The stigma of having a parent with mental illness: Genetic attributions and associative stigma

Abstract

Children often report associative stigma because they are ‘contaminated’ by association with a parent who has a mental illness. An exploratory study was conducted to investigate the role of genetic attributions in the aetiology of associative stigma. The first hypothesis was that genetic attributions would predict associative stigma over and above the contribution of biochemical and stressful‐event attributions, while the second hypothesis was that the relationship between genetic attributions and associative stigma would be mediated by the perceived likelihood that children would develop the same disorder as their parents. Two‐hundred‐and‐two individuals were asked to read a hypothetical scenario describing a teenage girl whose mother had been diagnosed with either schizophrenia or depression. Both hypotheses were supported. The findings of the study have implications for a number of professions working in the community such as teachers and psychologists. Additional avenues for future research are also explored.

Key words:

• associative stigma
• genetic attributions
• psychological disorders
• social issues
• social psychology
• stigma of mental illness

According to the World Health Organisation (WHO), the stigma associated with a mental illness is currently the ‘most important barrier to overcome in the community’ (WHO, 2001, p. 98). A substantial amount of research supports the claim that the stigma associated with a mental illness is a major problem that must be overcome because individuals who have been diagnosed with a mental illness often experience a considerable degree of ostracism, prejudice, and discrimination in the community (Angermeyer & Matschinger, 2005; Dickerson, Sommerville, Origoni, Ringel, & Parente, 2002; Morgan, Mitchell, & Jablensky, 2005; Morselli & Elgie, 2003; Wahl, 1999). The stigma of a mental illness, however, does not just affect those with a mental illness. A number of studies also indicate that family members of the individual who has a mental illness also experience what has been called associative stigma (Angermeyer, Schulze, & Dietrich, 2003; Chang & Horrocks, 2006; Dore & Romans, 2001; Lefley, 1989; Mehta & Farina, 1988; Phelan, Bromet, & Link, 1998; Veltman, Cameron, & Stewart, 2002). Whereas primary stigma refers to individuals who are ‘marked’ in some way, associative stigma refers to the stigma experienced by individuals who have some degree of association with the ‘marked’ individual (see Corrigan, Watson, & Miller, 2006, p. 239). For example, a girl with a mother who suffers from schizophrenia may be bullied and teased in class, not selected for group games at lunch time, and may be avoided by classmates after school hours because of her mother's illness. Associative stigma has also been used to explain stigma directed at the friends of homosexual individuals (Sigelman, Howell, Cornell, Cutright, & Dewey, 1991), children of lesbian mothers (King & Black, 1999), dating partners of students with physical disabilities (Goldstein & Johnson, 1997), and family members of people with HIV‐AIDS (Poindexter & Linsk, 1999).

Research has provided evidence to demonstrate that associative stigma is a useful concept when attempting to understand the discrimination that can be experienced by children when a parent has a mental illness (e.g. Fudge & Mason, 2004; Mehta & Farina, 1988). For example, Mehta and Farina examined the levels of associative stigma experienced by college‐aged children when their father suffered from a mental illness. One‐hundred‐and‐twenty undergraduate students were presented with a vignette depicting a hypothetical roommate with a father who was labeled as depressed, alcoholic, incarcerated, old, frequently absent, or having only one leg and were asked to make judgments about their roommate in terms of their academic performance, friends, career, and family. The study provided evidence to support the concept of associative stigma because children were ‘contaminated’ by their father's diagnosis (p‐value for all main effects was <.01). For instance, children who had a father with a psychiatric illness (e.g., depression, alcohol abuse) were subject to extremely high levels of associative stigma.

