Development of the Family Stigma Stress Scale (FSSS) for Detecting Stigma Stress in Caregivers of People With Mental Illness

Abstract

People with mental illness and their family caregivers often perceive public stigma, which may lead to stigma-related stress (or stigma stress). However, no instruments have been developed to measure this stress for family caregivers of people with mental illness. We modified an instrument that measures the stigma stress of people with mental illness (i.e., the cognitive appraisal of stigma as a stressor) and examined the psychometric properties of the scores of the newly developed instrument: the Family Stigma Stress Scale (FSSS). Primary family caregivers of people with mental illness in Southern Taiwan (n = 300; mean age = 53.08 ± 13.80; 136 males) completed the FSSS. An exploratory factor analysis showed that the FSSS score had two factors; both factor scores had excellent internal consistency (α = .913 and .814) and adequate test–retest reliability (r = .627 and .533; n = 197). Significant correlations between FSSS factor scores and other instruments supported its concurrent validity and the ability of the FSSS to differentiate between clinical characteristics, for example, having been previously hospitalized or not. The FSSS is a brief and effective measure of the stigma stress of family caregivers of people with mental illness.

Keywords

caregiver Chinese culture family member severe mental illness stigma stress

People with mental illness often perceive public stigma (Parcesepe & Cabassa, 2013), and in many countries, public negative attitudes toward people with mental illness are common (Angermeyer & Dietrich, 2006; Angermeyer, van der Auwera, Carta, & Schomerus, 2017; Jacobsson, Ghanea, & Törnkvist, 2013). In addition to public stigma, people with mental illness may endorse negative stereotypes and turn them against themselves (Corrigan, Watson, & Barr, 2006; Mak & Cheung, 2010), which is called self-stigma. Self-stigma has profoundly negative effects on people with mental illness (Boyd, Emerald, Otilingma, & Peters, 2014; Corrigan, Larson, & Rüsch, 2009; Lin, Chang, Wu, & Wang, 2016).

People with mental illness may also need to cope with stigma-related stress (or stigma stress; Hinshaw, 2007; Link & Phelan, 2001; Thornicroft, 2006). Based on the works of Major and O’Brien (2005) and Lazarus and Folkman (1984), a stress-coping model of mental illness stigma was developed (Rüsch et al., 2009a, 2009b). In this model, stigma stress occurs if the cognitive appraisal of stigma as harmful exceeds the individual’s perceived resources to cope with stigma (Rüsch et al., 2009a). On the other hand, if persons with mental illness feel they have the resources to cope with stigma-related threats, their stigma stress will be low. In studies among people with severe mental illness or among individuals at risk of psychosis, increased stigma stress was associated with social anxiety, shame (Rüsch et al., 2009b), suicidality (Xu et al., 2016), transition to schizophrenia (Rüsch et al., 2015), and reduced well-being (Rüsch et al., 2014a). Therefore, stigma stress is important for mental health professionals to consider in their treatment plans for people with mental illness.

In addition to the stigma faced by people with mental illness, their family caregivers may also be stigmatized. This has been called “courtesy stigma” (being stigmatized because of one’s relationship to a person with a stigmatizing mark; Chang et al., 2015; Chang, Wu, Chen, & Lin, 2016; Goffman, 1963) and “affiliate stigma” (internalizing the stigma because of the relationship; Chang, Su, & Lin, 2016; Chang et al., 2015; Mak & Cheung, 2008). The stigma issue for family caregivers should be addressed because such stigma may induce heavy stress, especially in the context of Chinese culture (Chang, Yen, Jang, Su, & Lin, 2017). Family caregivers in this culture are likely to be distressed because according to Chinese culture, family members are often the cause of their relatives’ mental illness and responsible for the illness (Kleinman & Lin, 1981; L. H. Yang, 2007). In general, responsibilities for an individual’s well-being are shared by the family members in collectivistic cultures. Thus, family caregivers of people with mental illness are at a high risk of stigma stress, which may subsequently jeopardize their health (Rüsch et al., 2009b). If family caregivers cannot cope with their stigma stress, they too might develop psychological problems (Bevans & Sternberg, 2012; Schulz & Sherwood, 2008).

Moreover, family caregivers often collaborate with mental health professionals to take care of their relatives with mental illness; however, caregivers may be less able to focus their efforts on caregiving when they encounter stress (Bevans & Sternberg, 2012; Schulz & Sherwood, 2008). Hence, it is important for mental health professionals to monitor and detect the stigma stress of family caregivers to help optimize the effectiveness of treatment.

While several self-stigma instruments have been developed for people with mental illness (Boyd et al., 2014; Corrigan et al., 2012; Mak & Cheung, 2010) and for their caregivers (Chang et al., 2015, Chang, Su & Lin, 2016; Mak & Cheung, 2008), there is only one instrument of stigma stress in people with mental illness (Rüsch et al., 2009a, 2009b) but none for stigma stress among caregivers. We therefore aimed to develop an instrument for measuring stigma stress in family caregivers of people with mental illness. We also evaluated the psychometrics of the Family Stigma Stress Scale (FSSS) to determine whether it is suitable for clinical use.

Method

The study was approved by the institutional review boards (IRBs) of the following hospitals: Chi Mei Medical Center (IRB No. 10402-002); Jianan Psychiatric Center, Taiwan Ministry of Health and Welfare (IRB No. 15-016); and Chang Gung Memorial Hospital (IRB No. 104-1716B).

