Stigma as a Barrier to Recovery

Adverse Effects of Perceived Stigma on

Social Adaptation of Persons Diagnosed
With Bipolar Affective Disorder
Deborah A. Perlick, Ph.D.
Robert A. Rosenheck, M.D.
John F. Clarkin, Ph.D.
Jo Anne Sirey, Ph.D.
Jamelah Salahi, B.S.
Elmer L. Struening, Ph.D.
Bruce G. Link, Ph.D.

R
Objective: The purpose of this study was to evaluate the effect of concerns esearch over the past four dec-
about stigma on social adaptation among persons with a diagnosis of bipo- ades has compellingly dem-
lar affective disorder. Methods: The sample comprised 264 persons who onstrated that individuals di-
were consecutively admitted to a psychiatric inpatient or outpatient serv- agnosed as having mental illness are
ice at a university-affiliated hospital and who met research diagnostic cri- socially stigmatized or discriminated
teria for bipolar I disorder, bipolar II disorder, or schizoaffective disor- against on several dimensions by key
der, manic type. Patients were evaluated with use of the Schedule for Af- individuals in their social networks and
fective Disorders and Schizophrenia, Lifetime Version (SADS-L), the communities (1). For example, studies
Brief Psychiatric Rating Scale (BPRS), and a measure of perceived stig- have found that employers (2–4), fam-
ma. Social adjustment was measured at baseline and seven months later ilies of patients (5), mental health work-
with the Social Adjustment Scale (SAS). Results: As predicted, patients ers (6), and prospective landlords (7)
who had concerns about stigma showed significantly more impairment at all endorsed devaluing statements
seven months on the social leisure subscale but not on the SAS extended about or discriminated against mental-
family subscale, after baseline SAS score and symptom level had been ly ill individuals. The degree of stigma-
controlled for. More refined models using SAS-derived factors as de- tization has been found to be positive-
pendent variables indicated that concerns about stigma predicted higher ly associated with the manifest severi-
avoidance of social interactions with persons outside the family and psy- ty of the mental disorder (8); however,
chological isolation at seven-month follow-up, after baseline SAS and even persons who have minimal signs
BPRS scores had been controlled for. Conclusions: Concerns about the of mental illness, that is, those who ap-
stigma associated with mental illness reported by patients during an acute pear “troubled,” may be stereotyped
phase of bipolar illness predicted poorer social adjustment seven months and rejected (9).
later with individuals outside the patient’s family. Greater attention to pa- Link and colleagues (10,11) have
tients’ concerns about stigma is needed from both researchers and clini- argued that because persons with
cians. (Psychiatric Services 52:1627–1632, 2001) mental illness internalize the devalu-
ing or discriminatory attitudes of soci-
ety at large, they anticipate discrim-
ination or rejection by others and de-
Dr. Perlick and Dr. Rosenheck are with the Northeast Program Evaluation Center at the velop coping strategies, such as secre-
Veterans Affairs Medical Center in West Haven, Connecticut, and Yale University School of cy about their illness or withdrawal
Medicine in New Haven. Dr. Clarkin and Dr. Sirey are with the department of psychia- from social interaction, in an effort to
try of the New York Presbyterian Hospital, Westchester Division, of the Joan and Sanford avoid the rejection they anticipate.
I. Weill Medical College of Cornell University in White Plains, New York. Ms. Salahi is in
Goffman (12) has suggested that to
the master’s program at the Social Work School of the University of Connecticut in West
avoid discrimination and rejection,
Hartford. Dr. Struening and Dr. Link are with the epidemiology of mental disorders pro-
gram at Columbia University and New York State Psychiatric Institute in New York City. persons with mental illness may limit
Send correspondence to Dr. Perlick, Northeast Program Evaluation Center, Department their social interaction to individuals
of Veterans Affairs, VA Connecticut Healthcare System, 950 Campbell Avenue, West who are similarly stigmatized or who
Haven, Connecticut 06516 (e-mail, deborah.perlick@ yale.edu). This paper is part of a spe- are aware and accepting of the stig-
cial section on stigma as a barrier to recovery from mental illness. ma—for example, family members.
PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12 1627
 To test Goffman’s theory, Link and Methods scale on which 1 indicates not present
colleagues (11) evaluated the associ- Subjects and 7 indicates extremely severe. Pos-
ation between reports of withdrawal The sample comprised 264 persons sible scores range from 7 to 168. An
from social interaction by persons aged 16 years or older who were con- intraclass correlation coefficient was
with mental illness in response to secutively admitted to a university-af- calculated for the four raters on the
concerns about stigma and their re- filiated psychiatric inpatient or outpa- basis of two videotaped interviews,
liance on individuals inside their tient service and who had a lifetime using all 24 items. Coefficients for
households rather than those out- diagnosis of bipolar depression with tape 1 were .83, .82, .85, and .96; for