Genetic attributions

Attribution theory may be a useful perspective to adopt when attempting to understand the aetiology of associative stigma. The primary focus in attribution theory is to examine the attributions that people make when considering the causes of events, outcomes, or illnesses (Heider, 1958). According to Heider (1958), attributions are important because attributions of causality are ‘everyday occurrences that determine much of our understanding of and reaction to our surroundings’ (p. 16). A number of researchers have speculated that attributions play a significant role in the aetiology of associative stigma, and, in particular, the tendency to cite genetic attributions about the causes of mental illness (Hinshaw, 2005; Mehta & Farina, 1988; Phelan, 2002, 2005). The first aim of the study was to test the speculation already reported in the research literature that the tendency to endorse genetic attributions is related to higher levels of associative stigma. A review of the research literature reveals, however, that biochemical and stressful‐event attributions may also be related to associative stigma because biochemical and negative life events are often cited by scientists as potential causes of mental illness (see Brown & Harris, 1989; Friedman & Silver, 2007; Wedding & Stuber, 2006). The first hypothesis in the present study was therefore designed to provide a more thorough test of the hypothesis that genetic attributions play a significant role in the aetiology of associative stigma. If genetic attributions are central to the aetiology of associative stigma when considering the children of parents with a mental illness, then genetic attributions will predict associative stigma over and above the contribution of biochemical and stressful‐event attributions.

In addition to examining the relationship between genetic attributions and associative stigma, the present study provides a further test of the proposal that genetic attributions have a significant role to play in the aetiology of associative stigma. Mehta and Farina (1988) formulated but did not test the hypothesis that genetic attributions play a role in the aetiology of associative stigma as ‘people may think something similar could happen to the offspring’ and that ‘some people probably do not want to be friends with a potentially mentally ill person’ (p. 199). The second hypothesis was based on the speculation by Mehta and Farina, and predicts that genetic attributions are related to associative stigma because a tendency to endorse genetic attributions will result in an increased belief that a child will develop the disorder, which in turn will result in higher levels of associative stigma. An experimental study by Phelan (2005) provides preliminary support for the hypothesis to be evaluated in the present study, although the findings are of limited relevance because Phelan only asked subjects to consider the siblings of individuals who had a mental illness. Phelan interviewed 641 subjects and asked them a series of questions about a vignette in which an individual is suffering from a serious illness (i.e., schizophrenia, depression, ruptured disc). The cause of illness was experimentally manipulated and ascribed to genetic, partly genetic, or non‐genetic causes. Phelan reports that the tendency to endorse genetic attributions was not only related to an increased belief that people will develop the same disorder as their siblings, but was also related to an increasing tendency to avoid people who have a sibling who suffers from a mental illness. In conclusion, the second hypothesis was that the relationship between genetic attributions and associative stigma would be mediated by the perceived likelihood that children will develop the same disorder as their parents.

METHOD

Participants

Data were collected from 202 individuals who were surveyed at a range of locations around Adelaide. Thirty‐six participants (17.8%) were first‐year psychology students at Flinders University, 46 participants (22.8%) were recruited from the general population at Flinders University, and 120 (59.4%) were members of the wider community. Two participants failed to provide demographic information. Of the remaining 200 participants, 128 were female and 72 were male, with ages ranging from 18 to 82 years (M = 36.23, SD = 15.96).

Design

The design used in the present study was loosely based on a number of studies that have been reported in the research literature (e.g., Corrigan, Watson, & Miller, 2006; Phelan, 2005). Participants were assigned to one of two disorder conditions (schizophrenia or depression) on a rotational basis to ensure no systematic bias in the allocation process (i.e., first person to one condition, second person to the other condition and so on until all the participants were allocated). Each condition was represented by a vignette that described a 14‐year‐old girl whose mother had been diagnosed with either schizophrenia or depression.

Materials

Demographics

Participants were asked to provide details about their age and sex.

Vignette scenarios

The vignette descriptions used in the present study were based on the vignettes employed by Corrigan et al. (2006) and Phelan (2005). The illnesses were portrayed as long‐term illnesses that had a significant impact on the parent's life to prompt participants to think about the effect on the child. The decision to include descriptions of behaviours associated with the illnesses (as used in previous studies) was based on the suggestion by Link, Yang, Phelan, and Collins (2004, p. 515) that introducing ‘the actual behaviors indicative of mental illness’ may help ‘elicit stigma processes such as labeling and stereotyping.’ The vignettes were designed to contain minimal information about the child so that participants would respond to questions based on their own preconceptions. The child was aged 14 years because, according to Phelan (p. 318), ‘issues of genetic contamination are most salient for younger relatives, individuals who will be seeking mates and who may be seen as squarely in the risk period for developing the illness themselves’. (Copies of the vignettes employed in the present study are available from the first author.)

