Content uploaded by Michael Canute Lambert
Author content
All content in this area was uploaded by Michael Canute Lambert on Mar 05, 2015
Content may be subject to copyright.
Self-Esteem of Adolescents Who Were Born Prematurely
Saroj Saigal, MD, FRCP(C)*; Michael Lambert, PhD#; Chad Russ, BS#; and Lorraine
Hoult, BA*
From the *Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada; and #Department of
Psychology, Michigan State University, East Lansing, Michigan.
PEDIATRICS Vol 109, n° 3 marzo 2002
ABSTRACT. Objective. To determine whether there are any diff erences between the self-
esteem of a cohort of adolescents who were extremely low birth weight (ELBW) in
comparison with term controls (cs); and to test the associations between self-esteem and
several predefined predictor variables.
Background. Self-esteem is considered to play a significant role in psychological
adjustment and scholastic success. Little information exists on how adolescents who were
ELBW regard themselves.
Design/Methods. Longitudinal follow-up of a regional cohort of 1321169 (78%) ELBW
survivors and 1271 145 (88%) sociodemographically matched Cs, born between 1977
and 1982. Measures: Harter Adolescent Self-Perception Profile (1988) with 9 dimensions,
including Global Self-Worth, socioeconomic status (Hollingshead), height-for-age and
weight-for-age z scores, and Wide Range Achievement Test-Revised (WRAT-R; Reading,
Spelling, and Arithmetic). Data Analysis: General linear model multiple analyses of
covariance were performed to determine whether significant relationships existed
between the 9 self-esteem dimensions and the independent variables of birth weight
status and gender, and the covariates of age, socioeconomic status, physical
development, and academic achievement.
Results. Global Self Worth was similar for ELBW and Cs (means: 3.1 and 3.2).
Multivariate effects revealed no interactions, but significant main effects emerged for
birth weight status, gender, weight-for-age z scores, age in months, and for all 3 WRAT-
R subtests, all effect sizes medium to large.
Follow-up, analysis of covariance revealed medium~ size gender effects for athletic
competence (means: 3.1 and 2.6), and physical appearance (means: 2.9 and 2.5),
where boys rated themselves significantly higher on both domains; and age effects,
where older teens rated themselves better for job competence. Significant but small
effect sizes emerged for the following: 1) weight-for-age z scores, where heavier youth
rated themselves higher on close friendships, 2) gender, where girls had higher ratings
for close friendships, 3) birth weight, where Cs rated themselves higher on athletic
competence, and 4) WRAT-R math effect, where children with higher math scores rated
themselves better on scholastic competence.
Conclusions. Overall, ELBW teenagers do not differ significantly from C teenagers on
most dimensions of self-esteem. Gender effects emerged on some Harter domains.
Pediatrics 2002;109:429-433; self-esteem, adolescence, self-concept, extremely low
birth weight, controls.
ABBREVIATIONS. VLBW, very low birth weight; GSW, global self-worth; ELBW, extremely
low birth weight; C, control; SPPA, Harter Self-Perception Profile; WRAT-R, Wide Range
Achievement Test-Revised; SES, socioeconomic status; MANCOVA, multiple analysis of
covariance; ANCOVA, analysis of covariance; SD, standard deviation; ES, effect size.
