p < .001). For contact quantity, MHW had more contact with PMI in different contexts than FM and CR (p < .001) and FM had more contact with PMI than CR (p < .001). For contact quality, MHW had significantly better contact quality than CR (p < .001), but non-significantly better contact quality than FM (p = .219). Specifically, on the rating of positive-negative nature of contact (cutoff: >4), a higher percentage of MHW (83%) than FM (66.1%) or CR (56.8%) reported having positive contact with PMI.

Stigma

The between-group differences in all components of AQ-9 scores were significant (p < .05) except fear (p = .367). Post-hoc analyses revealed that MHW scored significantly lower than FM in a range of stigma factors, including perceived responsibility (p = .003), perceived dangerousness (p < .001), anger (p = .008), segregation (p < .001), coercion (p < .001) and pity (p = .030). MHW also scored significantly lower than CR in a range of stigma factors, namely perceived responsibility (p < .001), perceived dangerousness (p < .001), anger (p = .011), segregation (p < .001), coercion (p < .001) and avoidance (p < .001). MHW scored significantly higher in help than FM (p = .039) and CR (p < .001). Meanwhile, FM and CR did not significantly differ in all AQ-9 component scores.

Correlations among key variables

Bivariate analyses found knowledge and contact quality to have significant negative correlations to stigma components in all groups (p < .05). Contact level and contact quantity showed significant negative associations with stigma components among MHW and CR (p < .05), but not FM (p > .05). Low education, low income and short working experience were linked to higher stigma and less knowledge among MHW (p < .05). Older age was linked to less pity among FM (p < .05). Low education, low income, older age and longer residence in the current community were linked to higher stigma among CR (p < .05). Significant positive correlations between knowledge and contact quality were also found among FM and CR (p < .05).

Multivariate analyses of influencing factors for AQ-9 components

Multiple regression results of influencing factors are presented in Tables 3 and 4.

Table 3.

Influencing factors for AQ-9 components among MHW (n = 141).

	Responsibility		Danger		Pity		Anger		Fear		Segregation		Coercion		Avoidance		Help
	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p
1. Age	−.06	.645	.03	.818	−.03	.828	.11	.431	−.09	.526	.14	.318	−.002	.987	−.16	.255	.08	.546
2. Gendera	.09	.310	−.01	.873	.05	.597	.07	.404	.14	.086	.14	.104	.01	.866	.03	.687	−.16	.049*
3. Education	−.23	.055	−.13	.244	.13	.271	.05	.693	−.18	.140	−.04	.761	−.004	.976	−.21	.079	.10	.394
4. Income	−.09	.507	−.24	.057	−.30	.026*	−.17	.178	−.002	.991	.03	.847	.17	.221	−.09	.493	.16	.199
5. Work experience	.11	.435	.26	.065	.20	.183	−.02	.904	−.13	.381	−.06	.697	−.20	.188	.20	.168	−.12	.388
6. Knowledge	−.10	.272	−.13	.122	.14	.117	−.24	.008**	.03	.747	−.05	.594	−.13	.169	−.20	.024*	.18	.035*
7. Contact level	−.12	.165	.06	.443	.02	.820	−.20	.018*	−.04	.639	−.11	.223	−.14	.116	−.14	.105	.12	.158
8. Contact quantity	.12	.201	−.06	.469	.02	.804	.10	.272	.09	.332	.04	.652	−.03	.753	.06	.506	.12	.156
9. Contact quality	−.01	.908	−.18	.031*	−.08	.365	.004	.960	−.18	.035*	−.03	.708	.04	.656	−.09	.310	.04	.596
R²		.100		.177		.064		.159		.114		.058		.061		.136		.181
F for R² change	1.61		3.13**		.990		2.75**		1.87		.903		.949		2.29*		3.23**

Note. Multiple regression.

p < .05. **p < .01. ***p < .001.

0 = male, 1 = female.

Table 4.