Attributions

Participants' attributions about the cause of the disorder depicted in the vignette were assessed by responses to three items using a 7‐point Likert‐type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Genetic attributions were assessed by response to the question ‘[Schizophrenia/Depression] may be caused by genetic or hereditary factors’. The question ‘[Schizophrenia/Depression] may be caused by a chemical imbalance in the brain’ measured biochemical attributions. Stressful‐event attributions were measured with the question ‘[Schizophrenia/Depression] may be caused by stressful life events’. For each item, scores ranged from 1 to 7 with higher scores indicating stronger belief in the cause of the illness.

Perceived likelihood the child will develop the disorder

One question was included to measure participants' beliefs regarding the likelihood the child will develop the disorder from which the parent suffers (‘How likely is it that Anna will experience the same illness as her mother?’). A 7‐point Likert‐type scale was used, ranging from 1 (much less likely than the average child) to 7 (much more likely than the average child). Higher scores indicated a stronger belief that the child would develop the same mental illness as their parent.

Associative stigma

The measure of associative stigma was based on a number of measures that had been used in previous research to evaluate associative stigma when children have parents with a mental illness. Three common themes were evident across the various measures, with children stereotyped as having negative personal characteristics (e.g., sad, unfriendly), viewed as dysfunctional (currently and in the future), and as experiencing social rejection (e.g., Burk & Sher, 1990; Corrigan & Miller, 2004; Corrigan et al., 2006; Fink & Tasman, 1992; Fudge & Mason, 2004; Gladstone, Boydell, & McKeever, 2006; Mehta & Farina, 1988).

The scale consisted of three sections. The first section was a semantic differential scale designed to assess stereotypes regarding the child's personal characteristics. Participants were asked to rate the child in terms of adjective pairs such as sad/happy, gentle/aggressive, and sane/crazy. The section included nine questions that were rated on a scale of 1–7, with two of the items reverse scored, and higher scores indicating more negative attitudes. The second section asked participants to rate the functioning of the child in various domains (e.g., ‘how likely is it that Anna is functioning well?’). The scale was designed to measure stereotypes about the functioning of the child, with higher scores indicating greater dysfunction. The subscale included 12 items (three were reverse scored) with items rated on a scale of 1 (much less likely than the average child) to 7 (much more likely than the average child). The third section was designed to measure social distance and consisted of 10 questions. The scale asked participants how willing they would be to interact with Anna in various situations on a scale from 1 (very willing) to 7 (very unwilling). Items included ‘how willing would you be to become close friends with Anna?’ Higher scores indicated greater desire for distance from the child. The final score was derived by summing all 31 items with reverse coding where appropriate. Potential scores ranged from 31 to 217, with higher scores reflecting higher levels of associative stigma. The measure of associative stigma developed and used in the present study was of high internal consistency (Cronbach's α = 0.92).

Procedure

Participants were given a copy of the questionnaire booklet, which contained instructions, vignette, demographic questions, and a series of items designed to evaluate attributions, perceptions that the child will develop the same disorder as their parents, and associative stigma. Participants were given the option of returning the questionnaire directly to the researcher, or posting it to the researcher via a reply‐paid envelope.

RESULTS

Genetic attributions and associative stigma

The first hypothesis was that genetic attributions would predict associative stigma over and above the contribution of biochemical and stressful‐event attributions. A simultaneous regression analysis was conducted separately for the schizophrenia and depression conditions (see Table-1). In the condition where a parent has schizophrenia, the first hypothesis was supported, because genetic attributions were a significant predictor of associative stigma (with genetic attributions explaining 11.9% of the variance in associative stigma), and there was no evidence of a relationship between biochemical or stressful‐event attributions and associative stigma. Consistent with the speculations noted by a number of researchers (Hinshaw, 2005; Mehta & Farina, 1988; Phelan, 2002, 2005), the findings of the regression analysis indicated that the tendency to endorse genetic attributions was related to higher levels of associative stigma. The first hypothesis was again supported when the analyses were repeated, but with a parent who has depression (with genetic attributions explaining 8.41% of the variance in associative stigma).

Table 1 Summary of simultaneous regression analyses for causal attributions to predict associative stigma

Variable	Schizophrenia		Depression
	B	ß	B	ß
Genetic attributions	5.75**	0.37**	4.20*	0.31*
Biochemical attributions	−3.10	−0.14	−4.03	−0.21
Stressful‐Event attributions	−2.37	−0.17	−0.21	−0.01

* p < .05,

** p < .01.

Mediation hypothesis

The second hypothesis in the present study was that the relationship between genetic attributions and associative stigma would be mediated by the perceived likelihood that the child will develop the same disorder as their parents. The preconditions required in order to test mediation are that significant correlations are evident between the predictor, mediator, and outcome variables (Baron & Kenny, 1986). A series of correlations were calculated, which indicated that the required assumptions were met for both vignettes that involved mothers who had either schizophrenia or depression. With regards to the correlations in the schizophrenia condition, genetic attributions were correlated with the perceived likelihood that the child will develop the disorder, r(66) = 0.69, p < .001, and with associative stigma, r(66) = 0.31, p = 0.01. The likelihood that the child will develop the disorder was also correlated with associative stigma, r(66) = 0.42, p < .001. With regards to the correlations in the depression condition, genetic attributions were correlated with the perceived likelihood that the child will develop the disorder, r(66) = 0.42, p < 0.001, and associative stigma, r(66) = 0.26, p = .035. The likelihood that the child will develop the disorder was also correlated with associative stigma, r(66) = 0.32, p = .008.

The mediation hypothesis was tested using regression analyses for the schizophrenia and depression conditions separately. First, a simple regression analysis was conducted to test the relationship between genetic attributions and associative stigma. The results indicated that genetic attributions were a significant predictor of associative stigma when a child had a parent who suffered from schizophrenia, accounting for 9.5% of the variance, R² = 0.095, F(1, 66) = 6.92, p = 0.01, as well as when a child had a parent who suffered from depression, accounting for 6.6% of the variance, R² = 0.066, F(1, 66) = 4.65, p = 0.035.

A hierarchical multiple regression analysis was then performed to examine whether significantly less variance in associative stigma would be found after controlling for the perceived likelihood that the child will develop the disorder as their parents (see Table-2). The perceived likelihood that the child will develop the same disorder as their parents was entered in Step 1. Genetic attributions and the perceived likelihood that the child will develop the same disorder as their parents was then entered in Step 2. For the schizophrenia condition, in Step 1, the perceived likelihood that the child will develop the same disorder as their parents accounted for a significant 18% of the variance in associative stigma. At Step 2, genetic attributions did not explain any additional variance in associative stigma, supporting the speculation that the relationship between genetic attributions and associative stigma is mediated by the perceived likelihood that the child will also develop schizophrenia. The results were replicated when analyses were repeated, but for a child who had a parent with depression (see Table-3).

Table 2 Hierarchical regression to evaluate mediation hypothesis (schizophrenia condition)

Variable	B	ß
Step 1
Likelihood child develops disorder	8.76***	0.42***
R^2 = 0.18, F(1, 66) = 14.48, p < .001
Step 2
Likelihood child develops disorder	8.36**	0.40**
Genetic attributions	0.44	0.03
R^2change = 0.00, Fchange(1, 65) = 0.03, p = .86

* p < .05,

** p < .01,

*** p < .001.

Table 3 Hierarchical regression to evaluate mediation hypothesis (depression condition)

Variable	B	ß
Step 1
Likelihood child develops disorder	6.96**	0.32**
R^2 = 0.10, F(1, 66) = 7.60, p < .01
Step 2
Likelihood child develops disorder	5.62*	0.26*
Genetic attributions	2.00	0.15
R^2change = 0.02, Fchange(1, 65) = 1.35, p = .25

* p < .05,

** p < .01,

*** p < .001.

DISCUSSION

The primary aim of the present study was to examine hypotheses about the role of genetic attributions in the aetiology of associative stigma when considering the children of parents with a mental illness. The data were consistent with the first hypothesis, and supports the speculation that genetic attributions predict associative stigma over and above the contribution of biochemical and stressful‐event attributions. The second hypothesis was also consistent with the data and supports the speculation that the relationship between genetic attributions and associative stigma is mediated by the perceived likelihood that children will develop the same illness as their parent. Overall, the findings support the speculation by a number of authors (e.g., Mehta & Farina, 1988; Phelan, 2002, 2005) that genetic attributions are an important consideration when attempting to understand the aetiology of associative stigma.