Patients and Procedures

Eligible patients—those diagnosed with schizophrenia, bipolar disorder, or major depressive disorder based on Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) criteria—were first identified by several psychiatrists who ensured the capacity of the patients with severe mental illness to understand the research purposes and give consent. All patients were recruited from psychiatric settings: outpatients (n = 206, 68.7%), inpatients (n = 53, 17.6%), day-care patients (n = 26, 8.7%), and home-care patients (n = 15, 5.0%) of three general hospitals and a psychiatric center in Southern Taiwan. Afterward, several experienced research assistants approached the patients and explained the objectives and nature of the study in detail. They also assessed the cognitive capacity of the patients using the following methods: Capacity to give informed consent was assessed through the patients’ verbal responses when the research assistants spoke to them about participation, and the issues of incapacity reflected in the patients’ verbal responses (e.g., attention problems) and behaviors (e.g., continuously asking similar questions about the same instructions) were also assessed. However, no standardized screening tools were used to test their eligibility. Finally, the research assistants did not exclude any patients based on this cognitive assessment.

After informed consent was obtained from the patient, the same research assistants contacted and provided a detailed introduction of the study to the primary family caregiver (n = 300), a family member who, according to the patient, devoted the most time and effort on their care. The patients provided written consent for the researchers to contact their primary family caregivers and collect their data. We let each patient define their primary family caregiver or identify the family caregiver who spent the most time on caregiving if they had a problem identifying their primary family caregiver. If the nominated family caregiver was not available or declined to participate, we did not invite additional family caregivers. The collected patient data included demographics from their medical records and three clinical characteristics (hospitalization, suicidal behaviors, and compulsory admission) from their caregivers. We collected no other patient data.

To protect patients’ rights to unbiased health care regardless of whether they agreed or declined to participate in the study, we did the following. First, all the psychiatrics who assisted in transferring patients to the study were informed and understood that there would be no effect on the patients’ health-care services whether or not they were participating in the study. Second, the psychiatrists did not know after the transfer whether the transferred patients had agreed to participate in the study. Third, the informed consents for the patients and for the family caregivers clearly stated that declining and withdrawing from the study would not influence their health-care services. Fourth, patient confidentiality was protected by delinking the patients’ and family caregivers’ private information and collected information. Hence, we did not know which set of data belonged to whom.

The research assistants screened the eligibility of family caregivers using the inclusion criteria: (a) ≥20 years old and (b) understood Taiwanese Mandarin Chinese (similar to Beijing dialect) or Taiwanese (similar to Hoklo or Southern Min). The exclusion criterion was having an obvious cognitive impairment, for example, a clinically observed difficulty maintaining attention, memory problems, or repetitious verbal expressions. Cognitive impairments were determined by well-trained research assistants. The first author, who is a psychiatrist with extensive experience in stigma research, trained the research assistants to ensure the quality of the standardized recruiting procedure. Two of the research assistants obtained a master degree in science, and the other research assistant had a bachelor degree in science. Except for the family caregivers of home-care patients (research assistants visited their homes to collect data), all other participants completed the interview in a comfortable and private room in hospitals without disturbance.

Structured questionnaires (which asked for sociodemographics and clinical characteristics and the instruments described in the Study Instruments section) were administered by well-trained and experienced research assistants. For those participants who had low levels of literacy, the research assistants read out the questions to maximize comprehension and reliability. The questionnaire required about 30 min to complete. Between 2 and 12 weeks later, 197 of the 300 patients completed the FSSS again. The 2- to 12-week intervals were not planned. Participants completed the second FSSS when they visited the outpatient again after an interval of 2–12 weeks.

Study Instruments

FSSS

The FSSS (Online Supplementary Table 1) was developed based on an instrument that measures cognitive appraisal of stigma stress for people with mental illness; it was designed by Rüsch et al. (2009a) and adapted from Kaiser, Major, and McCoy (2004). The instrument consists of two domains (the primary appraisal of perceived stigma-related harm and the secondary appraisal of perceived resources to cope with stigma); each domain has 4 items. We revised the instrument to use with family caregivers of people with mental illness, and the procedure of the development is detailed below.

We first developed an English version of the FSSS and replaced the original term “prejudice against people with mental illness” with “prejudice against my relative with mental illness.” No other modifications were made at this stage. Afterward, two Taiwanese psychiatrists fluent in English independently translated the English version of the FSSS into Mandarin for the Taiwan population. The two independently translated Taiwan versions were then integrated into a single forward-translated version with consensus between the two psychiatrists. A native English speaker back-translated this version and a group of experts (the two psychiatrists who did the forward translation and one psychologist) then compared the two English FSSS versions to ensure that the item meanings were consistent. Next, we held an eight-person focus group of caregivers (three men and five women; age range: 27–63 years old) for people with mental illness to test the readability of the Taiwan Chinese version. We made minor revisions of the wordings. The linguistic validity of the FSSS was confirmed, and we ensured that the patients would have no difficulties completing the FSSS.

The FSSS contains 8 items rated on a 4-point Likert-type scale (1 = strongly disagree, 2 = agree, 3 = agree, and 4 = strongly agree), and we hypothesized that the FSSS had two underlying factors as suggested by Rüsch et al. (2009a). The original instrument developed by Rüsch et al. (2009a) uses a 7-point Likert-type scale (from 1 = strongly disagree to 7 = strongly agree), and we decreased the scale to four to ensure that all the participants could provide valid answers. Clark and Watson (1995) suggested that more rather than fewer response alternatives might jeopardize the validity of the instrument, when the respondents do not have sufficient ability to accurately discriminate between many choices. Because 15 family caregivers were illiterate, we believe that using a 4-point scale was justified.