Influencing factors for AQ-9 components among CR (n = 95).b

	Responsibility		Danger		Pity		Anger		Fear		Segregation		Coercion		Avoidance		Help
	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p
1. Age	.51	.003**	.27	.144	.14	.463	.47	.012*	.24	.211	.54	.001**	.44	.012*	.31	.085	.20	.291
2. Gendera	−.27	.006**	−.02	.834	.03	.794	−.11	.293	.07	.513	−.22	.017*	−.03	.733	−.18	.079	.08	.480
3. Education	.08	.621	.05	.762	.04	.821	.30	.078	.07	.706	.23	.118	.07	.683	.25	.137	.00	.997
4. Income	−.07	.503	−.12	.305	−.21	.083	−.05	.689	−.08	.503	−.23	.021*	−.11	.336	−.11	.366	.10	.418
5. Residence years	−.04	.749	.09	.520	−.10	.459	.03	.820	−.07	.601	.04	.733	.09	.454	.13	.330	−.12	.369
6. Knowledge	−.13	.190	−.11	.326	−.13	.261	−.20	.066	.01	.931	−.22	.022*	−.03	.784	−.08	.440	.01	.927
7. Contact level	.22	.050	.10	.426	.01	.927	.07	.584	−.04	.743	−.14	.158	−.05	.687	.12	.322	−.01	.946
8. Contact quantity	.002	.987	.03	.821	−.25	.054	.04	.729	.04	.740	.18	.089	.03	.824	−.05	.700	.14	.294
9. Contact quality	−.11	.321	−.09	.430	.14	.250	−.10	.371	−.12	.321	−.17	.085	−.13	.225	−.34	.004**	.22	.078
R²		.299		.147		.150		.162		.067		.411		.266		.203		.106
F for R² change	4.02***		1.63		1.67		1.82		.676		6.59***		3.42**		2.40*		1.12

Note. Multiple regression.

p < .05, **p < .01, ***p < .001.

0 = male, 1 = female.

Participants who scored 5 on Contact Quantity (minimum score indicating no contact with PMI across contexts) were excluded with listwise deletion as they provided no rating for Contact Quality.

MHW group

For MHW, multiple regression indicated that knowledge showed significantly negative links to anger (β = −0.24, p = .008), avoidance (β = −.20, p = .024) and positive link to help (β = .18, p = .035) (Table 3). For contact dimensions, contact level was negatively associated with anger (β = −.20, p = .018). Contact quantity was not a significant influencing factor of any of AQ-9 factors (p > .05). Contact quality was negatively associated with perceived dangerousness (β = −.18, p = .031) and fear (β = −.18, p = .035). Female gender was negatively related to help (β = −.16, p = .049). Income level was negatively related to pity (β = −.30, p = .026). The overall model was statistically significant for perceived dangerousness (R² = 0.177, F = 3.13, p = .002), anger (R² = 0.159, F = 2.75, p = .006), avoidance (R² = 0.136, F = 2.29, p = .020) and help (R² = 0.181, F = 3.23, p = .001).

FM group

For FM, multiple regression revealed that knowledge was negatively linked to anger (β = −.35, p = .023). For contact dimensions, contact level and contact quantity were not significantly related to any of AQ-9 stigma factors (p > .05), although contact level showed close-to-significant positive links with perceived dangerousness (β = .23, p = .098) and coercion (β = .24, p = .090). In contrast, contact quality was positively related to help (β = .41, p = .003). Contact quality also showed close-to-significant negative links to avoidance (β = −.27, p = .068) and segregation (β = −.27, p = .076). For sociodemographics, only education level showed significant positive link to help (β = .28, p = .039). The overall model was statistically significant for help (R² = 0.276, F = 2.53, p = .021).

CR group

For CR, multiple regression indicated that knowledge showed significantly negative link to segregation (β = −.22, p = .022) and close-to-significant negative link with anger (β = −.20, p = .066) (Table 4). For contact dimensions, contact level had a marginally significant positive relationship with perceived responsibility (β = .22, p = .050). Contact quantity was not significantly associated with any AQ-9 factor. Contact quality showed significant negative association with avoidance (β = −.34, p = .004). Its positive association with help (β = .22, p = .078) and negative association with segregation (β = −.17, p = .085) were also approaching significance. Age was positively related to a range of stigma factors, namely perceived responsibility (β = .51, p = .003), anger (β = .47, p = .012), segregation (β = .55, p = .001) and coercion (β = .44, p = .012). Female gender was negatively related to perceived responsibility (β = −.27, p = .006) and segregation (β = −.22, p = .017). Income level was negatively related to endorsement of segregation (β = −.23, p = .021). The overall model was statistically significant for perceived responsibility (R² = 0.299, F = 4.02, p < .001), segregation (R² = 0.411, F = 6.59, p < .001), coercion (R² = 0.203, F = 3.42, p = .001) and avoidance (R² = 0.203, F = 2.40, p = .018).