The primary focus of most studies on the outcome of children who were born
prematurely has been to report on neurodevelopmental morbidity, cognitive functioning,
and school difficulties.1-5 A few investigators recently have reported on the behavioral
and emotional adjustments in early to mid-childbood.6-8 It is apparent from these
studies that infants who were very low birth weight (VLBW) are at a significantly greater
risk for a wide array of neurodevelopmental, cognitive, behavioral, and emotional
difficulties than children born at term.1-8 These differences seem to persist even when
children with neurologic impairments are excluded from the study population. 5,7
In the last 2 decades, there has been increasing recognition of the profound role that
self-esteem plays on all aspects of children's development.9 It is generally agreed that
self-esteem is the evaluative component of the self-concept.10 Self-esteem has been
variously defined by researchers and clinicians as an "expression of approval or
disapproval, involving the extent to which a person believes himself or herself
competent, successful, significant and worthy~ "11' Self-esteem also refers to the
"discrepancy between a person's actual and ideal self,"10 and how adequately an
individual performs in domains in which that individual considers success to be
important.12 Thus, low self-esteem would ensue if there is a significant discrepancy
between one's expectations and perceived adequacy. 13 However, despite the extensive
literature, research on self-esteem is flawed because of ambiguous definitions of the
construct, lack of appropriate theoretical models, and inadequate instruments to measure
the same. 14,15
Several studies suggest that self-esteem plays a significant role in personal adjustments,
quality of peer relationships, motivation, as well as in athletic and scholastic success.9
Harter 16 has shown that among teenagers, physical appearance, particularly among
girls, correlates most high1y with global self-worth (GSW). This is followed closely by
peer social acceptance, whereas scholastic competence, athletic abilities, and conduct are
less influential. It is also often assumed that people with a physical disability have lower
self-esteem than the able-bodied. 17 Given that there is little information on how
adolescents who were extremely low birth weight (ELBW) regard themselves, we propose
to compare first the self-esteem of ELBW and term control (C) adolescents; second, to
determine whether there are any differences in self-esteem between adolescents with
and without neurosensory impairments; and finally, to test the association between self-
esteem and several predefined predictor variables. We hypothesized that there would be
no differences in self-esteem. dimensions between ELBW and C teenagers, nor between
ELBW teenagers with and without neu~ rosensory impairments.
METHODS
Respondents ELBW Cohort
The ELBW survivors, 501 to 1000 g birth weight, were born between 1977 and 1982 Io
residents of a geographically defined region in central-west Ontario and followed
longitudinally from birth.4 At the time of the current assessment, the children ranged
between 12 and 16 years of unadjusted age.
Controls (Cs)
Cs were recruited at 8 years old from a random list of children obtained through the
Directors of the Hamilton Public and Roman Catholic Separate School Boards and
matched for sex, age, and social class to each index child (1977-1981 births).4 These
children were also followed longitudinally and reassessed at the same age as the ELBW
adolescents.
Measures
We used the Harter Self-Perception Profile (SPPA) for adolescents. 111 This scale is
designed to measure teenagers' perception of their competence in, and the importance to
them in the following 8 domains: scholastic competence, social acceptance, athletic
competence, physical appearance, job competence, romantic appeal, behavioral conduct,
and close friendship. These domains were discriminable through factor analytic
procedures. In addition, the Harter SPPA provides a GSW score. This is a number
between 1 and 4 and is the mean score given to 5 items that tap the extent Io which the
adolescent likes himself/herself as a person and is happy with the way he/she is. Thus, it
constitutes a global judgment of one's worth as a person, rather than domain-specific
competency or adequacy.
The SPPA is comprised of a total of 45 items and is recommended for use for children in
the 9th through the 12th grades. The reponse format includes both positively and
negatively worded phrases, designed to eliminate the "pull" for socially desirable
responses: for example, "some kids have a lot of friends, but others don't have many
friends." Harter also introduced a 2-step response format whereby the adolescents must
first choose the direction and then the intensity of their response. This instrument is used
widely and has been found to have sound psychometric properties.19 The internal
consistency ranged from 0.74 to 0.92 (median: 0.81); 6 of the 9 subscales had good
internal consistency.
The questionnaire was self-completed by the teenagers in a private room at McMaster
University after they completed direct interviews for a study on their health-related
quality of life.20 The teenagers also participated in standardized psychometric
assessments,5 including the Wide Range Achievement Test-Revised (WRAT-R).21
Physical growth measures, height, weight, and head circumference were obtained by
standard techniques. Hollingshead 4-factor index (AB Hollingshead, unpublished data,
1975) was used to describe parental socioeconomic status (SES).
Informed Consent
The study was approved by the Ethics Committee of Hamilton Health Sciences
Corporation, and written informed consent was obtained from the parents of all
participants.
Statistical Analyses
We performed a general linear model multiple analysis of covariance (MANCOVA) to
reduce the probability of Type 1 error (which could have resulted if separate analysis of
covariance [ANCOVA] analysis was performed on each self-esteem variable). The
independent variables were gender and birth weight status (ELBW vs Cs); the dependent
variables were each of the 8 Harter Self-Esteem dimensions and GSW. The following
variables were included as covariates in the model, based on a review of the literature
that indicates their association with self-esteem: age, physical development, academic
achievement, and social class.7,16 Thus, besides the age of the children (calculated in
months), standardized weight-for-age and height-for-age z scores were considered as
proxies for physical development, and the WRAT-R21 reading, spelling, and math scores
were entered separately in the model as proxies for academic achievement. In all, there
were 7 continuous independent variables (SES, chronological age, weight-for-age z
scores, height-for-age z scores, WRAT-R reading. spelling, and math scores), 2
categorical independent variables (gender and birth weight status), and the 9 Harter
indices, including GSW, as multivariate dependent variables.
RESULTS
Study Respondents
ELBW Adolescents
Between 1977 and 1982 inclusive, 179 ELBW infants survived to hospital discharge (48%
survival rate), and 10 children subsequently died leaving 169 available survivors. Of
these, 13 children were lost; 5 lived too far away, 6 refused to participate, and 9
severely impaired children were untestable, leaving 141 available survivors. One hundred
thirty-two children (78%) completed the study questionnaire.
Control Adolescents
A total of 145 Cs were recruited at 8 years of age. Of these, 10 children were lost, 9
refused, and 2 lived too far away. Of the remaining children, 117 (81%) completed the
assessment.
Demographic Data
Table 1 shows the demographic data on both ELBW and C adolescents. The mean birth
weight of the ELBW infants was 839 (standard deviation [SD]: 125) g and the mean
gestational age was 27 (SD: 2) weeks. There were no significant differences in the SES
between the 2 cohorts. Neurosensory impairments were present in 24% of the ELBW
participants. One C child had mild cerebral palsy. The mean age at assessment was 14.1
(SD: 1.6) years for the ELBW cohort and 14.5 (SD: 1.3) years for Cs.
Comparisons of Adolescents With and Without Impairments
Before conducting the MANCOVA analyses, we compared 31 ELBW teenagers with and
101 teenagers without impairments. Our tests revealed no differences in the mean scores
between the groups on any of the 8 Harter domains or in the GSW scores, nor was there
any interaction between gender and neurosensory impairments.
MANCOVA Effects
All significant MANCOVA and follow-up ANCOVA effects were interpreted using Cohen's
IS22 criteria for effect sizes (ES), where effects are deemed small, medium, and large if
they account for 1% to 5,8%, 5,9%,to 13,8% of the variance, respectively. The 2 (birth
wieght status) X 2 (gender) MANCOVA revealed significant multivariate main effects for
birth weight status. No Interaction emerged, indicating that birth weight status, gender,
weight-for-age z score, age in month, and the WRAT-R reading, spelling, and math
scores are associated independently with the multivariate measure of self-esteem.
Univariate Effects
The 2 categorical independent variables, birth weight status and gender, and 5 of 7
continuous independent variables (weight-for-age z score, age in months, and the 3
WRAT~R subtest scores), were used as covariates in an ANCOVA, to test the effects on
each of the 9 dependent variables considered separately. The Bonferroni procedure, 23
which set a at .005, was used to reduce the chance of Type 1 error.
Birth Weight
The analyses revealed significant birth weight status effects on athletic competence F
(1226) = 13.78, P = .0001, ES = 5.7%. Table 2 indicates that C children rated
themselves as having higher athletic competence than their ELBW peers.
Gender
Gender effects emerged for the following domains: athletic competence, F(1226) =
16.73, P = .0001, ES = 6.9%; physical appearance, F(1226) = 15.56, P = .0001, ES =
6.4%; and close friendships, F(1226) = 11.62, P = .001, ES = 4.9%. The means listed in
Table 2 show that boys rated themselves significantly higher than girls on athletic
competence and physical appearance. The converse was true for close friendships.
Weight for Age
Weight-for-age z score effects emerged for close friendships only (F(1226) = 10.87, P
= .001, ES = 4.6%) indicating lower close friendship scores for children with higher
weight-for-age z scores.
Achievement
There was a WRAT-R math effect for scholastic competence (F(1226) = 8.13, P = .005,
ES = 3.5%) indicating that children with higher math achievement scores rated
themselves higher on scholastic competence. No significant achievement effects emerged
for reading or spelling.
Age
An age effect emerged for job competence (F(1226) = 20.23, P < .001, ES = 8.2%)
revealing that older children rated themselves higher in job competence than their
younger peers, but no age effects emerged for the other self-esteem domains.
DISCUSSION
Adolescence is an interesting period of transition that poses unique challenges for the
individual and the family. During this period, important decisions have to be made
regarding future education, work, and independent living, which are like1y to have a life-
Long impact. The literature suggests that aspirations and subsequent achievement may
be related to a person's level of self-esteem.17 Although a considerable body of research
is available on the neurodevelopmental and scholastic sequelae of extreme prematurity,'-
' the question of whether these children differ from Cs with respect to their self-concept
is a relatively neglected area.
Recent attempts to address this issue have resulted in studies that are not strictly
comparable. The key deterrents have been the lack of a theoretical framework and
differing operational definition of self-esteem .14,1' Other limitations include
inconsistency in the measures used, self-report or parent report, wide age range of the
respondents, inclusion and/or exclusion of children with disabilities, and variations in the
racial and sociodemographic composition of the study participants. It is therefore not
surprising that there is conflicting information on whether children who were premature
have lower levels of self-concept and whether those who have disabilities are doubly
compromised.
We have shown that at adolescence there were no differences between our ELBW and
sociodemographically matched term Cs on most perceived self-esteem domains,
including GSW, nor in the importance of such domains. The only differences which
emerged were in athletic competence where ELBW adolescents rated themselves
significantly lower than Cs. This is consistent with objective measurements of poorer
motor performance among VLBW children, which persist even when children with
neurosensory impairments are excluded .24 We found no differences by gender, except
for the physical appearance domain, where females rated themselves significantly lower
than males. This is also consistent with other published reports where adolescent girls
are reported to have lower self-esteem. 15,16 Overall, birth weight group, gender, and
neurosensory impairments did not contribute significantly to most Harter SPPA domain
scores.
Only 1 other study has addressed the issue of self-concept of 13- to 14-year-old
adolescents who were VLBW, using multiple respondents.7 Premature children had lower
scores on the Tennessee Self-Concept Scale and were reported by their parents and
teachers to manifest more behavioral disturbances than Cs, both at home and at school.
These findings persisted even when children with neurologic impairments were excluded.
In a study of all admissions to the neonatal intensive care unit, Speechley et a125 have
shown that the admission itself had a greater impact on adolescent girls than boys.
Admission was associated with lower social competence and self-esteem than was
admission to the normal nursery.
McCormick et al 26 obtained parental ratings of the child's competence on the Harter
Scale at age 8 to 10 years, on a multisite sample of premature children (<1000 g, 1001-
1500 g, and 1501-2500 g) and Cs. Highly significant differences were found between
birth weight groups, with the lower birth weight children being rated as less competent.
These differences were seen for all subscales of the Harter Scale, except for physical
appearance. However, the overall scores for both cohorts were higher than expected,
based on the published reference samples, so that ELBW children were considered by
their parents to be similar to the general population. The limitations of this study include
the biases inherent in a multisite sample, a very high attrition rate with disproportionate
losses of the disavantaged group, and lack of direct responses from children. It is also
not clear whether children with neurosensory impairments were excluded. Taylor et a127
found no group differences in self-perceived competence, parental ratings of
competence, or teacher ratings of social skills for infants <750 g and 750 to 1499 g,
when assessed at a mean age of 11 years. Zelkowitz et a1211 reported no differences in
self-concept between nonimpaired VLBW children and Cs at 6 years old. However, at 9
years old, the VLBW cohort perceived themselves as less scholastically competent than
Cs. Furthermore, although social status was not related to self-concept, birth weight and
lower academic achievement were associated with more negative self-perception in the
area of scholastic competence. In another study of a more mature cohort of
underprivileged prematurely born infants, social class was an overwhelming factor that
accounted for the variance in the self-perception of cognitive competence at late
adolescence.29
The literature on the self-esteem of adolescents with physical disabilities is equally
conflicting. In a commentary on the self-esteem of children with physical disabilities,
Llewellyn 30 reports that 3 studies found low self-esteem in young people, whereas 7
other studies found no significant differences compared with C participants. This
inconsistent picture was felt to be attributable to methodological problems in carrying out
disability research. For example, variables such as age of onset of the disability, severity,
and the visual appearance of an impairment, selection criteria of the inception cohort,
sample size problems, and whether the children attended a special school or were in
main stream education, could have all played a role in the self-perception of children.
Nevertheless, the majority of studies suggest that adolescents with physical disabilities
report good self-esteem 31-35 strong family relationships, and as many close friends as
adolescents in the national sample .31 However, adolescents with physical disabilities
participated in fewer social activities and had less intimate relationships with their
friends .31 Crocker and Major 36 speculate that members of a stigmatized group might
attempt to protect their self-image by comparing their outcomes with other individuals
who are more severely disabled, selectively devaluate dimensions in which they fare
poorly, or attribute negative feedback by others to prejudice. Whether there is a certain
lack of realism or overcompensation to achieve among individuals with disabilities can
only be resolved through qualitative studies.
We believe that the strengths of our study are that the participants are population-based,
and therefore the findings are generalizable to populations with similar
sociodemographics; the measurement tool, ie, the Harter,18 is widely accepted; and, the
attrition rate was relatively low. We should, however, point out that 9 severely impaired
ELBW teenagers were unable to participate in the self-assessments, and this might have
contributed to some of the lack of differences observed between the impaired and
nonimpaired teenagers. We also found no differences in self-esteem between ELBW and
C children at 8 years old~6 but unfortunately, because a different measure (Piers-Harris)
was used, the data cannot be considered strictly longitudinal. Our findings of no
differences in self-esteem between the 2 groups at adolescence are consistent with the
overall high self-reported health-related quality of life scores, which were provided
concurrently by the same cohort of ELBW teenagers and reported in our previous
publications.20,37
The self-esteem and social and emotional wellbeing of children are important domains
that should be monitored closely in VLBW children who are considered to be at high risk
for problems with adjustment. There is now a consensus that all aspects of a child's daily
activities, motivation, and behavior are impacted by the child's self-esteem.9 It appears
that the self-perception profiles change over time, and adolescence is a particularly
vulnerable period.16 Chronically low self-esteem may contribute to anxiety, depression,
lack of motivation, and poorer achievement in both occupational goals and interpersonal
relationships at adulthood.14 Additional longitudinal studies on self-esteem and
emotional adjustments of infants who are premature are encouraged to establish a
trajectory for children's long-term mental well-being, to elucidate antecedents of
maladjustments, and to plan for timely intervention strategies to foster better self-
esteem.9
REFERENCES
L Vohr BR, Wright LL, Dusick AM, et al. Neurodevelopmental and funetional outcomes of
extremely low birth weight infants in the National Iristitute of Child Health and Human
Development Neonatal Research Network, 1993-1994. Pediatrics. 2000;105:1216-1226
2. Woocì NS, Marlow N, Costeloe K, Gibson AT, Wilkinson AR for the EPICure Study Group.
Nemologic and developmental disability after extremely preterm birth. N Engl J Med.
2000;343:378-384
3. Hack M, Taylor HG, Klein N, Eiben R, Schatsclmeider C, MercuriMinich N. School-age
outeomes in children with birth weights under 750 g. N Engl 1 Med. 1994;331:753-759
4. Saigal S, Szatmari P, Rosenbaum P, Campbell D, King S. Cognitive abilities and school
performance of extremely low birth weight children and matched term control children at
age 8 years: a regional study. 1 Pediatr. 1991;118:751-760
5. Saigal S, Hoult LA, Streiner DL, Stoskopf BL, Rosenbaum PL. School cifficulties at
adolescence in a regional cohort of children who were extremely low birth weight. Pediatrics.