Higher scores on Items F1–F4 represent higher levels of stigma-related harm, and Items F5–F8 were reverse coded such that higher scores represent lower perceived coping resources. Unlike the difference score used by Rüsch et al. (2009a, 2009b), we used a total sum score for the FSSS: The total sum score for the FSSS ranged between 8 and 32; the score for each factor ranged between 4 and 16. Because stigma stress occurs when stigma-related harm exceeds perceived coping resources, we combined the harm and the coping scores into one FSSS total sum score with higher scores representing more stigma stress.

Affiliate Stigma Scale (ASS)

The ASS, a 22-item instrument, measures the self-stigma among caregivers of family members with mental illness (Mak & Cheung, 2008). All items are rated on a 4-point Likert-type scale; higher scores indicate higher levels of self-stigma. In addition, the ASS consists of three domains: affect and cognitive (7 items each) and behavior (8 items). The psychometric properties of the ASS score are satisfactory in the original Hong Kong version (α = .94; Mak & Cheung, 2008) and the Taiwan version (α = .82 to .93; Chang et al., 2015, Chang, Su & Lin, 2016).

Multidimensional Scale of Perceived Social Support (MSPSS)

The MSPSS, a 12-item instrument, measures subjectively perceived social support from three sources (4 items each): family, friends, and significant others (Zimet, Dahlem, Zimet, & Farley, 1988; Zimet, Powell, Farley, Werkman, & Berkoff, 1990). All items are rated on a 7-point Likert-type scale, and higher scores indicate greater perceived support. The three-structure framework of the MSPSS score has been affirmed using confirmatory factor analysis (Zimet et al., 1988, 1990), and the MSPSS score showed high internal consistency (α = .85 to .95) across different ethnic samples (Canty-Mitchell & Zimet, 2000). Moreover, the internal consistency was high (α = .80 to .89) in Chinese samples (Chou, 2000; Zhang & Norvilitis, 2002).

Center for Epidemiologic Studies Depression Scale (CES-D)

The CES-D, a 20-item instrument, was developed to clinically screen depression (Radloff, 1977). All the items are rated as 0 if rarely or none of the time (<1 day), 1 if some or a little of the time (1–2 days), 2 if occasionally or a moderate amount of the time (3–4 days), and 3 if most or all of the time (5–7 days) during seven days. There are four domains in the CES-D: depressed affect (7 items), somatic symptoms (7 items), interpersonal problems (2 items), and positive affect (4 items). The psychometric properties of the CES-D Taiwan version score accurately screened people with depression (area under receiver operating characteristic curve = .88 to .90; H. J. Yang, Soong, Kuo, Chang, & Chen, 2004) and its theoretical structure (Cheng, Yen, Ko, & Yen, 2012). A higher CES-D sum score indicates higher level of depression.

Rosenberg Self-Esteem Scale (RSES)

The RSES (Rosenberg, 1965), a 10-item instrument, measures the self-esteem with all items rated on a 4-point Likert-type scale with mean scores between 1 and 4. All 10 items are embedded in the same concept, and the unidimensionality has been supported by Rasch analysis in the Taiwan version score (Wu, Lai, & Shih, 2013). A higher RSES mean score indicates lower self-esteem.

Data Analysis

All the data were analyzed using SPSS 23.0 (IBM Corp., Armonk, NY). The demographics of family caregivers and of patients with severe mental illness, and the clinical characteristics of the ill relatives were analyzed using descriptive statistics: mean and SD for continuous variables; frequency and percentage for categorical variables. Exploratory factor analysis (EFA) was used to examine whether the FSSS score had a two-factor structure as proposed by Rüsch et al. (2009a). We used principal axis functioning with the promax oblique rotation method (κ set at 4) to extract the possible factors in the FSSS score. We determined whether a factor existed based on an eigenvalue > 1. Factor loadings > .4 were used to support an item embedded in the factor. After the factors were determined, Cronbach’s α and corrected item–total correlation were used to verify the internal consistency of each factor score: the cutoffs were >.7 for Cronbach’s α and >.4 for corrected item–total correlation. Test–retest reliability using Pearson correlation was examined for each item score, each factor score, and the overall FSSS score. Moreover, Cronbach’s α was used for the overall FSSS score.

Several multiple regression models were constructed to examine the concurrent validity of the FSSS score (including the factor score and the total score). Each regression model included one of the following independent variables: ASS total score, MSPSS total score, CES-D score, and RSES score. In addition, all models controlled for the demographics of the family caregivers: age, gender, marital status (currently married vs. currently unmarried), years of education, and living status (living together vs. not living together). The dependent variables were the factor scores of the FSSS identified using our EFA results. We used a standardized coefficient and an increased R ² to detect the associations between the FSSS score and the scores of other criteria.