Discussion

To our knowledge, this is the first study to compare stigma and its influencing factors among three stakeholder groups (MHW, FM, and CR) in Hong Kong. This is also the first study to compare how contact level, contact quantity and contact quality are respectively associated with AQ-9 components. This study found notable between-group differences that distinguish the stigma phenomenon of each group. This study facilitates the understanding of the role of contact, especially positive contact, on reducing the stigma of mental illness from perspectives of three stakeholder groups.

Mostly consistent with previous Western studies using AQ to assess stigma in different groups (Corrigan et al., 2014; Del Olmo-Romero et al., 2019), our results suggested that MHW held more supportive attitudes characterized by help and pity whereas FM and CR held more mixed and stigmatizing attitudes characterized by pity, coercion, help, perceived dangerousness and fear. Our findings also showed that MHW had higher knowledge, contact quantity and contact quality than FM or CR. MHW also showed significantly higher help and lower stigma than FM and CR in terms of personal responsibility, perceived dangerousness, anger, segregation and coercion. Compared to other Hong Kong studies (Mak et al., 2015), our research studied community sample and found more positive attitudes towards mental illness among MHW. In community-based rehabilitation facilities (e.g. ICCMW), service users typically show less severe symptoms. More emphasis is placed on recovery-oriented community care, creating more opportunities of equal and cooperative interactions with PMI (Chui et al., 2012). MHW are well-trained with adequate mental health knowledge. These factors may combine to foster their relatively low stigma (Fang et al., 2020).

Although FM had a higher contact level with PMI than CR, they were not significantly different from CR in knowledge, contact quality and all AQ-9 component scores. While this fits the picture of pervasive intra-familial and public stigma in Hong Kong (Chan et al., 2016; Lee et al., 2005), our findings added more complexities beyond this picture. First, with more nuanced measures of stigma components with AQ-9, attitudes of FM and CR were found to be mixed rather than plainly negative (Fang et al., 2020). Previous studies found pity a paradoxical construct positively linked to both help and stigmatizing attitudes (Corrigan & Bink, 2005; Corrigan et al., 2003). Hence, attitudes of FM and CR were best characterized as a mixture of fear, sympathy for PMI’s misfortunes and support for mandatory treatment to help them. As these patronizing attitudes risk promoting paternalism that undercuts PMI’s self-determination (Corrigan et al., 2003), future anti-stigma interventions should develop strategies to target these stigmatizing beliefs. Second, FM showed comparable stigma to CR despite having more intimate contact with PMI. This may be accounted by their non-significant difference in knowledge and contact quality, which are crucial factors related to stigma level.

The results of this study revealed low education and low income to be major risk factors for stigma among MHW and CR in bivariate analyses (Fang et al., 2020). Old age represented another risk factor for stigma for CR. In multivariate models, old age and male gender of CR were found to be prominent risk factors of more than one stigma component.

For knowledge, our findings aligned with research that supported negative knowledge-stigma relationship (Fang et al., 2020; Griffiths et al., 2014; Morgan et al., 2018). Specifically, in multivariate analyses, knowledge was related to more help, less avoidance and less anger in MHW. It was also linked to less anger in FM and lower endorsement of segregation in CR. The results of this study contrasted with the positive knowledge-stigma relationship previously reported in Hong Kong (Chan et al., 2016), which may be accounted by our adoption of MAKS to measure stigma-related mental health knowledge (e.g. recovery, employment, support), as opposed to general knowledge of symptoms. Our bivariate analyses showed significant positive associations between knowledge and contact quality among FM and CR, which supports that knowledge deficit can be related to negative contacts with PMI (Fang et al., 2020; Wan & Wong, 2019).

Our findings showed the stronger link of contact quality than contact quantity or contact level to lower stigma (Allport, 1954; Islam & Hewstone, 1993). Contact quality was linked to lower perceived dangerousness and fear in MHW. It was also associated with higher help and lower avoidance in FM and CR. As such, personalized positive contacts featuring equality, support and friendliness help challenge stigma and promote supportive relationships (Islam & Hewstone, 1993; Pettigrew et al., 2011). Particularly, this study highlighted important associations of contact quality with behavioural tendencies of FM and CR, which bear especially significant impacts on real-life consequences of PMI. Conversely, contact level and contact quantity were less consistently associated with most AQ-9 components. Although contact level was significantly associated with less anger in MHW, it showed potential positive links to stigma components in FM and CR. Moreover, while contact quantity was insignificantly related to all AQ-9 components in all three groups, it showed potential links to lower pity and higher segregation score among CR. Contact quality is more centrally linked to lower stigma than contact level and contact quantity, supporting the multidimensional conceptualization of contact (Corrigan & Nieweglowski, 2019; Fang et al., 2020; Islam & Hewstone, 1993).