2000;105:325-331
6. Szatrnari P, Saigal S, Rosenbauro P, et al. Psychopathology and adaptive funetioning
among extremely low birthweight children at eight years of age. Dev Psychopath. 1993;5:345-
357
7. Levy-Schiff, Einat C, Har-Even D, et al. Emotional and behavioral adjustment in children
born prematurely. 1 Clin Child Psychol. 1994;23: 323-333
8. Botting N, Powls A, Cooke RWL et al. Attention deficit hyperactivity
disorders and other psychiatric outeomes in very low birthweight children at 12 years. 1 Child
Psychol Psychiatry. 1997;38:931-941
9. Brooks RB. Self-esteem during the school years. Its normal development and hazardous
decline. Pediatr Cffli North Am. 1992;39:537-549
10. Pope AW, McHale SM, Craighead WE. Self-Esteem Enhancement With Children and
Adolescents. Oxford, England: Pergamon; 1988
11. Coopersmith S. Self-Esteenì Inventories. Palo Alto, CA: Consulting Psychologists Press; 1981
12. James W. Psychology: The Briefer Course. New York, NY: Holt, Rinehart, & Winston; 1982
13. Rosenberg M. Conceiving the Self. New York, NY: Basie Books; 1979
14. Wylie R. The self-concept. In: A Review of&lethodological Considerations aìid Measuriiig
Instruments. Vol. I. Lincoln, NE: University of Nebraska Press; 1974
15. Harter S. Self and identity development. In: Feldman SS, Elliot C, eds. At the Threshold; The
Developing Adolescent. Cambridge, MA: Harvard University Press; 1990
16. Harter S. Causes, correlates, and the functional role of global self-worth: a life-span
perspective. In: Kolligian J, Sternberg R, eds. Perceptions of Competence aud Incompetence
Across the Ufe-Spaìi. New Haven, CT: Yale University Press; 1989
17. Arnold P, Chaproan M. Self-esteem, aspirations and expectations of adolescents with
physical disability. Dev Med Child Neurol. 1992;34: 97-102
18. Harter S. Self-Perception Profilefor Adolescents. Denver, CO: University of Denver Press;
1988
19. Keith LK, Bracken BA. Self-concept instrumentation: a historical and evaluative review. In:
Bracken BA, ed.'Handbook of Self-Concept. Developmental, Social, and Cliffical
Considerations. New York, NY: John Wiley & Sons, Inc; 1996
20. Saigal S, Feeny D, Rosenbaum P, Furlong W, Burrows E, Stoskopf B. Self~perceived health
status and health-related quality of life of extremely low-birth-weight infants at adolescence.
JAMA. 1996;276: 453-459
21. Jastak S, Wilkinson C. Wide Range Achievertient Test-Revised (WRAT-R). Wilmington, DE:
jastak Associates; 1984
22. Cohen j. Statistica] Power AìiWysisfor the Behavioral Sciences. 2nd ed. New York, NY:
Academic Press; 1988
23. Cliff N. Analysing Multivariate Data. New York, NY: Harcourt Brace Iovanovich, 1987
24. Powls A, Botting N, Cooke RWI, et al. Motor impairment in children 12 to 13 years old with a
birthweight of less than 1250 g. Arch Dis Child. 1995;73:F62-F66
25. Speechley KN, Avison WR. Admission to a neonatal intensive care unit as a predictor of
long-term health: a 12-year follow-up. J Dev Beliav Pediatr. 1995;16:397-405
26. McCormick MC, Workman-Daniels K, Brooks-Gunn j. The behavioral and emotional well-
being of school age children with different birth weights. Pediatrics. 1996;97:18-25
27. Taylor HG, Klein N, Minich NM, Hack M. Middle school age outcomes in children with very
low birthweight. Child Dev. 2000;7L1495-1511
28. Zelkowitz P, Papageorgiou A, Zelazo PR, Weiss MJS. Behavioral adjustment in very low arid
normal birth weight children. 1 Clin Child Psychol. 1995;24:21-30
29. Cohen S. Biosocial factors in early infancy as predictors of competence in adolescents
who were born prernaturely. 1 Dev Behav Pediatr. 1995; 16:36-41
30. Llewellyn A. Self-esteern in children with physical disabilities Iletterj. Dev Med Child Neurol.
2001;43:70-71
31. Stevens SE, Steele CA, lutai JW, KaInins IV, Bortolussi JA, Biggar WD. Adolescents with
physical disabilities: some psychosocial aspects of health. 1 Adolesc Health. 1996;19:157-164
32. Magill J, Hurlbut N. The self-esteem of adolescents with cerebral palsy. Am 1 Occup Ther.
1986;49:402-407
33. King CA, Shultz IZ, Steel K, Cilpin K, Cathers T. Self-evaluation and self-concept of
adolescents with physical disabilities. Am [ Occup Ther. 1993;47:132-140
34. Appleton PL, Minchom P, Ellis NE, Elliott CE, Boll V, jones P. The self-concept of young
people with spina bifida: a population-based study. Dev Med Child Neurol. 1994;36:198-215
35. Ostring H, Nieminen S. Concept of self and the attitude of school age CP children towards
their handicap. Int 1 Rehabil Res. 1982;5:235-237
36. Crocker j, Major B. Social stigma and self-esteern: the self-protective properties of stigma.
Psychol Rev. 1989,96:608-630
37. Saigal S, Stoskopf BL, Feeny D, et al. Differences in preferences for neonatal outcomes
among health care professionals, parents, and adolescents. IAMA. 1999;281:1991-1997