Known-group validity, an ability to discriminate between two groups known to differ on the variable of interest, was examined using three clinical characteristics (had been hospitalized, had suicidal behaviors, and had been compulsorily admitted) for each FSSS factor score identified using our EFA results and for the total FSSS score. Using the three clinical characteristics provided by our family caregivers is appropriate because studies have found that people with mental illness who had been hospitalized, had presented with suicidal behaviors, and had been compulsorily admitted had higher levels of self-stigma than did their counterparts (Chang, Wu, Chen & Lin, 2016; Rüsch et al., 2014b). Therefore, similar conditions were generalized to the primary family caregivers of family members with mental illness. Specifically, to investigate known-group validity, we used several independent t tests to compare the differences in FSSS scores (including factor and total scores) between family caregivers whose relatives with mental illness had the clinical characteristics and those whose family members with mental illness did not have the clinical characteristics. We hypothesized that the family caregivers who cared for a relative with these clinical characteristics would have poorer FSSS scores than those who cared for a family member without these characteristics. Although each clinical characteristic stands alone, the factor and total scores of the FSSS were dependent. Therefore, we had to adjust for a type I error based on the number of factors in the FSSS. For example, if we find two factors in the FSSS, we must divide the type I error by three (two factor scores plus one total score). Hence, the α level was set at <.016 (.05 divided by 3) to indicate significance.

Results

Less than half (n = 136; 45.3%) of the family caregivers were men (mean age = 53.08 ± 13.80 years), and almost the same percentage of patients with mental illness (n = 140; 46.7%) were men (mean age = 46.70 ± 13.19 years; Table 1). All family caregivers were the primary family caregiver, and nearly 60% (n = 179) shared caregiving duties with other family members. Nearly 90% (n = 262) lived with their relatives with mental illness, half of whom had schizophrenia, one fourth of whom had bipolar affective disorder, and one fourth of whom had major depression. Moreover, most ill relatives (n = 195, 65.0%) had been hospitalized, some (n = 112, 37.3%) had suicidal behaviors, and some (n = 103, 34.3%) had been compulsorily admitted.

Table 1.

Characteristics of Caregivers and Patients with Mental Illness.

Variable	n (%)	Mean ± SD
Caregiver
Age in years		53.08 ± 13.80
Years of education		10.38 ± 4.33
Years of caregiving		11.33 ± 9.11
Gender (male)	136 (45.3)
Marital status (currently married)	197 (65.7)
Employment status
Full-time employment	141 (47.0)
Part-time employment	28 (9.3)
Unemployed or retired	131 (43.7)
Religious belief (yes)	249 (83.0)
Relationship with patient
Parents	119 (39.7)
Spouse	90 (30.0)
Siblings	43 (14.3)
Children	39 (13.0)
Others	9 (3.0)
Living with the ill relative (yes)	262 (87.3)
Affiliate Stigma Scale		2.23 ± 0.48
MSPSS		4.65 ± 0.90
CES-D		15.39 ± 11.31
RSES		2.15 ± 0.45
Patients with mental illness
Age (years)		46.70 ± 13.19
Education (years)		10.27 ± 4.03
Age at onset (years)		34.87 ± 14.17
Gender (male)	140 (46.7)
Marital status (currently married)	120 (40.0)
Employment status
Full-time or part-time employment	56 (18.7)
Supported or sheltered employment	4 (1.3)
Unemployed or retired	240 (80.0)
Diagnosis
Schizophrenia	150 (50.0)
Bipolar disorder	75 (25.0)
Major depressive disorder	75 (25.0)
Hospitalized (yes)	195 (65.0)
Suicidal behaviors (yes)	112 (37.3)
Compulsory admission (yes)	103 (34.3)

Note. MSPSS = Multidimensional Scale of Perceived Social Support; CES-D = Center for Epidemiologic Studies Depression Scale; RSES = Rosenberg Self-Esteem Scale.

EFA showed that two factors—Items F1–F4 clustered together to represent perceived harm (eigenvalue = 3.611 with 45.14% of explained variance), and Items F5–F8 clustered together to represent perceived coping resources (eigenvalue = 1.447 with 18.09% of explained variance)—were extracted from the FSSS Taiwan version score and that the two factors identically corresponded to the original FSSS. In addition, factor loadings and corrected item–total correlations were satisfactory both for perceived harm (factor loadings = .815 to .885; corrected item–total correlations = .783 to .829) and for perceived coping resources (factor loadings = .588 to .826; corrected item–total correlations = .504 to .703). Furthermore, there were no cross-loadings in the 8 items: The 4 items of perceived harm had low loadings on perceived coping resources (.311 to .362); the 4 items of perceived coping resources also had low loadings on perceived harm (.145 to .366). The two factors were moderately correlated (r = .41). The mean score for each factor was 8.72 (perceived harm) and 8.41 (perceived coping resources), which yielded a mean score of 17.13 for the entire FSSS. Internal consistency and test–retest were adequate for the overall FSSS score (α = .850; test–retest Pearson’s r = .664) and each factor score (α = .913 and r = .627 for perceived harm; α = .814 and r = .533 for perceived coping resources). Moreover, the test–retest reliability for each item score was fair to good (Table 2). However, the wide time frame seemed to affect test–retest reliability (Online Supplementary Table 2). The test–retest reliability shown by the time frame of 4–8 weeks (n = 47) was better than the time frames of 2–4 weeks (n = 112) and 8–12 weeks (n = 37).

Table 2.

Item Properties, Exploratory Factor Analysis, and Reliability of the Family Stigma Stress Scale.