Enhancing contact model (ECM) for reducing stigma of mental illness

Combining our findings, higher contact level was linked to lower stigma for the group with more positive contact (e.g. MHW), whereas higher contact level or quantity was potentially linked to higher stigma for the group with less positive contact (e.g. FM, CR). Based on the results of this study, authors of this study support the ECM as the model (first proposed by M.S. Ran) to illustrate how multiple contact dimensions may interact to reduce the stigma of mental illness. As our empirical evidence pointed to the central role of contact quality in stigma reduction, the ECM hypothesizes that enhancing quantity and level of positive contacts is expected to reduce stigma whereas increasing quantity and level of negative contacts would be counterproductive (Fang et al., 2020; Li et al., 2020). With a carefully planned ECM intervention, stigma of mental illness in MHW, FM and CR may be reduced through enhancing the level and quantity of positive contact. Further studies need to be conducted to test the model.

Implications for policy and services

Our findings support that contact quality is a more crucial factor linked to lower stigma, as compared to contact level and contact quantity. The rationale of the proposed ECM thus emphasizes that contact quality should take priority over other contact dimensions (e.g. quantity, level) in designing anti-stigma interventions. Given the close relationship between the mental health knowledge and contact quality, the use of knowledge-contact interventions should also be further explored, such as what type of knowledge can best enhance contact quality and lower stigma (Chan et al., 2016; Fang et al., 2020; Mak & Cheung, 2012).

Health policies should be developed for addressing the contact, especially positive contact, among CR, MHW, FM and persons with mental illness. Cultural-specific anti-stigma interventions should also account for the unique characteristics of different stakeholder groups (Ran et al., 2021). This study indicated that knowledge was significantly associated with the behaviours of MHW (e.g. help, avoidance) and attitudes of FM and CR (e.g. anger, endorsing segregation) whereas contact quality was significantly linked to attitudes of MHW (e.g. fear, dangerousness) and behaviour of FM and CR (e.g. help, avoidance). The study has supported the ECM, especially the impact of positive contact on reducing the stigma of mental illness. Hence, different anti-stigma strategies should be deployed to challenge different stigmatizing aspects in each group (e.g. FM, MHW and CR).

Limitations of the study

First, the small sample size may have underpowered some statistical analyses and obscured potentially significant results. Future studies may increase sample size to ascertain positive links between contact level and stigma among groups with lesser contact quality. Second, findings from our cross-sectional study are correlational and direction of causality has to be determined in future research with longitudinal designs. Third, this study mainly relied on self-report measures. Although surveys were administered anonymously, stigma level may still be underestimated due to social desirability. Fourth, the representativeness of the CR may be limited due to the snowball sampling. Fifth, the behavioural measures may have assessed behavioural intentions instead of actual behaviour.

Footnotes

Acknowledgements

The authors are extremely grateful for all Integrated Community Centers of Mental Wellness (ICCMWs) and Relative Resource Centres (RRCs) in Hong Kong for their kind assistance offered to our study. Our gratitude also extends to each family member, mental health worker and community resident for their time and effort in joining our research and giving us valuable insights.

Author contributions

Ran designed and conducted the study. Yau conducted the data analysis. Ran and Yau wrote the first draft of the paper. Peng played an important role on revising the paper. Data collection: Ran, Li, Yau. Critical revision of the manuscript: all authors actively participated in revision.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research project is funded by the Public Policy Research Funding Scheme (PPR) from Policy Innovation and Co-ordination Office of the Hong Kong Special Administrative Region Government (Grant no.: 2019.A8.081.19A. PI: Ran MS).

ORCID iD

Mao-Sheng Ran

References

Alexander

Link

B.G.

(2003). The impact of contact on stigmatizing attitudes toward people with mental illness. Journal of Mental Health, 12(3), 271–289.

Allport

G. W.

(1954). The nature of prejudice. Perseus Books.

Angermeyer

M. C.

Matschinger

Corrigan

P. W.

(2004). Familiarity with mental illness and social distance from people with schizophrenia and major depression: Testing a model using data from a representative population survey. Schizophrenia Research, 69(2–3), 175–182.