side for emotional and practical sup- mania (bipolar I disorder), hypoma- tape 2, coefficients were .88, .87, 1,
port. As predicted by Goffman’s the- nia (bipolar II disorder), or schizoaf- and .90. Internal consistency for the
ory, they found that reports of with- fective disorder, manic type, accord- expanded BPRS was acceptable (Cron-
drawal in response to concerns ing to research diagnostic criteria (14). bach’s alpha=.76).
about stigma were positively associ- Because the study was carried out as Patients’ concerns about stigma
ated with reliance on individuals part of a study on family burden in were evaluated with use of a scale
within the household for support bipolar illness (15), only patients composed of eight items that meas-
but were negatively associated with whose family members consented to ure withdrawal as a coping mecha-
reliance on individuals outside the be studied were included. Fifty-seven nism designed to avoid rejection (10)
household. In other words, persons percent of eligible patients agreed to and 12 items from Link’s Beliefs
with mental illness who reported participate. Additional details about About Devaluation–Discrimination
avoiding social interaction to avoid sample selection are available else- Scale (11). Withdrawal was measured
exposure to rejection because of where (15). Institutional review board by asking the respondents to indicate
their mental illness also indicated approval was obtained for this study, the extent to which they agreed with
that they turned to members of their and all the participants gave informed statements such as “It is easier for me
own family rather than to persons consent. to be friends with people who have
outside the family for emotional and Patients were enrolled in the study been psychiatric patients” and “After
practical support. between October 1993 and Septem- being in psychiatric treatment, it is a
One implication of these findings is ber 1995; seven-month follow-up data good idea to keep what you are think-
that the adoption of coping strategies were collected approximately through ing to yourself.”
that reduce the stigmatized person’s April 1996. The attrition rate over the Beliefs about devaluation and dis-
range of social contacts—for exam- seven-month follow-up period was 20 crimination were measured by having
ple, withdrawal—may in fact further percent. A comparison of the socioe- the respondents report the extent to
handicap social adaptation and delay conomic and clinical characteristics of which they agreed with statements
recovery or limit the prospects of re- the participants who remained in the such as “Most people feel that enter-
covery. To investigate this possibility, study at seven months with those lost ing a mental hospital is a sign of per-
we evaluated the impact of concerns to follow-up found that the latter sonal failure” and “Most employers
about stigma among persons with a were more likely to be inpatients at will pass over the application of a for-
diagnosis of bipolar affective disorder baseline (χ2=4.52, df=1, p=.03). No mer mental patient.” All statements
on their social adaptation over time other significant differences were were rated on a 4-point scale with an-
within and outside of the family. found. chors ranging from “agree strongly”
Most studies of the stigma associat- to “disagree strongly.” The scale
ed with mental illness have focused Measures demonstrated adequate internal con-
on persons with schizophrenia or Baseline assessments were conducted sistency (Cronbach’s alpha=.83).
chronic mental illness; however, within one week of discharge from The interview version of the Social
there are indications that persons the index inpatient admission or with- Adjustment Scale (SAS) (19) was
with bipolar illness may also be ex- in one week of entry into a new epi- used at baseline and at seven-month
posed to stigma (13). On the basis of sode of outpatient care. The Schedule follow-up to evaluate participants’
findings from previous studies, we for Affective Disorders and Schizo- adaptive social functioning over the
hypothesized that strong concerns phrenia, Lifetime Version (SADS-L) previous three months. The extended
about stigma at baseline would pre- (16) was used to establish lifetime di- family subscale and the social leisure
dict impaired social functioning over agnosis and to rate the nature of the subscale of the SAS were used to
time among persons diagnosed as patient’s index episode of illness— evaluate participants’ adaptive social
having bipolar illness. We further manic versus depressed. The expand- functioning with their family and with
predicted that after baseline social ed version of the Brief Psychiatric their broader social network, respec-
functioning had been controlled for, Rating Scale (BPRS) (17), developed tively. The extended family subscale
concerns about stigma would be by Lukoff and colleagues (18) to in- assesses the quality of the respon-
found to have a more serious impact corporate the psychotic and affective dent’s relationships with his or her
on patients’ social interaction with in- symptoms associated with bipolar dis- parents, siblings, in-laws, and chil-
dividuals outside their families than order, was used to evaluate symptom dren living away from home along
on their interaction with family severity. The 24-item, interviewer-ad- eight dimensions: friction, reticence,