The findings in the present study have implications for a number of professions who work in the community with parents who have been diagnosed with a mental illness. Although the exact figures are unknown, it has been estimated that 21–23% of Australian children live with at least one parent who has been diagnosed with a mental illness (Maybery, Reupert, Patrick, Goodyear, & Crase, 2005). The findings of the present study lend some support to the claim that not only do children have to cope with the trauma of having a parent who is experiencing a mental illness (Angermeyer & Matschinger, 2005; Morgan et al., 2005; Wahl, 1999), children also have to cope with the experience of associative stigma (Fudge & Mason, 2004; Mehta & Farina, 1988; Meiser et al., 2007). A notable consequence is that professionals who work in the community, such as teachers, psychologists, social workers, and government agencies, need to be aware that when a parent is diagnosed with a mental illness, consideration and support may be needed for the children as well as for the parents. For example, teachers may need to guard against the emergence of associative stigma and be willing to offer consideration and support to school children who have parents with a mental illness. An important proviso, of course, is that not all children who have a parent with a mental illness will want or need assistance at school (see Fudge & Mason, 2004). The most suitable response would be for teachers, psychologists, and other professionals to be aware of the problems raised by associative stigma, and be ready to offer support if requested or needed by the child (for further details, see Couzins, 1999). Another important proviso is that professionals need to be careful not to draw undue attention to the children who have a parent with a mental illness in case the undue attention initiates or promotes the chances that the child may experience associative stigma (again, see Couzins, 1999).

The findings in the present study are limited by a number of considerations that provide a potential focus for future research. For example, a notable limitation of the present study is that the methodology was correlational and limits the opportunity to make causal claims. An obvious avenue for future research would be to further investigate the relationship between genetic attributions and associative stigma by experimentally manipulating the perceived likelihood that a child will develop the same disorder as their parents. The benefit of an experimental manipulation would be to provide a more thorough test of the causal relationships believed to generate the association between genetic attributions and associative stigma (for specific examples, see Phelan, 2005). Another limitation of the study is that no consideration was given to other variables that may be involved in the aetiology of associative stigma, such as attributional style, health beliefs, and social norms (see Angermeyer & Matschinger, 2005; Corrigan et al., 2006; Hinshaw, 2005; Wahl, 1999). Studies focusing on primary stigma with regards to mental illness have found evidence, for example, to indicate that higher levels of contact with individuals who have a mental illness are correlated with lower levels of stigma (Corrigan, Edwards, Green, Diwan, & Penn, 2001; Corrigan, Green, Lundin, Kubiak, & Penn, 2001; Link & Cullen, 1986). Hence, another avenue for future research would be to explore the possibility that a range of other variables, such as contact with people who have a mental illness, may be helpful in reducing the levels of associative stigma in the community.

Another opportunity for future research concerns the presentation of genetic information in the media. The findings of genetic research are frequently publicised in the Australian media and often highlight the positive role that ‘genetic breakthroughs’ have for people who have a mental illness (e.g., McLean, 2008; Macey, 2008). A number of journalists have argued, for instance, that genetic breakthroughs should bring ‘comfort’ to the families of people with a mental illness (see McLean, 2008). Phelan (2002) has suggested, however, that genetic breakthroughs when advertised widely in the media may unexpectedly have a detrimental impact on the children of parents with a mental illness. For example, there is the possibility that reports, such as those by McLean and Macey, may actually increase the levels of associative stigma experienced by children because the reports inadvertently promote the perception that the children who have a parent with a mental illness will inevitably develop the disorder themselves. Nevertheless, the proposal of a relationship between genetic information in the media and the levels of associative stigma experienced by children remains speculative, and at this stage requires further empirical investigation.

In conclusion, the present study provides support for the proposal that genetic attributions are an important consideration when attempting to understand the aetiology of associative stigma. The most notable implication of the present study is that professionals such as teachers and psychologists need to be aware that when parents are diagnosed with a mental illness, consideration and support may be necessary for the children as well as for the parents.

ACKNOWLEDGEMENT

We would like to extend our thanks to the Associate Editor and two anonymous referees who provided constructive feedback on an earlier draft of this paper.