Burks

D. J.

Youll

L. K.

Durtschi

J. P.

(2012). The empathy-altruism association and its relevance to health care professions. Social Behavior and Personality, 40(3), 395–400.

Chan

S. K. W.

Tam

W. W. Y.

Lee

K. W.

Hui

C. L. M.

Chang

W. C.

Lee

E. H. M.

Chen

E. Y. H.

(2016). A population study of public stigma about psychosis and its contributing factors among Chinese population in Hong Kong. International Journal of Social Psychiatry, 62(3), 205–213.

Chien

W. T.

Yeung

F. K.

Chan

A. H.

(2014). Perceived stigma of patients with severe mental illness in Hong Kong: Relationships with patients’ psychosocial conditions and attitudes of family caregivers and health professionals. Administration and Policy in Mental Health and Mental Health Service Research, 41(2), 237–251.

Chui

W. W.

Mui

J. H.

Cheng

K. M.

Cheung

E. F.

(2012). Community psychiatric service in Hong Kong: Moving towards recovery-oriented personalized care. Asia Pacific Psychiatry, 4(3), 155–159.

Clement

Schauman

Graham

Maggioni

Evans-Lacko

Bezborodovs

Morgan

Rüsch

Brown

J. S. L.

Thornicroft

, et al. (2015). What is the impact of mental health-related stigma on help-seeking? A systematic review of quantitative and qualitative studies. Psychological Medicine, 45(1), 11–27.

Corrigan

Bink

A. B.

(2005). On the stigma of mental illness (pp. 11–44). American Psychological Association.

10.

Corrigan

Markowitz

F. E.

Watson

Rowan

Kubiak

M. A.

(2003). An attribution model of public discrimination towards persons with mental illness Journal of Health and Social Behavior, 44(2), 162–179.

11.

Corrigan

P. W.

Green

Lundin

Kubiak

M. A.

Penn

D. L.

(2001). Familiarity with and social distance from people who have serious mental illness. Psychiatric Services, 52(7), 953–958.

12.

Corrigan

P. W.

Morris

S. B.

Michaels

P. J.

Rafacz

J. D.

Rüsch

(2012). Challenging the public stigma of mental illness: A meta-analysis of outcome studies. Psychiatric Services, 63(10), 963–973.

13.

Corrigan

P. W.

Nieweglowski

(2019). How does familiarity impact the stigma of mental illness? Clinical Psychology Review, 70, 40–50.

14.

Corrigan

P. W.

Powell

K. J.

Michaels

P. J.

(2014). Brief battery for measurement of stigmatizing versus affirming attitudes about mental illness. Psychiatry Research, 215(2), 466–470.

15.

Couture

Penn

(2003). Interpersonal contact and the stigma of mental illness: A review of the literature. Journal of Mental Health, 12(3), 291–305.

16.

Del Olmo-Romero

González-Blanco

Sarró

Grácio

Martín-Carrasco

Martinez-Cabezón

A. C

Perna

Pomarol-Clotet

Varandas

Ballesteros-Rodríguez

Rebolleda-Gil

Vanni

González-Fraile

, & INTER NOS group. (2019). Mental health professionals’ attitudes towards mental illness: Professional and cultural factors in the INTER NOS study. European Archives of Psychiatry and Clinical Neuroscience, 269(3), 325–339.

17.

Evans-Lacko

Brohan

Mojtabai

Thornicroft

(2012). Association between public views of mental illness and self-stigma among individuals with mental illness in 14 European countries. Psychological Medicine, 42(8), 1741–1752.

18.

Evans-Lacko

Little

Meltzer

Rose

Rhydderch

Henderson

Thornicroft

(2010). Development and psychometric properties of the mental health knowledge schedule. Canadian Journal of Psychiatry, 55(7), 440–448.

19.

Fang

Zhang

T. M.

Wong

Y. L. I.

Yau

Y. Y.

X.-H.

Chui

C. H. K.

Tse

Chan

C.L.-W.

Chen

E. Y. H

Ran

M.-S.

(2020). The mediating role of knowledge on the contact and stigma of mental illness in Hong Kong. International Journal of Social Psychiatry. Advance online publication. https://doi.org/10.1177/0020764020975792

20.

Griffiths

K. M.

Carron-Arthur

Parsons

Reid

(2014). Effectiveness of programs for reducing the stigma associated with mental disorders. A meta-analysis of randomized controlled trials. World Psychiatry, 13(2), 161–175.