members. ministered instrument uses a 7-point withdrawal, dependency, rebellious-
1628 PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12
ness, worry, guilt, and resentment. The Table 1
social leisure subscale assesses the Characteristics at baseline of individuals diagnosed as having mental illness who par-
quantity and quality of social interac- ticipated in a study of the effects of concerns about stigma on social functioning1
tions outside the family, including the
number of close friends and social in- Characteristic N or mean±SD % Range
teractions and the experienced de-
gree of friction, social discomfort, Female 158 58
hurt feelings, boredom, loneliness, Age (years) 38.46±13.55 15–82
and ease of confiding. Ethnicity
White 205 84
The global ratings made immedi- Nonwhite 38 16
ately after the interview were used to Socioeconomic status2
measure these dimensions of social I to II 49 22
adaptation over the previous three III 72 33
months and to assess overall adjust- IV to V 98 45
Marital status
ment during this time frame. Patients Married or cohabiting 68 27
were rated on 7-point Likert scales on Single 184 73
which 1 indicates excellent adjust- Living with family 142 54
ment and 7 indicates very severe mal- Inpatient 165 64
adjustment. The SAS marital and SADS-L3 diagnosis
Bipolar I 135 53
parental subscales, which assess rela- Bipolar II 25 10
tionships with spouses and children Schizoaffective disorder, manic type 93 37
living at home, respectively, were not Age at onset of illness (years) 20.85±11.60 10–82
included in this study, because only a Lifetime number of psychiatric
relatively small number of patients admissions 5.58±8.73 0–75
had spouses and children. The work 1 Ns range from 219 to 264.
subscale was not used, because it 2 Based on the Hollingshead-Redlich two-point scale (30); I indicates higher socioeconomic status.
3 Schedule for Affective Disorders and Schizophrenia, Lifetime Version
combines data on adjustment within
the family—for example, functioning
as a homemaker—with data on ad-
justment outside the family—for ex- total BPRS score at baseline, the total were white, female, and single. Most
ample, adaptation to competitive em- stigma score at baseline, and the pa- had a primary diagnosis of bipolar I
ployment—and thus precludes exam- tient’s marital status were entered si- disorder and a relatively large number
ination of our hypothesis on the ef- multaneously to identify the contri- of psychiatric inpatient admissions.
fects of stigma on social adaptation bution of the stigma variable to social Onset of the illness had occurred at a
with family members compared with outcome while controlling for the ef- relatively early age. All the partici-
others. fects of other potential explanatory pants met criteria for a current affec-
factors. tive episode at baseline—49 percent
Analyses To further evaluate significant find- for a manic spectrum episode and 51
Bivariate correlations were calculated ings from the initial regression analy- percent for a depressive spectrum
with the Pearson r product-moment ses, the seven-month item-level data episode. About two-thirds were inpa-
correlation coefficient to identify so- from the global scales were subjected tients when they entered the study.
ciodemographic variables and clinical to a principal components analysis, The participants’ mean±SD BPRS
characteristics—for example, inpa- and additional, exploratory models score at baseline was 39.42±10.34.
tient versus outpatient and bipolar were run by using the resulting factor The mean global ratings assigned on
spectrum diagnosis—associated with scores as dependent variables. This an- the SAS at baseline and at seven-
the outcome variables for inclusion in alytic strategy enabled us to specify month follow-up, respectively, were
the multivariate models. Of all the more precisely the dimensions in so- 3.02±1.25 and 2.83±1.15 for the ex-
variables evaluated, only the partici- cial adjustment at seven months that tended family subscale, 3.49±1.41 and
pant’s marital status and BPRS score were affected by the patients’ percep- 3.27±1.36 for the social leisure sub-
at baseline were significantly correlat- tions of stigma at baseline. In these scale, and 3.91±1.14 and 3.58±1.21
ed with any of the measures of social models the global score for the rele- for the overall adjustment subscale.
adjustment; these were included in vant domain was used to control for
the multivariate models described the effects of baseline functioning on Predictors of social
below. seven-month adaptation. adjustment: global scales
Multiple regression models were As Table 2 shows, concerns about
then run with each of the three glob- Results stigma significantly predicted adjust-
al measures of social adjustment as an Sample characteristics ment at seven-month follow-up as
outcome variable. In each model, the Demographic and clinical character- measured on the social leisure sub-
corresponding baseline value for that istics of the sample are presented in scale (b=.151, t=2.37, df=177, p=
dimension of social functioning, the Table 1. Most of the participants .019), even after symptom level, base-
PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12 1629
Table 2
Baseline predictors of study participants’ scores on subscales of the Social Adjustment Scale (SAS) at seven-month follow-up