Factor^a Item#	Item Description	Mean (SD)	Eigenvalue	Cronbach’s α	Test–Retest Reliability^d
Factor^a Item#	Item Description	Mean (SD)	Factor Loading^b	Item–Total Correlation^c	Test–Retest Reliability^d
Perceived harm	Prejudice against my relative with mental illness (MI)…	8.72 (2.41)	3.611	.913	.627
F1	Has a negative impact on my future	2.16 (0.70)	.815/.358	.783	.492
F2	Has harmful or bad consequences for me	2.13 (0.63)	.867/.311	.803	.479
F3	Affects many areas of my life	2.27 (0.68)	.885/.362	.829	.530
F4	Has a severe impact on my life	2.16 (0.68)	.851/.332	.797	.504
Perceived coping resources	I am/I have /I will/I am able to…	8.41 (1.94)	1.447	.814	.533
F5^e	Prepared to deal with prejudice against my relative with MI	2.14 (0.61)	.781/.345	.688	.414
F6^e	The resources to handle prejudice problems against my relative with MI	2.23 (0.67)	.826/.315	.703	.439
F7^e	Do the best to deal with prejudice against my relative with MI	1.94 (0.51)	.588/.145	.504	.312
F8^e	Rise up and meet prejudice demands against my relative with MI	2.09 (0.63)	.697/.366	.651	.453

^aThe two factors were moderately correlated (r = .41). ^bThe former loadings indicate the loadings on their embedded factor (e.g., F1 on perceived harm); the later loadings indicate those on the other factor (e.g., F1 on perceived coping resources); The bold values in the column are eigenvalue; the non-bold values are factor loadings. ^cItem–total correlation for two factors (perceived harm and perceived coping resources) separately; The bold values in the column are Cronbach’s alph; the non-bold values are item-total correlation. ^dTest–retest reliability examined using Pearson correlations (n = 197), and all p values < .01. ^eReversely coded items.

The multiple regression models demonstrated the concurrent validity for the two-factor scores of the FSSS and the FSSS total score. In the separate regression models, the ASS total score, the MSPSS total score, the CES-D score, and the RSES score significantly correlated with the FSSS factor scores and total score when demographics (age, gender, marital status, educational status, and living status) of the family caregivers were controlled. In addition, strong associations were found between the ASS total score and perceived harm (ΔR ² = .46), between the CES-D and perceived harm (ΔR ² = .26), and between the RSES and perceived harm (ΔR ² = .27). The associations between all the criteria scores and the perceived coping resources score were weaker than those between all the criteria score and the perceived harm score (Table 3).

Table 3.

Concurrent Validity of the Family Stigma Stress Scale (FSSS).

Independent Variable^a	Dependent Variable
	Perceived Harm^b			Perceived Coping Resources^c			Total Stigma Stress^c
	β	p Value	ΔR ²	β	p Value	ΔR ²	β	p Value	ΔR ²
Affiliate Stigma Scale	.715	<.001	.46	.256	<.001	.06	.618	<.001	.35
MSPSS	−.277	<.001	.07	−.261	<.001	.06	−.327	<.001	.10
CES-D	.524	<.001	.26	.364	<.001	.13	.548	<.001	.29
RSES	.543	<.001	.27	.415	<.001	.16	.589	<.001	.32

Note: Higher scores in FSSS indicate more stigma stress, in Affiliate Stigma Scale indicate more affiliate stigma, in MSPSS indicate more perceived social support, in CES-D indicate higher level of depression, and in RSES indicate lower self-esteem. MSPSS = Multidimensional Scale of Perceived Social Support; CES-D = Center for Epidemiologic Studies Depression Scale; RSES = Rosenberg Self-Esteem Scale. ^aThe four scales were separately constructed in different regression models. ^bAge, gender, marital status, educational status, and living status were controlled in all regression models and together had an R ² of .04. ^cAge, gender, marital status, educational status, and living status were controlled in all regression models and together had an R ² of .07.

Known-group validity was tested using three clinical characteristics (had been hospitalized, had suicidal behaviors, and had been compulsorily admitted) of the patients with mental illness. There were significant differences in perceived harm scores, perceived coping resources scores, and FSSS total scores between patients had been hospitalized (mean = 9.01 ± 2.46, 8.61 ± 2.01, and 17.62 ± 3.73) and those had not been hospitalized (mean = 8.18 ± 2.22, 8.03 ± 1.78, and 16.21 ± 3.14; p = .003, .011, and .001). There were no significant differences in perceived coping resources scores and FSSS total scores between suicidal behaviors. There was a significant difference of perceived harm scores between patients with (mean = 9.20 ± 2.51) versus without (mean = 8.47 ± 2.33; p = .012) a history of compulsory admission but not in perceived coping resources scores and with a trend-level difference for FSSS total scores (mean = 17.74 ± 3.84 [with compulsory admission] vs. 16.77 ± 3.41 [without compulsory admission]; p = .03; Table 4).

Table 4.

Known-Group Validity of the Family Stigma Stress Scale.

Variable	Perceived Harm		Perceived Coping Resources		Stigma Stress
Variable	Mean (SD)	t (p Value)^a	Mean (SD)	t (p Value)^a	Mean (SD)	t (p Value)^a
Hospitalized		2.95 (.003)		2.57 (.011)		3.45 (.001)
Yes (n = 195)	9.01 (2.46)		8.61 (2.01)		17.62 (3.73)
No (n = 104)	8.18 (2.22)		8.03 (1.78)		16.21 (3.14)
Suicidal behaviors		0.37 (.71)		1.99 (.048)		1.38 (.17)
Yes (n = 112)	8.77 (2.36)		8.68 (2.13)		17.45 (3.55)
No (n = 184)	8.66 (2.40)		8.20 (1.77)		16.86 (3.51)
Compulsory admission		2.53 (.012)		0.92 (.36)		2.22 (.03)
Yes (n = 103)	9.20 (2.51)		8.53 (2.11)		17.74 (3.84)
No (n = 195)	8.47 (2.33)		8.31 (1.83)		16.77 (3.41)

^aWe adjusted the significance level to p < .016.