21.

Henderson

Thornicroft

(2009). Stigma and discrimination in mental illness: Time to Change. Lancet, 373(9679), 1928–1930.

22.

Islam

M. R.

Hewstone

(1993). Dimensions of contact as predictors of intergroup anxiety, perceived out-group variability, and out-group attitude: An integrative model. Personality and Social Psychology Bulletin, 19(6), 700–710.

23.

Knaak

Mantler

Szeto

(2017). Mental illness-related stigma in healthcare: Barriers to access and care and evidence-based solutions. Healthcare Management Forum, 30(20), 111–116.

24.

Lam

C. S.

Tsang

H. W.

Corrigan

P. W.

Lee

(2010). Chinese lay theory and mental illness stigma: Implications for research and practices. Journal of Rehabilitation, 76(1), 35–40.

25.

Lauber

Rössler

(2007). Stigma towards people with mental illness in developing countries in Asia. International Review of Psychiatry, 19(2), 157–178.

26.

Lee

Chiu

M. Y.

Tsang

Chui

Kleinman

(2006). Stigmatizing experience and structural discrimination associated with the treatment of schizophrenia in Hong Kong. Social Science & Medicine, 62(7), 1685–1696.

27.

Lee

M. T.

Chiu

M. Y.

Kleinman

(2005). Experience of social stigma by people with schizophrenia in Hong Kong. British Journal of Psychiatry, 186(2), 153–157.

28.

Thornicroft

Huang

(2014). Levels of stigma among community mental health staff in Guangzhou, China. BMC Psychiatry, 14(1), 231. https://doi.org/10.1186/s12888-014-0231-x

29.

Link

B. G.

Cullen

F. T.

Frank

Wozniak

J. F.

(1987). The social rejection of former mental patients: Understanding why labels matter. American Journal of Sociology, 92(6), 1461–1500.

30.

X. H.

Zhang

T. M.

Yau

Y. Y.

Wang

Y.-Z

Wong

Y.-L., I.

Yang

Tian

X.-L.

Chan

C. L.-W.

Ran

M.-S.

(2020). Peer-to-peer contact, social support and self-stigma among people with severe mental illness in Hong Kong. International Journal of Social Psychiatry. Advance online publication. https://doi.org/10.1177/0020764020966009

31.

Madianos

M. G.

Priami

Alevisopoulos

Koukia

Rogakou

(2005). Nursing students’ attitude change towards mental illness and psychiatric case recognition after a clerkship in psychiatry. Issues in Mental Health Nursing, 26(2), 169–183.

32.

Mak

W. W.

Cheung

R. Y.

(2012). Psychological distress and subjective burden of caregivers of people with mental illness: The role of affiliate stigma and face concern. Community Mental Health Journal, 48(3), 270–274.

33.

Mak

W. W. S.

Cheung

F. M. C.

Wong

S. Y. S.

Tang

W. K.

Lau

J. T. F.

Woo

Lee

D. T. F.

(2015). Stigma towards people with psychiatric disorders. Hong Kong Medical Journal, 21(Suppl 2), 9–12.

34.

Morgan

A. J.

Ross

Reavley

N. J.

(2018). Systematic review and meta-analysis of Mental Health First Aid training: Effects on knowledge, stigma, and helping behaviour. PLoS One, 13(5), e0197102.

35.

Moses

(2010). Being treated differently: Stigma experiences with family, peers, and school staff among adolescents with mental health disorders. Social Science & Medicine, 70(7), 985–993.

36.

Oexle

Ajdacic-Gross

Kilian

Müller

Rodgers

Rössler

Rüsch

(2017). Mental illness stigma, secrecy and suicidal ideation. Epidemiology and Psychiatric Sciences, 26(1), 53–60.

37.

Pettigrew

T. F.

Tropp

L. R.

Wagner

Christ

(2011). Recent advances in intergroup contact theory. International Journal of Intercultural Relations, 35(3), 271–280.

38.

Ran

M. S.

Hall

B. J.

T. T.

Prawira

Breth-Petersen

X.-H.

Zhang

T.-M.

(2021). Stigma of mental illness and cultural factors in Pacific Rim region: A systematic review. BMC Psychiatry, 21, 8. https://doi.org/10.1186/s12888-020-02991-5

39.

Ran

M. S.

Zhang

T. M.

Wong

I. Y. L.

Yang

Liu

C. C.

Liu

Luo

Kuang

W. H.

Thornicroft

Chan

C. L.

, & CMHP Study Group. (2018). Internalized stigma in people with severe mental illness in rural China. International Journal of Social Psychiatry, 64(1), 9–16.

40.

Reavley

N. J.

Jorm

A. F.

(2011). Recognition of mental disorders and beliefs about treatment and outcome: Findings from an Australian national survey of mental health literacy and stigma. Australian & New Zealand Journal of Psychiatry, 45(11), 947–956.

41.

Shin

S. K.

(2004). Effects of culturally relevant psychoeducation for Korean American families of persons with chronic mental illness. Research on Social Work Practice, 14(4), 231–239.

42.

Siu

B. W. M.

Chow

K. K. W.

Lam

L. C. W.

Chan

W. C.

Tang

V. W. K

Chui

W. W. H.

(2012). A questionnaire survey on attitudes and understanding towards mental disorders. East Asian Archives of Psychiatry, 22(1), 18–24.

43.

Thornicroft

Brohan

Rose

Sartorius

Leese

, & INDIGO Study Group (2009). Global pattern of experienced and anticipated discrimination against people with schizophrenia: A cross-sectional survey. Lancet, 373(9661), 408–415.

44.

Thornicroft

Mehta

Clement

Evans-Lacko

Doherty

Rose

Koschorke

Shidhaye

O’Reilly

Henderson

(2016). Evidence for effective interventions to reduce mental-health-related stigma and discrimination. Lancet, 387(10023), 1123–1132.

45.

Thornicroft

Rose

Kassam

Sartorius

(2007). Stigma: Ignorance, prejudice or discrimination? British Journal of Psychiatry, 190(3), 192–193.

46.

Van der Sanden

R. L.

Pryor

J. B.

Stutterheim

S. E.

Kok

Bos

A. E

. (2016). Stigma by association and family burden among family members of people with mental illness: The mediating role of coping. Social Psychiatry and Psychiatric Epidemiology, 51(9), 1233–1245.

47.

Van Dorn

R. A.

Swanson

J. W.

Elbogen

E. B.

Swartz

M. S

. (2005). A comparison of stigmatizing attitudes toward persons with schizophrenia in four stakeholder groups: Perceived likelihood of violence and desire for social distance. Psychiatry Interpersonal & Biological Processes, 68(2), 152–163.

48.

Wan

K. F.

Wong

M. M.

(2019). Stress and burden faced by family caregivers of people with schizophrenia and early psychosis in Hong Kong. Internal Medicine Journal, 49, 9–15.

49.

Weiner

(1988). Attribution theory and attributional therapy: Some theoretical observations and suggestions. British Journal of Clinical Psychology, 27(1), 99–104.

50.

Wei

McGrath

P. J.

Hayden

Kutcher

(2015). Mental health literacy measures evaluating knowledge, attitudes and help-seeking: A scoping review. BMC Psychiatry, 15(1), 291. https://doi.org/10.1186/s12888-015-0681-9

51.

Wong

D. F.

(2000). Stress factors and mental health of carers with relatives suffering from schizophrenia in Hong Kong: Implications for culturally sensitive practices. British Journal of Social Work, 30(3), 365–382.

52.

World Health Organization. (2013). Stigma and discrimination. Author, Retrieved June 10, 2020, from http://www.euro.who.int/en/health-topics/noncommunicable-diseases/mental-health/priority-areas/stigma-and-discrimination

53.

Yuan

Picco

Chang

Abdin

Chua

B. Y.

Ong

Yow

K. L.

Chong

S. A

Subramaniam

(2017). Attitudes to mental illness among mental health professionals in Singapore and comparisons with the general population. PLoS One, 12(11), e0187593.

54.

Zhang

T. M.

Wong

Y. L. I.

Y. H.

S.-G.

X.-S.

Bacon-Shone

Gong

Huang

C.-H.

Tang

M.-M.

Cao

Chan

C. L.-W.

Ran

M.-S.

, & CMHP Study Group. (2019). An integrative model of internalized stigma and recovery-related outcomes among people diagnosed with schizophrenia in rural China. Social Psychiatry & Psychiatric Epidemiology, 54, 911–918.

55.

Zuardi

A. W.

Ishara

Bandeira

(2011). Burden and stress among psychiatry residents and psychiatric healthcare providers. Academic Psychiatry, 35(6), 404–406.