Social leisure subscale1 Extended family subscale2 Overall adjustment subscale3

Predictor β b p β b p β b p

SAS score .491 .508 <.001 .393 .413 <.001 .522 .498 <.001
Brief Psychiatric
Rating Scale score .019 .006 .93 .164 .063 .37 .049 .018 .81
Marital status –.229 –.075 .25 –.056 –.022 .75 –.194 –.072 .27
Stigma total score .420 .151 .02 .250 .105 .13 .261 .106 .10
1 r2=.304, adjusted r2=.288
2 r2=.217, adjusted r2=.199
3 r2=.310, adjusted r2=.284

line functioning, and sociodemo- el, were not significant predictors in positive and significant predictors of
graphic covariates had been con- any domain, after baseline social psychological isolation and behavioral
trolled for. By contrast, concerns adaptation had been controlled for. avoidance. Baseline social leisure
about stigma at baseline were not a functioning was also a positive and
significant predictor of social adjust- Factor analysis and significant predictor of both factors,
ment at seven months on the SAS ex- exploratory models whereas baseline BPRS scores pre-
tended family subscale, either before The principal components analysis dicted poorer functioning on the psy-
or after symptom level and baseline for the SAS social leisure subscale chological isolation factor alone. In-
functioning in this domain had been produced three orthogonal factors. terestingly, a participant’s marital sta-
controlled for. Thus, as hypothesized, Factor 1, psychological isolation, had tus was differentially related to the
individuals who reported higher lev- high loadings (greater than .50) on so- psychological isolation and behavioral
els of concern about stigma at base- cial discomfort, loneliness, and bore- avoidance factors. Being married pre-
line had more impaired social func- dom. Factor 2, behavioral avoidance, dicted decreased psychological isola-
tioning in interactions with persons had high loadings on diminished con- tion but increased behavioral avoid-
outside their family but not in inter- tacts, reticence, and diminished inter- ance seven months after the index
actions with family members. actions. Factor 3, rejection sensitivity, episode of illness.
The effect of stigma in the model had high loadings on friction and hy-
predicting seven-month functioning persensitivity. Together the three fac- Discussion
on the overall adjustment subscale tors explained 53.7 percent of the cu- The findings of this study demon-
was not significant. Baseline social mulative variance. strate that concerns about stigma as-
adjustment was a significant predictor As Table 3 shows, the regression sociated with mental illness reported
of functioning at seven months in all models that used the three seven- by persons diagnosed as having bipo-
three domains. Conversely, marital month social leisure scale factors as lar affective disorder during an acute
status and baseline BPRS score, al- dependent variables demonstrated phase of their illness adversely affect-
though significant at the bivariate lev- that stigma concerns at baseline were ed an aspect of their social adaptation

Table 3
Baseline predictors of study participants’ scores on the social leisure subscale of the Social Adjustment Scale (SAS) at seven-
month follow-up

Social leisure subscale factors

Psychological isolation1 Behavioral avoidance2 Rejection sensitivity3

Predictor β b p β b p β b p

SAS score .171 .228 <.01 .332 .447 <.001 –.010 –.014 .87
Brief Psychiatric
Rating Scale score .454 .197 .01 .074 .032 .66 .265 .112 .18
Marital status –.420 –.189 .01 .321 .146 .04 –.199 –.088 .28
Stigma total score .301 .144 <.05 .286 .138 .05 .069 .032 .69
1 r2=.206, adjusted r2=.186
2 r2=.239, adjusted r2=.219
3 r2=.025, adjusted r2=.000

1630 PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12

seven months later. As predicted, pa- was associated with being married. ness should also be investigated.
tients with strong concerns about These findings suggest that these A limitation of this study was that it
stigma at baseline showed greater im- strategies may be employed by differ- did not address social adaptation rela-
pairment in their subsequent social ent subgroups of patients. tive to employment or members of
and leisure functioning, even after Additional research is needed to the immediate family. Future studies
symptom severity, baseline social replicate the findings of this study should examine the impact of con-
adaptation, and sociodemographic and to clarify the ways in which con- cerns about stigma on social function-
characteristics had been controlled cerns about stigma and coping strate- ing in these areas.
for. gies affect social behavior outside the The impairment in social and
This study was designed in part to family. Although the findings are con- leisure functioning associated with
test the hypothesis that stigma-relat- sistent with our hypothesis that pa- concerns about stigma has implica-
ed impairment in social adaptation tients exercise avoidant coping strate- tions for the health and well-being of
results from avoidant coping strate- gies selectively in anticipation of re- persons diagnosed as having bipolar
gies, such as secrecy and withdrawal jection by individuals outside the illness. First, the extent and quality of
(11), that patients use to minimize family, an alternative explanation of social interactions have an important
their exposure to discrimination from bearing on quality of life (20,21). Sec-
individuals outside their family or ond, research on social support has
peer group. As hypothesized, we consistently shown that the absence
found that concerns about stigma of close or confiding relationships is
were associated with poorer function- associated with greater risk of relapse
ing on the SAS social leisure subscale, Individuals or nonremission among individuals
which evaluates relationships with in- with depression (22–25). Because
dividuals outside the family, but not who reported chronic symptoms of depression are
with functioning on the extended associated with greater risk of med-
family subscale, which evaluates rela- higher levels of concern ical illness (26), the adverse impact of
tionships within the family. stigma on social functioning could af-
Our data thus represent an exten- about stigma at baseline fect the physical health of persons di-
sion of the cross-sectional findings of agnosed as having mental illness.
Link and colleagues (11) to a differ- had more impaired social Although our findings underscore
ent patient sample, to additional di- the need for interventions to reduce
mensions of social functioning, and to functioning in interactions the adverse impact of stigma for per-
a longitudinal time frame. The find- sons with bipolar disorders and other
ing that concerns about stigma were with persons outside mental illnesses, the results of previ-
not associated with poorer function- ous studies suggest that such inter-
ing on the SAS overall adjustment their family but not ventions need to be developed with
subscale, which is a composite gener- caution. For example, the results of
al measure of social adaptation, fur- in interactions national surveys indicate that most
ther supports the hypothesis that stig- people prefer to maintain a social dis-
ma specifically compromises social with family tance from individuals who have a
functioning in nonfamily relation- mental illness (9); therefore, interven-