Discussion

We developed the FSSS as a questionnaire to help mental health professionals understand stigma stress in family caregivers of people with mental illness. We then tested the psychometric properties of the FSSS and found it promising in a sample of 300 Taiwanese family caregivers. We showed that its construct validity, internal consistency, test–retest reliability, concurrent validity, and known-group validity were acceptable.

Our EFA yielded two factors of the FSSS that correspond to the framework proposed by Rüsch et al. (2009a): Family caregivers on the one hand evaluate whether stigma is personally harmful, and on the other hand, they evaluate their resources to cope with the threat of stigma. In addition to the EFA results, our results for internal consistency and corrected item–total correlations showed that each item score was significantly associated with other item scores in the same factor. Therefore, our psychometric results support the stress appraisal processes and confirm the two-factor framework of the FSSS.

We illustrated other promising psychometric properties of the FSSS including concurrent validity with associated measure scores (ASS, MSPSS, CES-D, and RSES) and known-group validity between groups with different clinical characteristics (had been hospitalized, had suicidal behaviors, and had been compulsorily admitted). However, test–retest reliability seemed to be somewhat unsatisfactory because of the wide range of retest intervals. Indeed, we found lower test–retest reliability among those with longer intervals as stigma stress might be affected by life events over time. The strongest association for concurrent validity was between the perceived harm domain score of the FSSS and the ASS total score because both focus on the negative effects of stigma (Mak & Cheung, 2008; Rüsch et al., 2009a).

We also found that both factor scores in the FSSS were moderately correlated with CES-D and RSES scores. This was reasonable because studies (Hammen, 2005; Lin et al., 2016; Link, Struening, Neese-Todd, Asmussen, & Phelan, 2001; Zuckerman, 1989) have reported that stress and stigma are related to depression and self-esteem. The FSSS differentiated between clinical characteristics (yes vs. no), which corresponds to earlier findings among people with mental illness (Chang, Wu, Chen & Lin, 2016; Rüsch et al., 2014b). Therefore, we are confident that the FSSS has sound psychometric properties.

This is the first study to describe the development of a stigma stress questionnaire, the FSSS, for family caregivers of people with mental illness. Because of its promising psychometric properties, researchers might want to consider using the scale to identify important factors for resisting family stigma stress. Mental health professionals might also want to use the FSSS to assess family caregiver stigma stress for the following reasons. By understanding this type of stigma stress, mental health professionals can determine and develop appropriate interventions that diminish its negative effects on the psychological health and emotional well-being of the caregivers. However, we suggest that the associations between the FSSS and psychological health should be confirmed before attempting to design appropriate interventions. For example, depression and self-esteem were related to the FSSS in our findings, and the interventions for stigma stress might also be efficacious for these psychological issues. Other benefits might be gained when the psychological health of the caregivers is well taken care of: The family caregivers can have the energy and willingness to cooperate with mental health professionals to care for the patients; the effectiveness of treating patients with mental illness might be improved because of good cooperation between family caregivers and mental health professionals. In contrast, the social burden might increase if the health of the family caregivers is jeopardized by their stigma stress.

This study has some limitations. First, our findings might not be applicable to the whole of Taiwan because we used a convenience sample of patients recruited in Southern Taiwan. Second, we recruited only family caregivers whose family members were diagnosed with one of the three types of mental illness. Future studies on family caregivers whose family members are diagnosed with other types of mental illness are thus warranted. Third, we did not use a standard research and review process to develop the FSSS: We simply modified an existing instrument of stigma stress for people with mental illness. Therefore, our results might not accurately reflect the differences in stigma stress between family caregivers and people with mental illness. Fourth, we adapted an instrument designed in the Western world (Rüsch et al., 2009a) in an East Asian country, and cultural differences might affect the suitability of the instrument. However, we tried to mitigate any cultural bias by ensuring linguistic validity and content validity in a focus group. Nevertheless, concepts and experiences are different between cultures, and future studies of stigma stress in different cultures are needed to confirm the psychometric properties of the FSSS.

Conclusion

We developed the FSSS, a concise 8-item instrument, to measure stigma stress in family caregivers of people with mental illness. The FSSS can be used regularly to monitor stigma stress and to help mental health professionals detect the emotional distress of family caregivers before they become overly anxious and depressed. Stigma stress treatments can be immediately provided to family caregivers; for example, intervention programs using psychoeducation (Martín-Carrasco et al., 2009), meditation (Waelde, Thompson, & Gallagher-Thompson, 2004), and mindfulness (Whitebird et al., 2013) can be used to reduce stress, and the FSSS can be used to evaluate the effectiveness of these intervention programs. However, additional studies to assess the applicability and properties of the scale in other contexts by examining the psychometric properties of the FSSS are warranted because we examined the reliability and validity of the FSSS only in a Taiwanese sample.

Supplemental Material

supplementary_material - Development of the Family Stigma Stress Scale (FSSS) for Detecting Stigma Stress in Caregivers of People With Mental Illness

supplementary_material for Development of the Family Stigma Stress Scale (FSSS) for Detecting Stigma Stress in Caregivers of People With Mental Illness by Chih-Cheng Chang, Jian-An Su, Kun-Chia Chang, Chung-Ying Lin, Mirja Koschorke, Nicolas Rüsch and Graham Thornicroft in Evaluation & the Health Professions

Footnotes

Authors’ Note

Chih-Cheng Chang and Jian-An Su share equal contributions to the work. The views expressed are those of the author(s) and not necessarily those of the National Health Service, the National Institute for Health Research, or the Department of Health.

Acknowledgment

We thank the staff of the Department of Psychiatry, Chi Mei Medical Center, Tainan, the Department of Psychiatry, Chang Gung Memorial Hospital, Chiayi and Jianan Psychiatric Center for their help with data collection.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Graham Thornicroft is supported by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care South London at King’s College London Foundation Trust. Graham Thornicroft acknowledges financial support from the Department of Health via the NIHR Biomedical Research Centre and Dementia Unit awarded to South London and Maudsley National Health Service (NHS) Foundation Trust in partnership with King’s College London and King’s College Hospital NHS Foundation Trust. Graham Thornicroft and Mirja Koschorke are supported by the European Union Seventh Framework Programme (FP7/2007–2013) Emerald project. This research was supported by grant CMRPG6E0291 from Chang Gung Memorial Hospital.

Supplemental Material

Supplemental material for this article is available online.

References

Angermeyer

M. C.

Dietrich

(2006). Public beliefs about and attitudes towards people with mental illness: A review of population studies. Acta Psychiatrica Scandinavica, 113, 163–179.

Angermeyer

M. C.

van der Auwera

Carta

M. G.

Schomerus

(2017). Public attitudes towards psychiatry and psychiatric treatment at the beginning of the 21st century: A systematic review and meta-analysis of population surveys. World Psychiatry, 16, 50–61. doi:10.1002/wps.20383

Bevans

M. F.

Sternberg

E. M.

(2012). Caregiving burden, stress, and health effects among family caregivers of adult cancer patients. Journal of the American Medical Association, 307, 398–403.

Boyd

J. E.

Emerald

P. A.

Otilingma

P. G.

Peters

(2014). Internalized Stigma of Mental Illness (ISMI) Scale: A multinational review. Comprehensive Psychiatry, 55, 221–231.

Canty-Mitchell

Zimet

G. D.

(2000). Psychometric properties of the Multidimensional Scale of Perceived Social Support in urban adolescents. American Journal of Community Psychology, 28, 391–400.

Chang

C.-C.

J.-A.

Lin

C.-Y.

(2016). Using the Affiliate Stigma Scale on caregivers of people with dementia: Psychometric evaluation. Alzheimer’s Research & Therapy, 8, 45.

Chang

C.-C.

J.-A.

Tsai

C.-S.

Yen

C.-F.

Liu

J.-H.

Lin

C.-Y.

(2015). Rasch analysis suggested three unidimensional domains for Affiliate Stigma Scale: Additional psychometric evaluation. Journal of Clinical Epidemiology, 68, 674–683.

Chang

C.-C.

T.-H.

Chen

C.-Y.

Lin

C.-Y.

(2016). Comparing self-stigma between people with different mental disorders in Taiwan. Journal of Nervous and Mental Disease, 204, 547–553.

Chang

C.-C.

Yen

C.-F.

Jang

F.-L.

J.-A.

Lin

C.-Y.

(2017). Comparing affiliate stigma between family caregivers of people with different severe mental illness in Taiwan. Journal of Nervous and Mental Disease, 205, 542–549.

10.

Cheng

C.-P.

Yen

C.-F.

C.-H.

Yen

J.-Y.

(2012). Factor structure of the center for epidemiologic studies depression scale in Taiwanese adolescents. Comprehensive Psychiatry, 53, 299–307.

11.

Chou

K.-L.

(2000). Assessing Chinese adolescents’ social support: The multidimensional scale of perceived social support. Personality and Individual Differences, 28, 299–307.

12.

Clark

L. A.

Watson

(1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7, 309–319.

13.

Corrigan

P. W.

Larson

J. E.

Rüsch

(2009). Self-stigma and the “why try” effect: Impact on life goals and evidence-based practices. World Psychiatry, 8, 75–81.

14.

Corrigan

P. W.

Michaels

P. J.

Vega

Gause

Watson

A. C.

Rüsch

(2012). Self-stigma of mental illness scale--short form: Reliability and validity. Psychiatry Research, 199, 65–69.

15.

Corrigan

P. W.

Watson

A. C.

Barr

(2006). The self-stigma of mental illness: Implications for self-esteem and self-efficacy. Journal of Social & Clinical Psychology, 25, 875–884.

16.

Goffman

(1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice Hall.

17.

Hammen

(2005). Stress and depression. Annual Review of Clinical Psychology, 1, 293–319.

18.

Hinshaw

S. P.

(2007). The mark of shame: Stigma of mental illness and an agenda for change. Oxford, England: Oxford University Press.

19.

Jacobsson

Ghanea

Törnkvist

(2013). Internalized stigma of mental illness in Sweden and Iran—A comparative study. Open Journal of Psychiatry, 3, 370–374.

20.

Kaiser

C. R.

Major

McCoy

S. K.

(2004). Expectations about the future and the emotional consequences of perceiving prejudice. Personality & Social Psychology Bulletin, 30, 173–184.

21.

Kleinman

Lin

(1981). Normal and abnormal behavior in Chinese culture. Dordrecht, Holland: Springer Science & Business Media.

22.

Lazarus

R. S.

Folkman

(1984). Stress, appraisal, and coping. New York, NY: Springer.

23.

Lin

C.-Y.

Chang

C.-C.

T.-H.

Wang

J.-D.