ships. members. tions that attempt to counter the social
Of particular interest are the re- withdrawal of people with mental ill-
sults of the factor analysis of the social ness may, paradoxically, expose them
leisure subscale, which identified to more experiences of rejection.
three different dimensions of social One possible strategy to alleviate
dysfunction in interactions with indi- the results is that family members this concern might be to link inter-
viduals outside the family. Regression compensate for their ill relatives’ so- ventions for stigma with existing mod-
analyses using these refined dimen- cial deficits in ways that people out- els for promoting the recovery of per-
sions may help to elucidate the spe- side the family do not. sons with mental illness, such as sup-
cific ways in which persons with con- Future research might address this ported employment (27). This strate-
cerns about stigma adapt their social alternative explanation by consider- gy would provide individuals with an
behavior to avoid exposure to rejec- ing the impact of the family’s attitudes opportunity to recognize and respond
tion or discrimination. These analyses and behavior toward the ill relative on to discrimination or rejection within
found that such concerns predicted his or her concerns or behavior in the supportive framework of the pro-
social dysfunction in two of the three coping with stigma as well as the po- gram, where they can rely on peers or
dimensions: psychological isolation tential contribution of family mem- professionals to support and guide
and behavioral avoidance. Psycholog- bers’ own concerns about stigma (5). them. Such a buffered exposure
ical isolation was associated with be- Sociocultural factors other than stig- might help inoculate them against the
ing single and having a higher symp- ma that may influence the social ad- adverse effects of future experiences
tom severity; behavioral avoidance justment of persons with mental ill- of discrimination or rejection.
PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12 1631
 Research on interventions also 2. Farina A, Felner RD: Employment inter- Manual for the expanded Brief Psychiatric
viewer reactions to former mental patients. Rating Scale (BPRS). Schizophrenia Bul-
must consider the individual cogni- Journal of Abnormal Psychology 82:268– letin 12:594–602, 1986
tive biases and coping styles that indi- 272, 1973
viduals employ. For example, studies 19. Weissman MM: The Depressed Woman: A
3. Olshansky S, Grob S, Malamud IT: Em- Study of Social Relationships. Chicago,
have demonstrated that some stigma- ployer’s attitudes and practices in hiring of University of Chicago Press, 1974
tized persons may preferentially at- ex-mental patients. Mental Hygiene 42:
tribute rejection and other negative 391–401, 1958 20. Lehman AF: A quality of life interview for
the chronically mentally ill. Evaluation and
outcomes of stigma to internal fac- 4. Farina A, Murray PJ, Groh T: Sex and Program Planning 11:51–62, 1988
tors—for example, ability—in order worker acceptance of a former mental pa-
tient. Journal of Consulting and Clinical 21. Rosenheck R, Evans D, Herz L, et al: How
to protect self-esteem or perceived Psychology 46:887–891, 1978 long to wait for a response to clozapine: a
social control (28). The results of the comparison of time course of response to
factor analyses in our study suggest 5. Struening EL, Perlick DA, Link BG, et al: clozapine and conventional antipsychotic
The extent to which caregivers believe most medication in refractory schizophrenia.
that persons with bipolar disorders people devalue consumers and their fami- Schizophrenia Bulletin 25:709–719, 1999
have more than one way of coping lies. Psychiatric Services 52:1633–1638,
with concerns about stigma in social 2001 22. Sherbourne CD, Hays RD, Wells KB: Per-
sonal and psychological risk factors for
situations outside the family. Because 6. Cohen G, Struening EL: Opinions about physical and mental health outcomes and
it is possible that different modes of mental illness in the personnel of two large course of depression among depressed pa-
coping with stigma are differentially mental hospitals. Journal of Abnormal and tients. Journal of Consulting and Clinical
Social Psychology 64:349–360, 1962 Psychology 63:345–355, 1995
associated with mental health and
well-being over time, empirical re- 7. Page S: Effects of the mental illness label in 23. Cronkite RC, Moos RH, Twohey J, et al:
attempts to obtain accommodations. Cana- Life circumstances and personal resources
search using a factor analytic ap- dian Journal of Behavioral Science 9:84–90, as predictors of the ten-year course of de-
proach might help identify the more 1977 pression. American Journal of Community
adaptive modes and thus inform re- Psychology 26:255–279, 1998
8. Farina A: Are women nicer people than
search on interventions. men? Sex and the stigma of mental disor- 24. George LK, Blazer DG, Hughes DC, et al:
ders. Clinical Psychology Review 1:223– Social support and the outcome of major
Conclusions 243, 1981 depression. British Journal of Psychiatry 154:
478–485, 1989
The Surgeon General’s report on 9. Link BG, Phelan JC, Bresnahan M, et al:
mental health (29) noted that stigma Public conceptions of mental illness: labels, 25. Dean A, Kolody B, Wood P: Effects of so-
“is among the many barriers that dis- causes, dangerousness, and social distance. cial support from various sources on de-
American Journal of Public Health 89: pression in elderly persons. Journal of
courage people from seeking treat- 1328–1333, 1999 Health and Social Behavior 31:148–161,
ment” for their condition. We found 1990
10. Link BG, Cullen F, Struening EL, et al: A
that concerns about stigma also serve modified labeling theory approach in the 26. Salovey P, Rothman AJ, Detweiler JB, et al:
as a barrier to meaningful social inter- area of the mental disorders: an empirical Emotional state and physical health. Amer-
actions with persons outside patients’ assessment. American Sociological Review ican Psychologist 55:110–121, 2000
54:400–423, 1989
own families. Thus our study high- 27. Drake RE, McHugo GJ, Bebout RR, et al:
lights the importance and potential 11. Link BG, Cullen F, Frank J, et al: The so- A randomized clinical trial of supported
benefits to consumers of reducing the cial rejection of ex-mental patients: under- employment for inner-city patients with se-
standing why labels matter. American Jour- vere mental disorders. Archives of General
stigma associated with mental illness. nal of Sociology 92:1461–1500, 1987 Psychiatry 56:627–633, 1999
To achieve this goal and promote op-
12. Goffman E: Stigma: Notes on Management 28. Crocker J, Major B, Steel C: Social stigma,
timal functioning for persons with of Spoiled Identity. Englewood Cliffs, NJ, in The Handbook of Social Psychology, 4th
mental illness, innovative research is Prentice-Hall, 1963 ed. Edited Gilbert D, Fiske S, Lindzey G.
needed to develop effective interven- New York, McGraw-Hill, 1998
13. Fadden G, Bebbington P, Kuipers L: Car-
tion strategies that do not unduly ex- ing and its burdens: a study of the spouses 29. Mental Health: A Report of the Surgeon
pose them to rejection and discrimi- of depressed patients. British Journal of General. Washington, DC, US Department
nation. At the same time, mental health Psychiatry 151:660–667, 1987 of Health and Human Services, US Public
practitioners, families, and consumer Health Service, 1999
14. Spitzer RL, Endicott J, Robins E: Research
advocates must join to promote more diagnostic criteria: rationale and reliability. 30. Hollingshead A, Redlich FC: Social Class
widespread efforts to educate the Archives of General Psychiatry 35:773–782, and Mental Illness: A Community Study.
1978 New York, Wiley, 1958
public about mental illness. ♦
15. Perlick D, Clarkin JF, Sirey J, et al: Burden
Acknowledgment experienced by caregivers of persons with
bipolar affective disorder. British Journal of
This study was supported by grant MH- Psychiatry 174:56–62, 1999
51348 from the National Institute of
16. Endicott J, Spitzer RL: A diagnostic inter-
Mental Health.
view: Schedule for Affective Disorders and
Schizophrenia. Archives of General Psychi-
References atry 35:837–844, 1978
1. Link BG, Phelan JC: Labeling and stigma, 17. Overall JE, Gorham DR: The Brief Psychi-
in The Handbook of the Sociology of Men- atric Rating Scale. Psychological Reports
tal Health. Edited by Aneshensel CS, Phe- 10:799–812, 1962
lan JC. New York, Kluwer Academic/ Plen-
um, 1999 18. Lukoff D, Nuechterlein KH, Ventura J:

1632 PSYCHIATRIC SERVICES ♦ December 2001 Vol. 52 No. 12