(2016). Dynamic changes of self-stigma, quality of life, somatic complaints, and depression among people with schizophrenia: A pilot study applying kernel smoothers. Stigma & Health, 1, 29–43.

24.

Link

B. G.

Phelan

J. C.

(2001). Conceptualizing stigma. Annual Review of Sociology, 27, 363–385.

25.

Link

B. G.

Struening

E. L.

Neese-Todd

Asmussen

Phelan

J. C.

(2001). The consequences of stigma for the self-esteem of people with mental illnesses. Psychiatric Services, 52, 1621–1626.

26.

Major

O’Brien

L. T.

(2005). The social psychology of stigma. Annual Review of Psychology, 56, 393–421.

27.

Mak

W. W.

Cheung

R. Y.

(2008). Affiliate stigma among caregivers of people with intellectual disability or mental illness. Journal of Applied Research in Intellectual Disabilities, 21, 532–545.

28.

Mak

W. W.

Cheung

R. Y.

(2010). Self-stigma among concealable minorities in Hong Kong: Conceptualization and unified measurement. American Journal of Orthopsychiatry, 80, 267–281.

29.

Martín-Carrasco

Martín

M. F.

Valero

C. P.

Millán

P. R.

García

C. I.

Montalbán

S. R.

… Vilanova

M. B.

(2009). Effectiveness of a psychoeducational intervention program in the reduction of caregiver burden in Alzheimer’s disease patients’ caregivers. International Journal of Geriatric Psychiatry, 24, 489–499.

30.

Parcesepe

A. M.

Cabassa

L. J.

(2013). Public stigma of mental illness in the United States: A systematic literature review. Administration and Policy in Mental Health and Mental Health, 40, 384–399.

31.

Radloff

L. S.

(1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401.

32.

Rosenberg

(1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

33.

Rüsch

Corrigan

P. W.

Powell

Rajah

Olschewski

Wilkniss

Batia

(2009a). A stress-coping model of mental illness stigma: II. Emotional stress responses, coping behavior and outcome. Schizophrenia Research , 110, 65–71.

34.

Rüsch

Corrigan

P. W.

Wassel

Michaels

Olschewski

Wilkniss

Batia

(2009b). A stress-coping model of mental illness stigma: I. Predictors of cognitive stress appraisal. Schizophrenia Research, 110, 59–64.

35.

Rüsch

Heekeren

Theodoridou

Müller

Corrigan

P. W.

Mayer

… Rössler

(2015). Stigma as a stressor and transition to schizophrenia after one year among young people at risk of psychosis. Schizophrenia Research, 166, 43–48.

36.

Rüsch

Müller

Heekeren

Theodoridou

Metzler

Dvorsky

… Rössler

(2014a). Longitudinal course of self-labeling, stigma stress and well-being among young people at risk of psychosis. Schizophrenia Research, 158, 82–84.

37.

Rüsch

Müller

Lay

Corrigan

P. W.

Zahn

Schönenberger

… Rössler

(2014b). Emotional reactions to involuntary psychiatric hospitalization and stigma-related stress among people with mental illness. European Archives of Psychiatry and Clinical Neurosciences, 264, 35–43.

38.

Schulz

Sherwood

P. R.

(2008). Physical and mental health effects of family caregiving. American Journal of Nursing, 108, 23–27.

39.

Thornicroft

(2006). Shunned: Discrimination against people with mental illness. Oxford, England: Oxford University Press.

40.

Waelde

L. C.

Thompson

Gallagher-Thompson

(2004). A pilot study of a yoga and meditation intervention for dementia caregiver stress. Journal of Clinical Psychology, 60, 677–687.

41.

Whitebird

R. R.

Kreitzer

Crain

A. L.

Lewis

B. A.

Hanson

L. R.

Enstad

C. J.

(2013). Mindfulness-based stress reduction for family caregivers: A randomized controlled trial. Gerontologist, 53, 676–686.

42.

P.-H.

Lai

Y.-C.

Shih

C.-L.

(2013). Evaluation of the Rosenberg Self-Esteem Scale using the rating scale model [in Chinese]. Psychological Testing, 60, 263–289.

43.

Mayer

Müller

Heekeren

Theodoridou

Dvorsky

… Rüsch

(2016). Stigma and suicidal ideation among young people at risk of psychosis after one year. Psychiatry Research, 243, 219–224.

44.

Yang

L. H.

(2007). Application of mental illness stigma theory to Chinese societies: Synthesis and new directions. Singapore Medicine Journal, 48, 977–985.

45.

Yang

H. J.

Soong

W. T.

Kuo

P. H.

Chang

H. L.

Chen

W. J.

(2004). Using the CES-D in a two-phase survey for depressive disorders among nonreferred adolescents in Taipei: A stratum-specific likelihood ratio analysis. Journal of Affective Disorder, 82, 419–430.

46.

Zhang

Norvilitis

J. M.

(2002). Measuring Chinese psychological well-being with Western developed instruments. Journal of Personality Assessment, 79, 492–511.

47.

Zimet

G. D.

Dahlem

N. W.

Zimet

S. G.

Farley

G. K.

(1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52, 30–41.

48.

Zimet

G. D.

Powell

S. S.

Farley

G. K.

Werkman

Berkoff

K. A.

(1990). Psychometric characteristics of the multidimensional scale of perceived social support. Journal of Personality Assessment, 55, 610–617.

49.

Zuckerman

D. M.

(1989). Stress, self-esteem, and mental health: How does gender make a difference? Sex Roles, 20, 429–444.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB