Adjustment of growth hormone deficient children

MISSING IMAGE

Material Information

Title:
Adjustment of growth hormone deficient children parent, teacher, peer, and self-perceptions
Physical Description:
vi, 93 leaves : ill. ; 29 cm.
Language:
English
Creator:
Lewis, Carol, 1959-
Publication Date:

Subjects

Subjects / Keywords:
Growth Disorders -- Child   ( mesh )
Growth Disorders -- Infant   ( mesh )
Self Concept -- Child   ( mesh )
Self Concept -- Infant   ( mesh )
Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- Clinical and Health Psychology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1986.
Bibliography:
Bibliography: leaves 88-92.
Statement of Responsibility:
by Carol Lewis
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000897815
oclc - 17159932
notis - AEK6498
sobekcm - AA00004866_00001
System ID:
AA00004866:00001

Full Text






THE ADJUSTMENT OF GROWTH HORMONE
DEFICIENT CHILDREN:
PARENT, TEACHER, PEER, AND
SELF-PERCEPTIONS







By

CAROL LEWIS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1986













ACKNOWLEDGMENTS


I would like to thank Suzanne Bennett Johnson, Ph.D., for her support and

encouragement throughout this project as well as during all of graduate school. Many thanks

are also extended to Carole Knuth, P.A.C., Janet Silverstein, M.D., and Arlan Rosenbloom,

M.D., for making this project possible and providing valuable assistance throughout. Lastly,

deepest appreciation is offered to my husband, Timothy J. Moses, for his patience during all

phases of this project.











TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...................... ii

ABSTRACT .......................... v

CHAPTERS

I INTRODUCTION .................. 1

Problem . . 1
Short Stature .................... 1
Psychosocial Problems Associated with Short
Stature....................... 3
Direct Psychological Effects of Human Growth
Hormone .. .. .. .. .. .. 4
Intellectual Functioning. . 4
Psychological Correlates: Growth Hormone
Deficient Adults. . . 6
Psychological Correlates: Growth Hormone
Deficient Children. . . 8
Self-Concept ..................... 10
Treatment with Growth Hormone: Psychological
Correlates . . 11
Behavioral Adjustment. . 14
Summary ... ... .. .... .. ... .. .. 17
Present Study ..................... 18



II METHOD ................... .. 20

Subjects... ................ .. .. 20
Measures...... .............. .. 20
Procedure ................... 26

III RESULTS ................... .. 27

Characteristics of the Sample. . ... .27
Height of the Growth Hormone Deficient
Children . . .. 29



III










Comparison of the Control Subjects from the
Growth Hormone Deficient Subjects'
Classrooms and Those Who Were Not. ...... .. 29
Comparison of the Growth Hormone Deficient
Subjects with Idiopathic Growth Hormone
Deficiency and Those with All Other Types of
Growth Hormone Deficiency. . ... 33
Covariates: Miscellaneous Statistics. ... 33
Ratings by Each Source. . ... 37
Correspondence Between Ratings of Growth
Hormone Deficient Children by Different
Sources.. ....................... 58
Correspondence Between Ratings of Control
Children and Total Sample by Different
Sources.. ................ ..... .. 71
Relationships Between Measures of Adjustment
and Current Perceptions of Height in the
Growth Hormone Deficient Children. ... 75

IV DISCUSSION.. ................ .... .. 81

BIBLIOGRAPHY.. ......................... 88

BIOGRAPHICAL SKETCH ....................... 93











Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


THE ADJUSTMENT OF GROWTH HORMONE DEFICIENT CHILDREN:
PARENT, TEACHER, PEER, AND SELF-PERCEPTIONS

By

Carol Lewis

December 1986

Chairman: Suzanne Bennett Johnson, Ph.D.
Major Department: Clinical Psychology


The general picture of the growth hormone deficient child that emerges from the

literature to date is one of an immature, socially-withdrawn child with a low self-concept

who has significant school problems despite at least average intelligence.

The present investigation looked at the adjustment of 45 growth hormone deficient

children being treated with human growth hormone from the perspectives of a number of

different sources. Parents, teacher, and the growth hormone deficient children themselves

completed measures concerning the behavioral adjustment and self-esteem of these children

and a group of normal height control children. Peers completed sociometric ratings which

provided data about the two groups of children.

The growth hormone deficient children rated themselves as having lower self-esteem

than did the normal height children. Mothers and peers rated them as more withdrawn than

the normal height children. Both mothers and fathers of the growth hormone deficient

children rated their children as significantly less socially competent than did the parents of











normal height children. Compared to normal height children, teachers rated the growth

hormone deficient children as having worse grades, putting forth less effort, learning less,

and being less happy. While the growth hormone deficient children were not disliked by their

peers relative to normal height children, neither were they popular.

Looking at the correspondence between the ratings by different sources, mother and

father ratings were highly related to each other. Peer and child self-ratings also showed a

pattern of significant relationships. Parental ratings of social competence were significantly

related to peer ratings of social withdrawal and being a relatively disliked classmate.

Lastly, growth hormone deficient children who overestimated their height relative to

peers also endorsed extremely elevated levels of self-esteem.

Overall, these results indicate that growth hormone deficient children have lower

self-esteem than normal height children, they are seen by a number of sources as less socially

competent than peers, and are rated by peers as socially withdrawn. Growth hormone

deficient children who overestimate their height also appear to have inflated self-esteem.









CHAPTER I
INTRODUCTION


Problem

The focus of the present investigation is the behavioral adjustment of growth hormone

deficient children. At the time of this study all of the growth hormone deficient children who

participated were receiving human growth hormone therapy as treatment for their disorder.

This treatment involved taking injections of growth hormone purified from the pituitary

glands of human cadavers. This form of treatment had been available for approximately 25

years. Prior to the availability of growth hormone replacement therapy, these children

remained about the size of an early grade school child and were called "midgets". Even with

treatment, however, growth rates were variable with most growth hormone deficient children

getting taller yet continuing to be noticeably short for their age and sex (Soyka, Bode,

Crawford, & Flynn, 1970; Shizume, 1984; Schaff-Blass, Burstein, & Rosenfield, 1984). In

1985 human growth hormone was taken off the market due to contamination of some lots with

a virus that later causes Creutzfeldt-Jakob disease. Later that year biosynthetic growth

hormone became available.

The current study compared perceptions of the behavioral adjustment and self-esteem of

a group of growth hormone deficient children being treated with human growth hormone and a

group of normal height control children. Mothers, fathers, teachers, peers, and the children

themselves participated in the study.


Short Stature

There are a number of reasons for short stature in childhood. Some are fairly innocuous

and merely represent variations of normal linear growth. For example, some children are the

offspring of relatively short parents and therefore may be short for familial or genetic







2

reasons. Other children may grow at a rate much slower than their peers but eventually will

catch up with them. Both of these are examples of variations in normal growth patterns.

Short stature may also result from several pathological conditions. Psychosocial

deprivation may cause delayed growth. Upon removal from the offending environment, these

children may attain relatively normal height. Physical problems also may result in short

stature. These include Turners' syndrome (a genetic disorder), achondroplasia (a bone and

cartilage disorder), craniopharnygioma (a tumor), and pituitary growth hormone deficiency.

Reports of the prevalence of growth hormone deficiency have varied greatly from 1 in

30,000 (Parkin, 1974) to 1 in 4,000 (Vimpani et al., 1981). The later study accounted for

the discrepancy between their findings and that of other investigations by suggesting that many

cases are missed due to professional inactivity and not because of lack of parental concern. The

authors concluded that growth hormone deficiency in a non-disadvantaged community may

account for 5% of total cases of short stature or 10% of all short but otherwise normal

children.

The criterion for short stature varies somewhat in different investigations. One

criterion commonly found is height below the third percentile for age and sex (see

Schaff-Blass et al., 1984). A diagnosis of growth hormone deficiency usually can be made

only after at least two diagnostic tests show a deficiency. These tests consist of evaluation of

serum growth hormone after either insulin-induced hypoglycemia, arginine infusion, L-dopa

stimulation, or glucagon administration (see Ad Hoc Committee on Growth Hormone Usage,

1983). The growth hormone deficient child may grow less than 2.5 + 0.8 cm. per year

(Kaplan, 1975) while the normal child generally grows at least 5 1.5 cm. per year

(Daughaday, 1974).









Psychosocial Problems Associated with Short Stature

Short children encounter a series of psychosocial problems as they age (see Drash,

1969). The major issue from birth to 4 years of age is the identification of growth hormone

deficiency. The adjustment of the child during this time period is in part determined by his

parents' personalities. Money and Pollitt (1966) introduced the concept of "complementarity

of pathology" to account for the observation that a child's adjustment to his growth problem is

partly determined by the response of his parents to the problem. For example, he may be

encouraged to act according to his size, not his age, a pattern due at least partially to parental

needs and conflicts.

During the ages of 5 and 6 the child may be introduced to a school setting for the first

time with all of the adjustments that entails. This may be a particularly difficult separation

for a child who has been encouraged to be relatively more dependent than his peers. Teasing,

ridicule and physical bullying by peers is often first encountered to any large extent in grades

2-4 (ages 7-9). It is during these years that the child may first become acutely aware of his

"difference." The child may react to teasing by passive resignation and withdrawal, mascotism

and laughing at himself, physical aggression, befriending a larger "protector," or using

humorous replies.

The adolescent years are filled with many potential "crises" for the short statured child.

His younger siblings may be passing him in height, he may become especially aware of his

condition as same sex competition increases in a number of areas including heterosexual

relations. If there are other associated pituitary deficiencies the child may also not develop

secondary sexual characteristics without appropriate hormone replacement. The common

personality "mechanisms" exhibited by short statured adolescents according to Drash (1969)

are immaturity and developmental retardation (not uncommonly at least two years delay),

denial, withdrawal, exaggeration and overcompensation, mascotism, and the use of humor. The







4

most common parental problems are treating the child according to his size rather than his

age, denial, and failure to look for psychological problems in the child.


Direct Psychological Effects of
Human Growth Hormone

Growth hormone is a pituitary hormone and is largely regulated by the central nervous

system. Growth hormone levels may vary according to the amount of stress one is

experiencing. Its usual pattern of responding may also be altered in the presence of

endogenous depression, possibly related to an underlying neurotransmitter problem.

However, reduction or elimination of growth hormone due to defective pituitary function does

not appear to produce direct psychological effects (Brown, Seggie, Chambers, & Ettigi,

1978).


Intellectual Functioning

Much of the earliest research on growth hormone deficient children of a psychological

nature asked whether intellectual retardation accompanied the physical growth retardation.

Pollitt and Money (1964) gave age-appropriate standardized intelligence tests (WISC, Binet,

or Gesell Development Schedules) to 15 growth hormone deficient children aged 3 to 15 years.

The mean full scale IQ was 103 for the WISC and Binet combined. These scores are equivalent

to the published normative means of 100. There was also no significant verbal-performance

IQ difference. School performance, however, was average or below average for all subjects

with teacher comments suggesting lack of interest and poor study habits. Pollitt and Money

concluded that for their sample the search for approval from peers and teachers seems to

interfere with school work.

Further work by this same group on intelligence (Money, Drash, & Lewis, 1967) found

a normal distribuiton of intelligence in their sample of 36 subjects aged 5 to 36 with growth







5

hormone deficiency. Drash, Greenberg and Money (1968) compared their previously

obtained IQ results for growth hormone deficient subjects with that from samples of people

with short stature due to three other causes. Subjects with short stature due to depreviation

syndrome had an IQ distribution indicating intellectual impairment. Turner's syndrome

subjects showed generally low performance IQ's with normal verbal and full scale IQ's. The

sample of subjects with achrondroplasia was small but seemed to show normal intelligence. At

least three other studies (Rosenbloom, Smith, & Loeb, 1966; Cacciaguerra, 1978;

Meyer-Bahlburg, Feinman, MacGillvray, & Aceto, 1979) have found normal intelligence in

their samples of growth hormone deficient subjects. The subjects in the first of these studies

also evidenced satisfactory academic achievement.

At least two studies have found impaired intelligence in samples composed of growth

hormone deficient subjects (Obuchowski et al., 1970; Frankel & Laron, 1968). Frankel and

Laron (1968) also found marked verbal-performance IQ discrepancies and marked subtest

scatter. Additionally performance on the Bender Visual-Motor Gestalt Test was generally

deficient. Visual-motor difficulties were also found by Abbott and her colleagues (Abbott,

Rotnem, Genel, & Cohen, 1982). The 11 children studied in their sample had a mean full

scale IQ (86.9) in the low average range. This was accounted for by the lower socioeconomic

status of this sample compared to the normative sample of the WISC-R (Wechsler, 1974).

More recently, Siegel and Hopwood (1986) studied the intellectual functioning and

academic achievement of 42 children with idiopathic growth hormone deficiency. Children

were given the Wechsler Intelligence Scale for Children-Revised (WISC-R), the Bender

Gestalt Test of Visual-Motor Integration, two subtests of the Wide Range Achievement Test

(Reading Sight Word Recognition and Math), the Reading Comprehension subtest of the Peabody

Individual Achievement Test (PIAT), and the Piers-Harris Self-Concept Inventory. Children

were categorized as low academic achievers if either their combined reading standard score,







6

math standard score, or both were <85. Twenty-two of the 42 children were low achievers.

The low achievers were then categorized into one of three psychometric profiles explaining

academic failure:

1. Cognitive Deficit theory--at least one WISC-R scale score (VIQ or
PIQ) falls within the average range (90-110); a V-P difference
>18 points and/or a significant visual-motor integration deficit.

2. Low Ability theory--both VIQ and PIQ scale scores fall below the
average range (<90).

3. Cognitive Underfunctioning--Low Self-Concept theory both VIQ
and PIQ scale scores fall within the average range; there is neither
a significant V-P difference nor a significant visual-motor deficit.

The mean Verbal IQ of the growth hormone deficient group was 93.9. The mean

Performance IQ was 94. Twenty-nine percent of the sample had a VIQ-PIQ discrepancy of >18

points in comparison to 16% of the normative sample (Kaufman, 1979). The mean

self-concept score was higher than the mean of the standardization sample (X=60.2, %=75).

Neither age at onset nor duration of growth hormone replacement treatment was significantly

related to self-concept scores. Seventy-four percent of the low achievers had at least one

cognitive atypicality (41%=low ability; 32%=cognitive deficit). Twenty-six percent of the

children had significant visual-motor deficits. The authors conclude that while growth

hormone deficient children have average cognitive functioning overall, they show specific

problems: significant cognitive variability (high incidence of large VIQ-PIQ differences) and

visual-motor integration difficulties.


Psychological Correlates:
Growth Hormone Deficient Adults

Personality characteristics of growth hormone deficient children and adults have been

studied. While the earlier work seems to be based on impressions, case studies, interview

material, or projective testing, it is nonetheless a very interesting background from which to







7

begin a more methodologically sound investigation. For example, in a sample of growth

hormone deficient adults Obuchowski and his colleagues (1970) found behaviors indicating

childishness, jocularity and carelessness. On more in-depth analysis, the authors noted

depression which they felt had been over-shadowed by well-developed defense mechanisms

such as hypercompensation.

Brust, Ford, and Rimoin (1976) looked at the adjustment of 16 short statured

adults--5 with growth hormone deficiency and 11 with achrondoplasia. Using interviews and

some personality testing they found no significant differences between the two groups. Their

subjects had generally achieved a satisfactory life adjustment despite their physical

conditions. They seemed to have secure identities as "little people" and successfully used

coping mechanisms such as a sense of humor or pleasant interpersonal style. Males tended to

experience more emotional distress than females. The authors concluded that these short

statured adults were psychologically well-adjusted and confident and generally lacked

psychiatric symptoms, excessive anxiety, and depression.

More recently, Mitchell and colleagues (Mitchell, Johanson, Joyce, Libber, Plotnick,

Migeon, & Blizzard, 1986) assessed the self-esteem and social, educational, and vocational

status of 58 growth hormone deficient adults (ages 16 to 46 years) who had previously been

treated with human growth hormone. The average was 26 years and the average final height

was 5'2". Most of the subjects reported average or above average academic performance and

satisfaction with employment status. In terms of self-concept, the growth hormone deficient

adults rated themselves as lower than norms on physical self and self-criticism, but higher on

self-satisfaction, personal worth, and sociality. Both heterosexual and same sex relationships

were rated as inadequate. Forty-three percent were married and 59% of the married subjects

had children.







8

A follow-up assessment of 116 growth hormone deficient adults who had been treated

with human growth hormone was also done recently in Canada (Dean, McTaggart, Fish, &

Friesen, 1986) where growth hormone has been distributed and clinical data collected

centrally since 1967. While these subjects generally grew with human growth hormone

treatment, post-treatment their average height remained 3 standard deviations below the

mean height of children their same age and sex. These adults showed a high rate of

unemployment (35%) and never having been married (85%). Of the 96 subjects who had

completed formal education 73% lived with their parents. Only 58% had a drivers' license

and 21% had received psychological counselling at some point. All of these factors combined

indicate the less than satisfactory adjustment of these growth hormone deficient adults.

Impairment in heterosexual relationships and activities was also found by Money, Clopper,

and several associates in groups of adult males with hypopituitarism of various diagnoses

(Money & Clopper, 1975; Clopper, Adelson, & Money, 1976; Money, Clopper, & Menefee,

1980). However, in a sample of 39 growth hormone deficient adults who had previously been

treated with growth hormone, Clopper and colleagues (Clopper, MacGillivray, Mazur,

Voorhess, & Mills, 1986) found somewhat more adequate adult adjustment. Only 8% of the

subjects were unemployed, 95% completed high school and 70% of these completed further

education. The sample as a whole reported spending an average of 55% of their free time with

at least one other person. Nonetheless, 67% were still living with their parents and only

10% were married. Half of the sample reported current juvenilization by others and only

44% were satisfied with their physical appearance.


Psychological Correlates:
Growth Hormone Deficient Children

Similar to the early psychological investigations of growth hormone deficient adults, the

early work with children tended to utilize case studies, interviews, and projective testing.







9

Krims (1968) found a tendency toward psychological infantilism in his sample of 12 growth

hormone deficient children aged 4 to 15 years. He noted that the reaction of the child and his

environment to the perception of his extreme shortness was progressive, becoming worse as

the child ages. Boys face more psychological difficulties than girls, he felt, particularly in

adolescence when issues of role identification and occupational choice become relevant. The

short child may experience feelings of being unacceptably different, inadequate, and

incompetent. Sadness and grief may result.

Spencer and Raft (1974) outlined typical adaptive maneuvers and defense mechanisms

used by short statured individuals who are often subject to infantilization by others. In

reaction to failure to keep pace with their peers in terms of growth, the short statured child

may withdraw socially and avoid competition with peers, particularly during adolescence.

Conversely, he may react with overt aggression. Spencer and Raft felt that these adolescents

might not be able to complete the normal tasks of adolescence such as formation of sexual and

work identities. Instead, the stresses encountered during adolescence may cause a return to an

earlier ("pre-genital") level of adaptation.

Further personality research on short statured children has been conducted by

Steinhausen and Stahnke (1976, 1977). They looked at intelligence and other psychological

variables as a function of age, sex and socioeconomic status in 16 short statured children

without endocrine disorder as compared to normal controls. There were no differences

between short children with or without endocrine disorder. Compared to normals, short

statured children (regardless of etiology) did poorly on tests of spatial orientation and speed

of closure but were not different from normals on other intellectual factors or general

intelligence. Personality-wise the short children were less aggressive, less excitable, less

dominant, more conscientious, less shrewd, more controlled and less tense than normal

controls. From these findings Steinhausen and Stahnke depicted the short statured child as











living in a secluded inner world of intensified feelings, sentiment, and emotions, withdrawn

and unsociable. Age, sex and socioeconomic status were not important factors in the results.

In another study (Drotar, Owens, & Gotthold, 1980) 16 growth hormone deficient

children were compared to 16 normal height children on general adjustment (as measured by

the Missouri Children's Picture Series; Sines, Parker, & Sines, 1971), body image, sex role

development, sex-related fantasy, and reactions to frustration. The two groups were different

only in their reactions to frustration. The growth hormone deficient children used less

adaptive, mature solutions to frustrating situations than did the normal height children.


Self-Conceot

Several researchers have looked at self-concept in growth hormone deficient children.

Apter and his associates (1981) studied adolescents with combinations of pubertal delay and

growth retardation. They found that the presence of delay in sexual maturation by itself had no

significant negative effect on self-image. Both males and females showed self-concept

differences according to height in the areas of family relationships, external mastery,

superior adjustment, social relations, and moral attitudes. Due to the negative effects on

self-image of growth retardation, the authors endorsed speeding up growth where appropriate

in order to avoid these psychological difficulties.

Self-concept and emotional stability were studied by Shurka and her colleagues (1983)

in four groups of growth retarded subjects with different diagnoses. Groups of subjects with

high immunoreactive growth hormone deficiency, craniopharyngioma and

pan-hypopituitarism had low self-concepts and high levels of emotional problems. Subjects

with isolated growth hormone deficiency were emotionally stable and had self-concepts

similar to normals. The authors accounted for the differences between the groups by the fact







11

that the isolated growth hormone deficient subjects have the potential for adequate medical

treatment to alleviate their condition and the other three groups do not.


Treatment with Human Growth Hormone:
Psychological Correlates

When growth hormone deficient children were treated with human growth hormone, a

number of questions arose concerning their psychological adjustment. For example, Pollitt

and Money (1964) asked whether the physical benefits from growth hormone treatment would

motivate improved school achievement in underachieving short statured children. What effect

does this treatment have on psychological adjustment? Krims (1968) noted that some short

statured children reported intense unhappiness only after growth hormone treatment started

and they began to grow.

Even with adequate growth, feelings of inadequacy and incompetency may remain. Brust,

Ford, and Rimoin (1976) noted the presence of a range of difficulties in accepting treatment.

These varied from feelings of pressure to perform better to outright refusal to be treated.

These responses were explained by the authors as due to an identity crisis brought about by

the prospect of change from a familiar situation (i.e., short stature). Others faced with

treatment showed decreased denial and immature behavior coupled with increased overt

depression and anger.

Money and Pollitt (1966) studied 17 short statured patients under treatment with

human growth hormone. Based on interview and observation they described a "readjustment

syndrome":


.. the patient's ill health represents a chronic state
of disability to which he has been long accustomed. In
this instance, therapeutic intervention brings about
changes in corporeal self and the expectancies of life
experience. Such rapid change toward normalcy may
be as difficult to adjust to as a deforming injury or








other rapid departure from normalcy. (Money &
Pollitt, 1966, p. 387)

They suggest that this readjustment syndrome can be avoided if hormonal treatment is begun

very early in life before the child has had to adjust to life as an extremely short child.

The psychological status of a group of 11 children being treated with human growth

hormone was documented over the course of 2-1/2 years by Kusalic and Fortin (1975). The

status of these children changed over the course of treatment from immaturity, low

self-esteem and lacking aggressive impulses pretreatment as compared to normals to

depressed, apathetic and verbally aggressive after 2 1/2 years. Each treatment involved a 6

month series of injections followed by a 6 month rest period. According to projective testing

pretreatment, the children were psychologically immature as compared to a normal

population. They had low self-esteem and showed a lack of aggressive impulses. The overall

impression was one of distress and dejection. After the first treatment low self-esteem

became the main feature of their psychological make-up. Aggressive drives began to be

manifested in the children's fantasies. After the second treatment immaturity was still

present but depression shown most clearly in low self-esteem was even more prominent.

Aggressive drives were evident in occasional verbal discharges.

After the third treatment low self-esteem remained and depression became manifested

more in affective withdrawal. Aggressive drives appeared to increase in direct proportion to

changes in the patient's height. After the fourth treatment immaturity was still present but

even more evident was low self-esteem and a state of helplessness and emotional detachment.

At this point some patients seemed to regress. After the fifth treatment the group as a whole

remained depressed and apathetic but verbally aggressive. Any regression previously

exhibited became even more pronounced.

All of the patients in the Kusalic and Fortin (1975) study reportedly had unrealistic

hopes of obtaining normal height. Additionally 9 of 11 families showed a high degree of







13

pathology. There were parental schism and ambivalent, over-protective, covertly-rejecting

parental attitudes toward the short child. These pathological attitudes increased over the

course of treatment. In this sample 8 of 22 parents showed a marked psychiatric disorder.

Kusalic and Fortin concluded that this is a group of vulnerable children at very high risk for

psychological problems. The older the patient the longer he has been exposed to the feelings of

low self-esteem associated with small size and the less able he is to adjust to the change

brought about by growth hormone treatment. The authors found that the younger the child at

the beginning of treatment the better the subsequent adjustment. Also, girls tended to have

fewer adjustment problems during treatment.

Rotnem and her associates (1977, 1979) also followed the personality and social

development of a group of growth hormone deficient children being treated with human growth

hormone. Projective testing indicated that the majority of the children saw themselves as

socially isolated, powerless, vulnerable, incompetent, and suffering from low self-esteem.

These feelings were related to repeated instances of social rejection or failure to reach goals.

The children also showed inhibition of aggression. They tended to be hypersensitive to

criticism, anticipated rejection, increasingly based their self-esteem on the opinions of

others and felt hopeless and empty.

Parents of these children tended to be overprotective, showed excessive control, and

underestimated the emotional and developmental difficulties encountered by their children.

After one year of treatment the children tended to regard their treatment as a relative failure

even though their rate of growth was increased over baseline levels. This may be attributed to

unrealistic expectations for growth with treatment. Emotional reactions to the perception of

treatment failure included anger, pessimism, guilt, negativism and feelings of unacceptability.

Grew and his associates (1983) developed a visual technique to measure expectations of

growth in children treated with human growth hormone. They then introduced an educational







14

program aimed at altering family members' unrealistic expectations of growth from human

growth hormone replacement therapy. Almost 45% of growth delayed children and their

families overestimated their height relative to peers. Approximately 80% of the short

children had unrealistic expectations of the results of growth hormone therapy. After an

intervention composed of feedback, clarification, reframing and redirecting of ideas about

treatment, more subjects who had received intervention than those who had not accurately

estimated the short child's future height.


Behavioral Adjustment

At least two groups of researchers within the last few years have conducted research on

short statured children using sound research methods and measures with proven validity and

reliability. This is in contrast to most of the research previously cited which has been in

large part based on case studies, anecdotal reports, or projective measures.

Gordon and his group (1982) looked at 24 short statured children aged 6 to 12 years

with constitutional delay of growth. This group was compared to a group of 23 normal height

children matched for age, IQ, sex, and socioeconomic status. Parents filled out Achenbach's

(1979) Child Behavior Checklist and objective measures of family functioning and

child-rearing attitudes as well as providing interview material. Children completed the

Piers-Harris Self-Concept Scale (Piers, 1969). The short statured and normal height

groups were not significantly different on social competence factors indicating similar school

performance and involvement in activities. The groups were significantly different, however,

on the behavior problem index, particularly with regard to somatic complaints, schizoidal

tendencies and social withdrawal. Scores for the short children were so high they approached

the level typically found in children referred for psychological evaluation.







15

Parents of the constitutional short statured children appeared to be less strict in

child-rearing than parents of normal height children. On the self-concept measures the short

children tended toward lower scores. They more often described themselves as unhappy and

unpopular. There was no relationship between the height deficit of the child among short

statured children and the amount of psychopathology or impaired self-esteem. Age and sex of

the child did not affect the results in any systematic fashion. From this pattern of results the

authors concluded that constitutionally short statured children have significantly more

behavior problems and lower self-esteem than a matched group of normal height children.

The short children seemed to be socially withdrawn and aloof and tended to express their

emotional concerns internally.

A group of 27, 8 to 15 year old short statured children with either growth hormone

deficiency, constitutional delay, or short stature of unknown etiology was studied by

Young-Hyman (1986). Parents and the children themselves served as the respondents.

Children with earlier onset of growth delay tended to have more friends and longer-standing

relationships. However, children with greater growth delay tended to have fewer friends and

shorter relationships.

Holmes, Hayford, and Thompson (1982a, 1982b) have published two investigations of

the behavioral adjustment of short statured children. One of these studies (1982a) compared

the personalities and behavioral functioning of constitutionally delayed and growth hormone

deficient boys. Parents completed the Achenbach (1979) Child Behavior Checklist and

children were given the Missouri Children's Picture Series (Sines, Parker & Sines, 1971), a

nonverbal empirically derived measure of personality for 5 to 16 year olds. Results did not

differentiate the children based on diagnosis. Parents rated older boys as showing more

obsessive/compulsive behaviors and less aggressive behaviors than younger boys. The

children's results indicated that older boys are more conforming and inhibited. Altogether the







16

pattern of results shows this group of boys to be more withdrawn than boys of normal height

regardless of age.

The role of age and sex in the behavioral adjustment of short statured children was also

studied by Holmes, Hayford, and Thompson (1982b). Three groups of 6 to 16 year old

children were included with differing diagnoses: constitutionally delayed, growth hormone

deficient or Turner's syndrome. Parents rated the children on the Achenbach (1979) Child

Behavior Checklist while teachers used the Quay and Peterson (1979) Behavior Problem

Checklist. Adolescent girls were rated by both teachers and parents as showing the greatest

degree of behavioral immaturity, emotional inhibition and school problems of the groups

studied. Significant school problems were noted on parent ratings for all groups except for

younger constitutionally delayed children. Teachers rated all of the children as showing a

relatively high incidence of immature/inadequate behaviors except the adolescent males.

There were also indications of significant amounts of peer teasing. Additionally, 25% of the

subjects had been retained sometime in kindergarten through second grade, possibly due to

small size and immaturity. Drash (1969) also noted that short statured children are often

held back in school for these reasons. He felt this might have more of a negative effect on their

social development than a positive one as they will not catch up in growth within one year.

In a later study Holmes, Thompson, and Hayford (1984) looked at factors that might be

related to grade retention in their sample of short statured children. All of the children were

of at least average intelligence upon initial testing. Despite repeating a grade level in the

primary grades, the retained children continued to function 6 months below grade expectation

according to both teacher and parent ratings as well as a standardized achievement test.

Forty-seven of the children reported on by Holmes and her group in 1982 were

re-evaluated approximately three years later (Holmes, Karlsson, & Thompson, 1986).

According to parental ratings, the children showed an age-related decline in adjustment during







17

the early years of adolescence (beginning at approximately 12-14 years). Ratings of school

and social competence were about one standard deviation below the mean during these years.

Both before and after these years school and social functioning were rated by parents near the

50th percentile. By approximately 17 years of age both school and social competencies are

near age expectancy. As in the previous report (Holmes et al., 1982b) older females showed

more school problems than did other age/sex groups. While this is a report over only 3

years, the findings point to the need for longitudinal research into the adjustment of growth

hormone deficient children.


Summary

From this review of the literature relevant to the psychosocial adjustment of growth

hormone deficient children it is apparent that there has not been much research done using

reliable, valid measures, adequate controls and relatively large samples. Notable exceptions

are the studies of Holmes, Hayford, and Thompson (1982a, 1982b). There is also a lack of

research from a multivariate perspective.

The general picture of the growth hormone deficient child that emerges from the

literature to date is one of an immature, socially withdrawn child with a low self-concept who

has significant school problems despite at least average intelligence. The possibility of an

increased growth rate with growth hormone therapy (previously human growth hormone,

currently synthetic growth hormone) may cause initial optimism but may also result in

unrealistic expectations for ultimate height. When this height is obviously not being reached,

the child may respond by developing an even lower self-image and allowing release of

previously inhibited aggressive behavior.

One or two of the previous studies have looked at self-concept in growth hormone

deficient children (Apter et al., 1981; Shurka et al., 1983). Several studies have adequately







18

looked at adjustment using parent and/or teacher ratings of the short statured child's

behavioral functioning on some relatively objective inventory such as Achenbach's (1979)

Child Behavior Checklist (i.e., Holmes, Hayford, & Thompson, 1982a & 1982b). Teasing and

social relations seem to be very important issues in the adjustment of the growth hormone

deficient child. Therefore it is rather surprising that no one has investigated the short statured

child's sociometric status within his classroom.


Present Study

The present study was designed to address several questions. The first question

concerned the sociometric status of the growth hormone deficient child (being treated with

growth hormone) within his classroom. The second question concerned the correspondence

between perceptions of the child's "adjustment" from four different sources--both parents,

teachers, peers, and the child himself. Behaviors of particular importance were aggression

and social withdrawal. The review of the previous literature in this area indicates that these

two behavior problems may be particularly prevalent in growth hormone deficient children

receiving growth hormone replacement treatment. The last question concerned the

relationship between the behavioral adjustment of the growth hormone deficient children and

how realistically they perceive their present height in relation to other children of the same

age and sex.

It was predicted that growth hormone deficient children would be viewed relatively

more negatively than their peers on sociometric ratings. It was also predicted that the

perceptions of the children's adjustment from the four different sources--parents, teachers,

peers, and the children themselves--would be moderately correlated. Previous research has

suggested that children's self-reports may show the poorest correspondence with the other

measures (Ullman, 1952; Powell, 1948; Cox, 1966). It was also predicted that







19

correspondence between these raters will be different for the growth hormone deficient

children and the normal height children. Correspondence between ratings of the growth

hormone deficient children could be more accurate than those of normal height children due to

a tendency to be more attentive to the behavior of a child with a chronic medical condition.

Lastly, it was predicted that growth hormone deficient children who have unrealistic

perceptions of their height will be rated as more maladjusted than short statured children

with more realistic perceptions. Previous research has shown associations between

inaccurate perceptions of one's height and growth rate and a number of problem behaviors

from social withdrawal to aggression (i.e., Kusalic & Fortin, 1975; Rotnem et al., 1979).











CHAPTER II
METHOD


Subjects

Subjects were 45 growth hormone deficient children who were patients in the Pediatric

Endocrinology Clinic of Shands Teaching Hospital, the University of Florida, Gainesville. All

were currently receiving human growth hormone replacement treatment. Duration of this

treatment was from less than 1 year to more than 13 years.

A control group of 40 nongrowth hormone deficient children consisted of children who

were either volunteers from the growth hormone deficient child's class, or in the absence of

volunteer classmates, were volunteers from the University of Florida Laboratory School.



Measures

Child Behavior Checklist

While the entire Child Behaivor Checklist was completed by informants, only several of

its scales were used in this investigation. Each of the three social competence scales on the

parent version were utilized. Total behavior problem scores were used from all versions, but

the only behavior problem scales used were social withdrawal and aggression. These two

behavior problems have been reported to be of particular relevance in this population.

This measure is designed to assess the behavioral problems and competencies of children

aged 4 to 16 years (CBCL; Achenbach & Edelbrock, 1983). It consists of 113 behavior

problem items rated as either "not true," "somewhat or sometimes true" or "very true or

often true." Separate forms are available for parents, teachers, and children (11-16 years

of age).








21

The CBCL yields scores on 3 social competence scales as well as several behavior

problem scales, second-order factors, and total behavior problem and social competence

scores. The separate scales are plotted in profile form using T-scores. The particular scales

vary on the different forms according to age of the child in question. The parent form consists

of scales measuring behavior characterized as schizoid or anxious, depressed,

uncommunicative, obsessive-compulsive, somatic complaints, social withdrawal,

hyperactive, aggressive, and delinquent. The teacher form has scales measuring behavior

described as anxious, socially withdrawn, unpopular, self-destructive,

obsessive-compulsive, inattentive, nervous-overactive, and aggressive.

Agreement between ratings by mothers and clinicians averaged 83% for the

internalizing/externalizing factors and 74% for the lower level profile types (see Achenbach

& Edelbrock, 1983). The internalizing/externalizing factors were determined through factor

analysis of the behavior problem scales. The 1 week test-retest reliability for the

internalizing factor is 0.82, and for the externalizing factor is 0.91 (Achenbach & Edelbrock,

1983). The internalizing factor correlates 0.58 with the anxiety scale from the Conners

Parent Questionnaire and 0.62 with the psychosomatic scale, while the externalizing factor

correlates 0.45 with the impulsive/hyperactive scale and 0.77 with the conduct problem

scale (Achenbach & Edelbrock, 1983).

One week test-retest reliability for the total social competence scores is 0.89

(Achenbach & Edelbrock, 1983). In terms of content validity, clinically-referred children

received lower scores on all social competence items (Achenbach & Edelbrock, 1983).

The Youth Self-Report Form (YSRF) has not yet been factor analyzed to produce

separate factors. Instead, a total behavior problem score is derived. Test-retest reliability

over 6 months for this total behavior problem score is 0.69 (Achenbach & Edelbrock, 1983).

In terms of validity, the total behavior problem score has correlated 0.55 with








22

clinician-rated behavior problem scores (Achenbach & Edelbrock, 1983). While social

competence items are included in the YSRF, they have not as yet been normed.

Pupil Evaluation Inventory

This peer rating measure (Pekarik, Prinz, Liebert, Weintraub, & Neale, 1976)

consists of a matrix with 35 items down the left hand column and the name of each pupil in a

given classroom along the top of the page. Each child then checks which classmates he feels are

described by a particular item. Three factors account for 65% of the variance in scores

(Pekarik et al., 1976). They are Aggression, Withdrawal and Likeability. Item scores are

the sum of endorsements on a particular item for a given child. These are then divided by the

number of raters to facilitate intergroup comparison. An individual's score on a factor is

derived by averaging the previously obtained percentages for all items significantly loading on

that factor.

Test-retest reliability over two weeks ranged from 0.81 to 0.95 for the different

factors with both male and female raters. Looking at individual items, the median test-retest

correlations for male and female raters were 0.71 and 0.76, respectively (Pekarik et al.,

1976).

Teacher and self-ratings were correlated with peer ratings as a measure of concurrent

validity (Pekarik et al., 1976). The teacher-peer rating correlations ranged from 0.28 to

0.73 with a median of 0.57. The self-peer rating correlations ranged from 0.09 to 0.59 with

a median of 0.39. Correlations within each of the factors for teachers and peers were 0.65,

0.53, and 0.52 for Aggression, Withdrawal, and Likeability, respectively. For self-peer

comparisons the correlations were 0.46, 0.39, and 0.27.

Positive/Negative Peer Nomination Inventory

This instrument requires each pupil in a classroom to indicate the three classmates he

likes the most and the three he likes the least. These scales were considered separately. The








23

number of nominations for each was divided by the total number of children completing the

form to yield a percent score. This facilitated comparison across classrooms with different

numbers of students.

The median of test-retest reliability coefficients from Kane and Lawler's (1978)

review of such techniques was 0.78. They account for such a high reliability by the method's

focus on nominations of extreme members (3 liked most, 3 liked least). Hollander (1956)

showed that reliability seems to develop early in the life of a group. The median criterion

validity coefficient from the Kane and Lawler (1978) review was 0.43. Criteria included

graduation, promotion, and judgment of superiors.

Piers-Harris Self-Concept Scale

This measure of children's self-image (Piers,1969) consists of 80 first person

statements. The child circles "yes" or "no" for each item indicating whether he considers it

true for him. It was standardized on 1,183 children in grades 4 through 12. A factor analysis

accounting for 42% of the variance yielded at least 6 factors behavior, intellectual and

school status, physical appearance and attributes, anxiety, popularity, and happiness and

satisfaction. An overall self-concept measure for each child was obtained as were the 6 factor

scores.

Test-retest reliability over both a 2 and a 4 month interval was 0.77 for 244 fifth

graders (Wing, 1966). Concurrent validity has been investigated by comparing

Piers-Harris scores with scores from other similar measures. When compared with scores

on Lipsitt's (1958) Children's Self-Concept Scale for 98 special education 12 to 16 year old

students (Mayer, 1965), a correlation of 0.68 was obtained.

Reliability and validity coefficients are not reported for the six factors. It is

recommended that they be used primarily as research instruments (Piers, 1969).








24

It is important to note here that children may respond to the Piers-Harris Self-Concept scale

with a "social desirability" mindset. Millen (1966) found a correlation from 0.25 to

0.45 between this scale and the Children's Social Desirability Scale (Crandall, Crandall, &

Katkovsky, 1965).

Children's Social Desirability Scale

This is a 47 or 48 item inventory (CSD; Crandall, Crandall, & Katkovsky, 1965)

consisting of questions to which a child responds by circling either "yes" or "no" indicating

whether the question is true for him. The form for children in grades 6 and above contains one

more item than the form for younger children. An overall social desirability score was

obtained for each child. The CSD was developed by simplifying the language of items from the

Crowne-Marlowe (1960) Social Desirability Scale.

The scale was standardized on 956 students in grades 3 through 10. Test-retest

reliability (Crandall, Crandall, & Katkovsky, 1965) was 0.63 over a one month interval. In

terms of validity, CSD scores correlated 0.51 with the Good Impression scale from the

California Personality Inventory (Crandall, Crandall, & Katkovsky, 1965).

Silhouette Apperception Test-Revised

The Silhouette Apperception Test (Grew, Stabler, Williams, & Underwood, 1983) was

revised for the present study. The measure that was used consisted of a question which was

answered by circling one of five human silhouettes. The silhouettes were arranged in

increasing order of height from the third to the ninety-fifth percentile. The question was:

Children grow at different rates. These children are all the
same age. Which child looks the most like you? (The child was
told that the children were his same age and sex.)

Each growth hormone deficient child completed this form. A similar form asking for opinions

about the child in question was filled out by the physician assistant who worked closely with

all of these children. She was aware of the child's actual height percentile for age and sex, as








25

well as the child's bone age, height age, and growth rate. An index of how realistic the

children's perceptions were concerning their present height was derived. Each silhouette was

assigned a number from 1 to 5 in order of increasing height. The child's rating was subtracted

from the physician assistant's rating. Therefore realism ratings could vary from -4.0 to

+4.0.

Attractiveness Ratings

Two Polaroid pictures of each child were taken--one of the child's face and the other of

the child's entire body in order to indicate height. The pictures were rated for attractiveness

on a scale from 1 (not at all) to 5 (very) by students in similar grades in schools in Jackson,

Mississippi. The attractiveness rating for any particular picture was the average of the

ratings given to that picture by all of the students in one classroom. Therefore these ratings

could vary from 1.0 to 5.0.

Peabody Picture Vocabulary Test

This measure of receptive vocabularly (Dunn & Dunn, 1981) consists of 150 plates

roughly ordered in increasing difficulty. The subject's task is to identify which of four

pictured alternatives matches the word spoken by the examiner. This test was used as a global

estimate of intelligence.

The median split-half reliability coefficients were 0.80 and 0.81 for Form L and Form

M, respectively. In terms of criterion validity, the median correlation between the PPVT and

10 different vocabulary tests was 0.71 (Dunn & Dunn, 1981). Median correlations with the

Verbal and Full Scale scores of the Wechsler Intelligence Scale for Children-Revised were

0.71 and 0.72, respectively (Dunn & Dunn, 1981).









Procedure

Children were given the Peabody Picture Vocabulary Test (PPVT; Dunn, 1965) either

in clinic, at home, or at school.

Questionnaires were filled out by both parents and children at home and mailed to the

experimenter.

The experimenter visited each child's school to administer the peer rating measures.

Schools were located from Miami to Pensacola, Florida, to Valdosta, Georgia. Children in

grades 2 through 11 completed both the Peer Nomination Inventory and the Positive/Negative

Peer Nomination Inventory. Children in kindergarten and first grade completed only the

Positive/Negative Peer Nomination Inventory as they were unable to read the Pupil Evaluation

Inventory. In several other cases only the Positive/Negative Peer Nomination Inventory or

only the positive nomination portion of that inventory were administered at the request of the

school principal or county research committee.

Additionally, two pictures were taken of each child to be used to obtain attractiveness

measures. One picture of each child showed only his face, the other the entire child standing

against a door to indicate his height. These pictures were rated for attractiveness on a 5-point

scale by children in similar grade classrooms in Jackson, Mississippi.

Teachers were given their questionnaires at school and provided with a stamped envelope

in which to return them.

The physician assistant who worked closely with the growth hormone deficient children

completed a Silhouette Apperception Test-Revised concerning each child concurrent with their

being seen in the clinic.









CHAPTER III
RESULTS


Sample sizes vary in the different analyses due to several factors. These factors include:

fewer fathers and teachers participated than mothers; only adolescents completed the Youth

Self-Report Form; some schools did not allow the peer rating measures to be administered;

and children in kindergarten and first grade did not complete the Pupil Evaluation Inventory.


Characteristics of the Samole

Characteristics of both the growth hormone deficient and non-growth hormone deficient

children are shown Table 1. The growth hormone deficient sample was 60% male and 40%

female. The average age of the growth hormone deficient children was 12.2 years, ranging

from 5 to 20 years. Seventy-one percent were white, 16% were black, and 13% were

hispanic. The mean grade level was 5.8 with a range from kindergarten through grade 12.

PPVT scores ranged from 61 to 148 with a mean of 92.8.

The control group was 58% male and 42% female. The average age of the control

subjects was 11.4 years, ranging from 5 to16 years. Eighty-three percent of the children in

the control group were white and 17% were black. The mean grade level was 5.5 with a range

from kindergarten though grade 11. PPVT scores ranged from 48 to 146 with a mean of

104.9.

The two groups were significantly different on PPVT scores (t=3.12, p<.01). The mean

of the growth hormone deficient group was significantly lower then the mean of the control

group 92.8 (S.D.=18.6) versus 104.9 (S.D.=16.1).

Chi square analyses were done on a number of demographic variables to determine if

they varied between the two groups. There were no differences between the groups on income

(chi2=8.76, df=5, p.>05), sex (chi2=0.05, df=1, p>.05), and marital status (chi2=4.78,

df=4, p>.05).











TABLE 1
Sample Characteristics


Growth Hormone Deficient Children (n=45)


Control Children (n=40)


X=12.2 (5-20)


M=60%
F=40%


White= 71%
Black= 16%
Hispanic=13%

$ 0 9,000= 6%
10-19,000=52%
20-29,000=16%
30-39,000=10%
40,000+ =16%


Parental
Marital Status:


Grade Level:

Grades in School:


Married=74%
Other=26%

X=5.8 (K-12)


A's & B's=14%
B's & C's=60%
C's & D's=23%
D's & F's= 3%


X=11.4 (6-16)


M=58%
F=42%


White=83%
Black=1 7%


$ 0 9,000= 0%
10-19,000=23%
20-29,000=27%
30-39,000=20%
40,000+ =30%


Married=84%
Other=l 6%

X=5.5 (K-11)


A's & B's=55%
B's & C's=35%
C's & D's= 7%
D's & F's= 3%


Repeated a Grade
in School:


Peabody Picture
Vocabularly Test:


X=105.1 (48-146)


Age:

Sex:


Race:


Income:


yes=43%
no=57%


yes=10%
no=90%


- - - - --__ _


X=92.8 (61-148)








29

The two groups were different, however, on race (chi2=10.97, df=3, p<.01). The

groups contained essentially the same percentages of black subjects (16% vs. 17% for the

growth hormone deficient and control children, respectively). The remainder of the control

subjects were white (83%). The remainder of the growth hormone deficient sample was 71%

white and 13% hispanic. There were also differences between the groups on two

school-related indices. The two groups were significantly different on what kind of grades

they were currently making in school (chi2=13.07, df=4, p<.005). The growth hormone

deficient children were making more B's, C's, and D's than the control children who were

making more A's and B's. Significantly more of the growth hormone deficient

children had also repeated a grade level at some point (chi2=8.78, df=2, p<.01).


Height of the Growth Hormone Deficient Children

The mean height percentile for age and sex in the growth hormone deficient (GHD)

children was 5.1% (n=40) with a range from 1% to 50% and a standard deviation of 3.1%.

Since the age range of the GHD sample is so broad an average height for the total sample

would not provide useful information. Instead, Table 2 shows the average height for each sex

in small age ranges. In only one of the 11 age ranges does the mean for the range fall on the

growth curve (8-10 year old girls, mean at the 5th percentile). The means of the other age

ranges are all below the fifth percentile for age and sex.


Comparison of the Control Subjects from the Growth Hormone Deficient Subiects' Classrooms
and Those Who Were Not

The control subjects from the same classrooms as the growth hormone deficient

subject's classrooms (n=22) and the control subjects from the laboratory school (n=18)

were compared on all adjustment measures using analysis of variance (ANOVA) or

multivariate analysis of variance (MANOVA). As indicated in Table 3 the two groups were not











TABLE 2
Mean Height for Age and Sex of the
Growth Hormone Deficient Group


Girls

Aae (years)

6-8


8-10


10-12


12-14


14 +


Mean Height

3' 4 1/2"
(<5th percentile)

4' 0"
(at 5th %tile)

3' 8"
(<5th %tile)

4' 0"
(<5th %tile)

4' 6"
(<5th %tile)


Age (years)

6-8


8-10


10-12


12-14


14-16


16 +


Mean Height

3' 7"
(<5th %tile)

3' 8 1/2"
(<5th %tile)

4' 0"
(<5th %tile)

4' 4"
(<5th %tile)

4' 10 1/2"
(<5th %tile)

5' 3"
(<5th %tile)











TABLE 3
F Statistics Comparing Control Subjects
from the Growth Hormone Deficient Subjects'
Classrooms and Those from the Lab School on
Each Dependent Measure

Dependent Variables F value df D value

Child ratings:
self-concept 0.16 1,29 >.05

behavior problems 1.10 1,19 >.05

Mother ratings:
withdrawal 1.19 5,21 >.05
aggression
activity competence
social competence
school competence

internalization 0.99 2,26 >.05
externalization

overall social competence 1.01 2,24 >.05
total behavior problems

Father ratings:
withdrawal 1.98 5,17 >.05
aggression
activity competence
social competence
school competence

internalization 0.42 2,23 >.05
externalization

overall social competence 0.30 2,19 >.05
total behavior problems

Teacher ratings:
withdrawal 2.35 2,14 >.05
aggression

internalization 2.05 2,16 >.05
externalization











Table 3--continued.


Dependent Variables F value df p value

Teacher ratings (cont'd.):
grades 2.56 5,16 >.05
appropriate behavior
effort exerted
amount of learning
happiness

overall school competence 0.21 1,21 >.05

total behavior problems 1.94 1,18 >.05

Peer ratings:
withdrawal 1.17 3,27 >.05
aggression
likeability

classmates liked most 0.17 1,38 >.05

classmates liked least 2.10 1,37 >.05







33

significantly different in any of these analyses. The two types of control subjects were also

not significantly different on the PPVT (t (37)=1.36, p>,05). The mean PPVT score for the

subjects from the GHD children's classrooms was 101.7 (s.d.=17.1); the mean for the

laboratory school children was 108.7 (s.d.=14.5).


Comparison of the Growth Hormone Deficient Subjects with Idiopathic
Growth Hormone Deficiency and Those with All Other
Types of Growth Hormone Deficiency

The children who are growth hormone deficient as a result of tumors, infections, and

other types of known insults (n=18) were compared to the remainder of the growth hormone

deficient children idiopathicc GH deficiency; n=27) on all adjustment measures using analysis

of variance (ANOVA) or multivariate analysis of variance (MANOVA). As indicated in Table 4

these two groups were not significantly different in any of these analyses. Table 5 shows the

means of these two groups on several of the variables of interest.


Covariates: Miscellaneous Statistics

Social desirability was significantly related to child ratings of self-esteem (r=0.37,

n=67 pairs, p<.01). Therefore it was used as a covariate in the analyses dealing with

self-esteem.

Social desirability was not significantly related to behavior problem scores from the

Youth Self-Report Form of the Child Behavior Checklist (r=0.23, n=44 pairs,p>.05). It was

not used as a covariate for this set of analyses.

The growth hormone deficient and control groups were not significantly different on the

social desirability measure (t(65)=1.36, p>.05). The means for the two groups,

respectively, were 20.6 and 17.5.

Both attractiveness ratings, face and whole body, were significantly related to one of the

sociometric factors, social withdrawal, on the Pupil Evaluation Inventory (face: r=-0.32,

n=58, p<.05; body: r=-0.35, n=58, p<.01). Therefore, they were used as covariates in the










TABLE 4
F Statistics Comparing Growth Hormone
Deficient Children with Idiopathic Growth
Hormone Deficiency and Those with All
Other Types of Growth Hormone Deficiency
on All Dependent Variables

Dependent Variables F value df p value

Child ratings:
self-concept 1.94 1,33 >.05

behavior problems 1.44 1,21 >.05

Mother ratings:
withdrawal 0.89 5,27 >.05
aggression
activity competence
social competence
school competence

internalization 0.00 2,32 >.05
externalization

overall social competence 0.07 2,30 >.05
total behavior problems

Father ratings:
withdrawal 0.90 5,18 >.05
aggression
activity competence
social competence
school competence

internalization 0.97 2,23 >.05
externalization

overall social competence 0.56 2,20 >.05
total behavior problems

Teacher ratings:
withdrawal 1.21 2,27 >.05
aggression

internalization 1.41 2,29 >.05
externalization










Table 4--continued.


Dependent Variables F value df D value

Teacher ratings (cont'd.):
grades 0.41 5,25 >.05
appropriate behavior
effort exerted
amount of learning
happiness

overall school competence 0.54 1,29 >.05

total behavior problems 1.18 1,30 >.05

Peer ratings:
withdrawal 0.59 3,19 >.05
aggression
likeability

classmates liked most 0.70 1,34 >.05

classmates liked least 2.37 1,32 >.05









TABLE 5
Means and Standard Deviations of Several Variables
for the Growth Hormone Deficient Children with Idiopathic
Growth Hormone Deficiency and Those with All Other Diagnoses


Diagnosis


l lA Qth rs (n 18)


12.6 (3.4)

94.1 (21.6)


Child Ratings:
self-esteem:
social desirability
total behav. probs.


73.0
20.1
43.4


11.6 (3.9)

90.0 (13.4)


(16.7)
(8.7)
(22.4)


65.9
21.2
56.6


(18.7)
(8.9)
(30.0)


Mother Ratings:
aggression
withdrawal
total social
competence

Father Ratings:
aggression
withdrawal
total social
competence

Teacher Ratings:
aggression
withdrawal
effort
amt. learned

Peer Ratings:
aggression
withdrawal
likeability
liked most
liked least


63.9 (9.9)
60.1 (6.5)

37.8 (10.1)


61.3 (8.3)
57.2 (11.5)

40.9 (11.3)


62.7
56.7
52.4
48.8


13.4
17.6
26.1
11.1
8.6


(9.3)
(3.0)
(7.5)
(11.4)


(9.3)
(11.2)
(17.6)
(10.7)
(9.8)


60.7 (6.8)
58.4 (4.5)

36.6 (7.8)


59.1 (4.3)
57.0 (3.0)

37.5 (9.1)


59.0
55.5
55.3
49.2


14.2
25.8
22.5
8.3
15.1


(4.3)
(1.2)
(8.1)
(9.6)


(11.5)
(12.5)
(18.3)
(8.8)
(15.2)


PPVT


1 pA, hi In 071 Al =I


idI hi (n 27)







37

analyses dealing with the Pupil Evaluation Inventory. Neither of the attractiveness ratings

was significantly related to nominations of peers liked the most nor peers liked the least.

Therefore, they were not used as covariates in analyses dealing with these variables. Table 6

shows the correlations between the sociometric and attractiveness ratings.

The growth hormone deficient and control groups were not significantly different on

either of the attractiveness ratings. A multivariate analysis of variance (MANOVA) was used

to compare the two groups on the attractiveness ratings (F(2,77)=0.73, p>.05).


Ratings by Each Source

Either analysis of variance (ANOVA) or multivariate analysis of variance (MANOVA)

was used to compare the growth hormone deficient children and nongrowth hormone deficient

children on ratings by each source. Age (divided at the overall mean: <12 years versus >12

years), sex, score on the Peabody Picture Vocabulary Test (divided at the overall mean: <98

versus >98), and interactions involving group (growth hormone deficient versus control) and

each of these separate factors were initially included in each model. Any of these factors which

were nonsignificant were dropped from the model. They will be discussed only where they

made significant contribution to the model. Therefore, unless specifically stated, analyses

will involve only the factor Group (growth hormone deficient [GHD] versus control).

The number of subjects included in each of the following analyses varied depending on

the number of respondents in the particular area in question. All figures indicate the number

of cases included in each analysis.

Ratings by the Child

Scores on the Children's Social Desirability Scale (CSD) were used as a covariate in the

two analyses of child's self-report of self-esteem, both total score and separate scales from

the Piers-Harris Self-Concept Scale. The two groups were significantly different on total

self-esteem scores from the Piers-Harris Self-Concept Scale (F(1,64)=7.51, p<.01). The











TABLE 6
Correlations between Peer Ratings
and Attractiveness Ratings

Attractiveness Rating


Peer Rating

PEI:
Aggression

Withdrawal

Likeability


Face


-0.02 (n=58)

-0.32 (n=58)*

0.10 (n=58)


Full Rndv


0.02 (n=58)

-0.35 (n=58)**

0.15 (n=58)


Peer Nomination Inventory:
Classmates Liked Most

Classmates Liked Least


0.05 (n=75)

-0.00 (n=72)


*=p<.05
=p<.01


-0.01

-0.13


(n=74)

(n=71)







39

mean of the GHD group ratings was relatively lower than that of the control group. The mean

of the control group was rather high, however, compared to the norms (Piers, 1969).

The groups were not significantly different on a comparison of self-ratings on the six

separate scales from the Piers-Harris (F(6,56)=1,46,p>.05) with social desirabiity taken

into effect. The differences found on the analyses of self-esteem are depicted in Figures 1 and

2. The figures show differences in the actual means.

There was a Group X PPVT interaction on total behavior problem scores from the Youth

Self-Report Form of the Child Behavior Checklist (F(1,40)=4,30,p<.05). Figure 3 shows

this interaction. While the GHD and control group means are similar for subjects with higher

PPVT scores, means for the two groups are very different for subjects with lower PPVT

scores. GHD children with lower PPVT scores rated themselves as having significantly more

(frequency and/or severity) behavior problems than control subjects with lower PPVT

scores.

Ratings by Parents

All ratings of adjustment by mothers and fathers are from the Child Behavior Checklist

(CBCL) and involve T scores. Ratings by mothers and fathers were analyzed separately. For

each parent's ratings, three multivariate analyses of variance were conducted. The first

analysis looked at scores from individual scales chosen because they were found to be of

particular relevance in this population--two behavior problem scales (withdrawal and

aggression) and three social competence scales (activities, social involvement, and school).

The second analysis looked at the second-order behavior problem factors of internalizing and

externalizing. The third analysis included both the overall behavior problem score and the

overall social competence score.

Ratings by Mothers

The first analysis of the five individual scales was significant overall (F(5,54)=8.99,

p<.001). Separate ANOVAs indicated that the two groups were significantly different on








40















too
I00 GHO Kids


80 17 n :3 Control Kids


n=36
70 1
70-

Percent --- (normative mean=65%,
Piers. 1969)
60


Items

50

Endorsed


40



30 -


Self-Esteem













FIGURE 1. Mean ratings by children: Self-esteem total.
(Actual means are depicted.)























GHO Kids


Control Kids






















n=TO
83 n=3.
3


Behavior Intell & Physical Anxiety Popularity Happiness .
School Appear Satisfact'or
Status & Attri-
butes






FIGURE 2. Mean ratings on each self-esteem factor.
(Actual means are depicted.)


Scores
























- GHD Kids

--- Control Kids


n=14
528


50 t-


n=7
243


PPVT










FIGURE 3. Mean ratings on Youth Self-Report Form.
(Group X PPVT interaction.)


60 -


Raw


40 --


Scores


n=9
40 6


304


x


39 2


20 t







43

mother-rated withdrawal (F(1,58)=5.07, p<.05), competence in activities (F(1,58)=9.51,

p<.01), in social functioning (F(1,58)=25.73, p<.001), and at school (F(1,58)=17.58,

p<.001). GHD children were rated by their mothers as more withdrawn and less socially

competent in all three areas than were control children. No significant difference between the

groups was found on maternal ratings of aggression (F(1,58)=0.18, p>.05). Figure 4

illustrates these comparisons.

The second analysis looked at mother-ratings of internalization and externalization. The

two groups were not significantly different in this analysis (F(2,61)=2.09, p>.05).

The third analysis comparing mother-ratings of overall behavior problems and social

competence was significant overall (F(2,57)=15.17, p<.001). Separate ANOVAs showed that

the groups were rated differently only on overall social competence (F(1,58)=30.67,

p<.001; overall behavior problems: F(1,58)=2.10, p>.05). The GHD children were rated as

significantly less socially competent overall than were the control group children. Figure 5

shows the comparisons between the GHD and control groups on these last two analyses.

Ratings by Fathers

The first analysis comparing the five individual scales was significant overall

(F(5,41)=2.66, p<.05). Separate ANOVAs showed that the groups were rated significantly

different by their fathers on two of the three social competence factors--activities

(F(1,45)=8.09, p<.01) and social functioning (F(1,45)=6.19, p<.05). On each of these

factors the GHD children were rated as less competent by their fathers. The groups were not

significantly different on paternal ratings of withdrawal (F(1,45)=0.30, p>.05), paternal

ratings of aggression (F(1,45)=0.63, p>.05) or school competence (F(1,45)=2.18, p>.05).

Figure 6 shows the comparisons between the groups in this analysis.

The second analysis looked at father ratings of internalization and externalization.

There was a significant Group X Age interaction (F(2,47)=3.72, p<.05). Separate F tests

indicated that the effect was more likely to be in the ratings of internalization


















GHD Kids


Control Kids













n-'25


Withdrawal Aggression Activities Social Scnool

Social Competence Factors













FIGURE 4. Mean ratings by mothers on withdrawal, aggression,
and all three social competence factors.


T

Scores

50




40

















SGHO Kids


Control Kids


T

Scores n=5
5 .4 n=35 n=34
=29 5!7 -219
S.=29 r=28n
3.4 3 42 8
s ^3 5018
44,

50-




40 37


30 ... .....', .xv' ... ..
30 Z4

Internal- External- Total behavior Overall Social
vzatidn ization Problems Competence














FIGURE 5. Mean maternal ratings of internalization, externalization,
total behavior problems and overall social competence.








46

(F(1,48)=3.46, p>.05; externalization: F(1,48)=0.00, p>.05). Figure 7 shows this effect.

Newman-Keuls tests comparing the means of of each cell indicated that they were not

significantly different. The pattern of the interaction must therefore be examined. Figure 7

shows that while the control group children were rated approximately the same in the two age

groups (<12 years and >12years) the older GHD children were rated as more internalizing

than were the younger GHD children.

The MANOVA comparing overall behavior problems and overall social competence

ratings by fathers indicated that the two groups were significantly different

(F(2,42)=4.93,p<.05). Separate ANOVAs showed that the groups were different on only

overall social competence (F(1,43)=9.50,p<.01). They were not different on total behavior

problems (F(1,43)=0.13,p>.05). Figure 8 illustrates these comparisons.

Teacher Ratings

All teacher ratings are from the Teacher Report Form of the Child Behavior Checklist.

The first analysis compared the two groups on the behavior problems scales of withdrawal and

aggression. While the overall MANOVA was significant (F(2,44)=3.32, p<.05), separate

ANOVAs on each of the two scales failed to find significant effects (withdrawal: F(1,45)=2.66,

p>.05; aggression: F(1,45)=2.08, p>.05). Figure 9 depicts these comparisons.

The second analysis compared teacher ratings on the second-order factors of

internalization and externalization. While there was a significant Group X Age interaction

(F(2,46)=3.80,p<.05) separate ANOVA's indicated that the two groups were not significantly

different on either teacher ratings of internalization or externalization (internalization:

(F(1,47)=0.57,p>.05; externalization: (F(1,47)=2.40, p>.05). Figures 10 and 11

illustrate the possible interactions for these two variables.

The third analysis compared the teacher ratings of the two groups on total behavior

problems. The groups were not rated by their teachers as having significantly different



















SGHD Kids


Control K

Withdrawal Aggression Activities Social SCI'o!

Social Competence Factors













FIGURE 6. Mean ratings by fathers on withdrawal, aggression
and all three social competence factors.


T

Scores








48















- tGHD Kids


Control Kids


70t-


n: 14
58 7
n= 16
53 2



n=12 n=10
503 527


50 -


i 12 years


ears
> 12 years


Age









FIGURE 7. Mean ratings by fathers on internalization.
(Group X Age interaction.)


T

Scores




















GHD KldS


\ Control KI


T


Scores
60 ns=5
55 n '25
54
n= 3
49 4
50-


39
40-




30


Total Behavior Overall Social
Problems Competence












FIGURE 8. Mean ratings by fathers on total behavior problem
and overall social competence.




















n. GHD Kids


SControl K ;l


n=20
5.79


n=32
563


Withdrawal Aggression


FIGURE 9. Mean ratings by teachers of withdrawal and aggression.


T


Scores







51

numbers of total behavior problems (F(1,41)=1.8, p>.05). Figure 12 illustrates this

comparison.

The fourth analysis of teacher ratings used a MANOVA to compare teacher ratings of the

two groups on five variables--grades, effort, appropriateness of classroom behavior, amount

of learning, and happiness. The overall MANOVA contained both significant effects of Group

(F(5,46)=3.03, p<.05) and PPVT (F(5,46)=3.53,p<.01). Separate ANOVAs on each of the

variables indicated that GHD children were rated as having lower grades (F(1,50)=10.19,

p<.01), putting forth less effort (F(1,50)=7.23, p<.05), learning less (F(1,50)=4.45,

p<.05), and being more unhappy (F(1,50)=6.05, p<.05) than control children. The groups

were not different on ratings of extent of inappropriate behavior in the classroom

(F(1,50)=0.01, p>.05). These comparisons are depicted in Figure 13.

In terms of PPVT, separate ANOVAs indicated that the effect of PPVT was significant only

on teacher ratings of grades in school (F(1,50)=9.08,p<.01). Subjects with lower PPVT

scores were rated by their teachers as having significantly lower grades than subjects with

higher PPVT scores. This effect is shown in Figure 14.

The final analysis of teacher ratings compared the two groups on total classroom

behavior. The groups were significantly different on teacher ratings of total classroom

behavior (F(1,52)=7.53,p<.01). The GHD group were rated as less adaptive on overall

classroom behavior. This comparison is depicted in Figure 15.

Peer Ratings

Three analyses were conducted on peer-generated data. The two attractiveness measures

were included as covariates in the analysis of the Pupil Evaluation Inventory as one of its

factors was significantly correlated with them.

The first analysis was a MANOVA which looked at differences between the groups on the

three factors of the Pupil Evaluation Inventory. Both Group status and PPVT score had

significant main effects (Group: F(3,50)=3.3, p<.05; PPVT: F(3,50)=4.0, p<.05). Separate





































T

Scores


- GHD Kids

Control Kids


70-


n=18
57 0


60-+


so+5


40 -


n=14


4


55 2


n=6
49 3


i 12 years


> 12 years


Age







FIGURE 10. Mean teacher ratings on internalization.
(Group X Age interaction.)


I -

























I GHD Kids

Control Kids


80-t


70-


n=18
56.9


60-4


n=14
516


50-t


n=13
50 1


n=6
515


40-t


s 12 years


> 12 years


FIGURE 11. Mean teacher ratings on externalization.
(Group X Age interaction.)


Scores




















GHD Kids


Control Kids
35




30 n=52
27.2


25-

Raw


20--* n0


Scores

15




10


Total Behavior Problems


FIGURE 12. Mean teacher ratings of total behavior problems.




















GHD Kids


Control Kids


Grades Effort Behavior Amount of
Learning


HapDiness


FIGURE 13. Mean ratings by teachers on various factors.


T

Scores








56












QjL- PPVT <98


80-- PPVT >98




S 70-


Scores
60- nr 26
58


50-
n= 29
43 4


40
....... .... ..


..............
....... ...... .

30 ff__

Marks in School












FIGURE 14. Mean teacher ratings of marks made in school.
Note: Higher T-scores=higher marks.
























SControGHD Kids


Control Kids


Overall Classroom Behavior


FIGURE 15. Mean teacher ratings of overall classroom behavior.


Scores







58

F tests indicated that both of these effects were significant only for the PEI factor Social

Withdrawal (Group: F(1,52)=7.7, p<.01; PPVT: F(1,52)=7.0, p<.05). GHD children were

rated by their peers as more withdrawn than were control children. Children with lower

PPVT scores were also rated as more withdrawn than were children with higher PPVT scores.

There were no significant differences between the two groups on either of the other PEI factors

(aggression: F(1,52)=0.4, p>.05; likeability: F(1,52)=3.8, p>.05). There was a trend,

however, for the GHD children to be rated as somewhat less likeable than the control children.

The group comparisons are illustrated in Figure 16. The PPVT comparison is shown in Figure

17.

The second analysis compared the two groups on peer nominations of the three

classmates liked the most. The two groups were nominated at significantly different rates

(F(1,74)=5.26, p<.05). The GHD children were nominated by significantly fewer of their

peers as one of the three classmates liked the most.

The third analysis compared the two groups on peer nominations of the three classmates

liked the least. The two groups were not nominated by significantly different rates of their

peers as classmates liked the least (F(1,71)=0.01, p>.05). Figure 18 depicts group

differences on the peer nomination data.


Correspondence Between Ratings of Growth Hormone
Deficient Children by Different Sources

Table 7 shows the correlations between the ratings of total behavior problems of growth

hormone deficient children by each source--mothers, fathers, teachers, and the child himself.

The only significant relationship was between the mothers' and fathers' ratings of their

children's total behavior problems (r=0.66, p<.01). The other relationships vary from 0.23

to 0.39.
















r GHD KKds
T


Control Kids


n



ri=


Aggression Social Withdrawal Likeability


FIGURE 16. Mean ratings by peers on all three PEI factors.


Factor


Scores





















61: PPVT <98


35 -- PPVT >9




30 -

Factor

25

Scores
n: 24
20- 12




15 13"




10


Social Withdrawal











FIGURE 17. Mean social withdrawal ratings by peers
according to PPVT scores.



























18
Percent

of
16

Classmates

14

Nominated


GHD Kids


Control Kids


Classmates Liked Classmates Liked
the Most the Least


FIGURE 18.


Mean percent nominated by peers as
one of three classmates liked the
most or least.







62

The mothers and fathers also showed significant agreement in their ratings of each social

competence factor and overall competence. On the Activities factor the correlation was 0.70

(p<.01, n=25 pairs). The Social factor showed a correlation of 0.71 (p<.01, n=25 pairs).

Correlation for the School factor was 0.82 (p<.01, n=21 pairs). On overall Social

Competence ratings between parents correlated 0.75 (p<.01, n=21 pairs).

Correlations between mother, father, and teacher ratings of the second-order behavior

problem factors internalization and externalization are shown in Table 8. All of the mother

and father ratings for both of these factors were significant correlated with each other

(r=0.55 for internalizing and r=0.75 for externalizing). Teacher ratings of internalization

and externalization were significantly related to each other (r=0.71,n=32 pairs,p<.01).

Teacher ratings were not significantly related to either maternal or paternal ratings of

internalization or externalization.

Table 9 shows the correlations between ratings of the growth hormone deficient children

on various individual behavior problem scales and overall self-esteem by the different

sources.

Mother-Father Correspondence

Ratings of withdrawal by mothers and fathers were significantly correlated (r=0.71,

p<.01), as were ratings of aggression (r=0.43, p<.05). Correspondence was greater for

withdrawal than aggression, however.

Mother-Teacher Correspondence

Ratings by mothers and teachers of withdrawal were not significantly correlated

(r=0.36, p>.05). Mother and teacher ratings of aggression were, however, significantly

related (r=0.49, p<.05).

















Child (n)

1.00 (23


TABLE 7
Correlations between Ratings
of Total Behavior Problems
by Each Source
(Growth Hormone Deficient Group 0

Mother (n) Father (n)

3) 0.37 (22) 0.34 (17)

1.00 (34) 0.66 (24)**

1.00 (25)


nly)

Teacher (n)

0.23 (16)

0.39 (25)

0.27 (19)

1.00 (32)


Child

Mother

Father

Teacher


*=p>.05
**=p<.01











TABLE 8
Correlations between Ratings by Each Source
of Internalization and
Externalization
(Growth Hormone Deficient Group)

Father


Internali-(n)
SMtin


Externali-(n)
f;tion


Internali-(n)
Stinn


Externali-(n)
7atinn


Internali-(n)
7atinn


Externali-(n)
7atinn


1.00(35) 0.75(35)** 0.55(25)** 0.61(25)** 0.18(25) 0.16(25)


1.00(35) 0.46(25)*


1.00(26)


0.75(25)** 0.36(25) 0.31(25)


0.78(26)** 0.27(18) 0.08(18)

1.00(26) 0.28(18) 0.21(18)


1.00(32) 0.71(32)*

1.00(32)


Mother


Mother:
Int.


Teacher


Father:
Int.


Teacher:
Int.


*=p<.05
**=p<.01


Issas U told W-1 955, X.^ to











TABLE 9
Correlations between Ratings by Each Source
of Separate Behavior Problems
(Growth Hormone Deficient Group Only)


Mother
With-
drawal(n) e


Aggr-
ssion(n)


Father
With- Aggr-
drawal(n\ pA.innfn


Teacher
With- Aggr-
drawal(n) ession(n)


1.00(35) 0.54(35)**


0.71(25)** 0.17(25)


0.36(26) 0.29(25)


Aggr-
ession 1.00(35) 0.34(25) 0.43(25)* 0.22(26)

Father:
With-
drawal 1.00(26) 0.51(26)** 0.21(19)

Aggr-
ession 1.00(26) -0.24(19)


Teacher:
With-
drawal


Aggr-
ession


0.49(25)*


0.25(18)


-0.22(18)


1.00(32) 0.53(30)**


1.00(32)


*=p<.05
**=p<.01


Mother:
With-
drawal


ssion(n) rlr;;w.qltnl P-,.qinnInI









Table 9--continued.

Child Peers
Self- Aggr-
Esteem(n) ession(n)


Mother:
With-
drawal -0.30(35)

Aggr-
ession -0.15(35)

Father:
With-
drawal -0.20(26)

Aggr-
ession -0.21(26)

Teacher:
With-
drawal -0.02(26)

Aggr-
ession 0.07(25)

Child:
Self-
Esteem 1.00(36)

Peers:
Aggr-
ession

Withdrawal

Likeability

Liked Most

Liked Least


-0.27(19)


0.29(19)



-0.33(15)


0.10(15)


With- Like- Liked Liked
drawal(n) abilitv(n) Most(n) Least(n)


-0.13(19)


0.04(19)



0.31(15)


0.37(15)


0.26(20) 0.16(20)


0.67(21)**-0.01(21)


0.08(19)


0.20(19)



-0.03(15)


-0.13(15)



-0.34(20)


-0.05(21)


0.04(26)


0.14(26)



-0.35(21)


-0.39(21)



-0.23(28)


0.12(28)


-0.15(20) -0.26(20) 0.31(20) 0.06(27)


1.00(25) 0.21(25)

1.00(25)


-0.03(25)

-0.13(25)

1.00(25)


-0.09(24)


0.07(24)



0.06(19)


0.01(19)



0.43(26)*


0.36(26)



-0.40(25)'


0.26(25) 0.52(25)"

-0.25(25) 0.59(25)"

0.79(25)**-0.36(25)

1.00(36) -0.07(34)

1.00(34)


*=p<.05
*=p<.Ol









Mother-Child Correspondence

Neither maternal ratings of withdrawal nor maternal ratings of aggression were related

to child ratings of self-esteem (withdrawal: r=-0.30, p>.05; aggression: r=-0.15, p>.05).

Mother-Peer Correspondence

Neither mother and peer ratings of withdrawal nor mother and peer ratings of

aggression were significantly related (withdrawal: r=-0.13, p>.05; aggression: r=0.29,

p>.05).

Father-Teacher Correspondence

Neither father and teacher ratings of withdrawal nor father and teacher ratings of

aggression were significantly related (withdrawal: r=0.21, p>.05; aggression: r=-0.22,

p>.05).

Father-Child Correspondence

Neither father ratings of withdrawal nor father ratings of aggression were significantly

related to child self-esteem ratings (withdrawal: r=-0.20, p>.05; aggression: r=-0.21,

p>.05).

Father-Peer Corresoondence

Neither father and peer ratings of withdrawal nor father and peer ratings of aggression

were significantly related (withdrawal: r=0.31, p>.05; aggression: r=0.10, p>.05).

Teacher-Child Corresoondence

Neither teacher ratings of withdrawal nor teacher ratings of aggression were

significantly related to child self-esteem ratings (withdrawal: r=-0.02, p>.05; aggression:

r=0.07, p>.05).

Teacher-Peer Corresoondence

Teacher and peer ratings of withdrawal were not significantly related (r=0.17, p>.05).

Teacher and peer ratings of aggression were significantly related, however (r=0.66, p<.01).









Child-Peer Correspondence

Neither peer ratings of withdrawal nor peer ratings of aggression were significantly

related to child self-esteem ratings (withdrawal: r=-0.26, p>.05; aggression: r=-0.15,

p>.05). There was a significant inverse relationship between child self-esteem ratings and

peer ratings of those classmates liked least (r=-0.40, p<.05). High self-esteem ratings were

associated with a low number of nominations as a classmate liked the least.

Table 10 shows the correlations between the six scales of the Piers-Harris

Self-Concept scale (behavior, intellectual and school status, physical appearance and

attributes, anxiety, popularity, and happiness and satisfaction) and the ratings of withdrawal

and aggression by all sources and all sociometric ratings. Higher scores on each self-concept

factor indicate more adaptive attitudes or behavior. Of the 66 correlations listed only six of

them reached significance. Four of these were relationships with peer nominations of

classmates liked the least. Lower self-ratings of behavior (r=-0.38,n=25 pairs,p<.05),

intellectual and school status (r=-0.46,n=25 pairs,p<.05), popularity (r=-0.48,n=25

pairs,p<.05), and happiness and satisfaction (r=-0.53,n=25 pairs,p<.01) were associated

with higher rates of nomination as a peer liked the least. Peer ratings of social withdrawal

were related to self-ratings of anxiety (r=-0.50,n=20 pairs,p<.05) such that higher ratings

of social withdrawal were related to higher levels of anxiety. Peer ratings of likeability were

significantly related to self-ratings of intellectual and school status (f=0.58,n=20

pairs,p<.01) such that higher ratings of likeability were related to higher self-ratings of

intellectual status.

A final approach to looking at the correspondence between ratings by the different

sources is illustrated in Table 11. Correlations were derived between each of the peer rating

measures and each of the parental social competence scales and total competence measure. The

notion between this approach involves conceptualizing the peer rating data as peer-generated










TABLE 10
Correlations between Ratings by Each Source
of Separate Behavior Problems and Child Ratings
of Each Self-Esteem Factor
(Growth Hormone Deficient Group Only)


Child Self-Esteem Factors
Intell. & School Phys.
Status(nl Anoear. (n An


Popul- Happin. &
xiptv(n) aritvln\ R.atic In\


-0.09(33) -0.11(33)


-0.09(33) -0.09(33)


-0.32(33)


-0.23(33) -0.13(33)


0.03(33) -0.07(33) 0.11(33)


-0.19(33)


-0.02(33)


0.26(24) 0.01(24)


0.00(24) -0.03(24)


0.02(25) -0.12(25)


-0.06(24) -0.01(24)


-0.23(24)


-0.15(24)


-0.08(25)


-0.30(24)


-0.20(24)


-0.01(24)


0.03(24)


0.39(25) -0.26(25)


0.35(24) 0.35(24) 0.13(24)


-0.04(24)


-0.08(24)


-0.19(25)


0.12(24)


Behav-
ior(n)


Mother:
With-
drawal


Aggr-
ession


Father:
With-
drawal

Aggr-
ession

Teacher:
With-
drawal


Aggr-
ession


*=p<.05
**=p<.01


io~n i-t- riW ntQ n










Table 10--continued.
Child Self-Esteem Factors
Behav- Intell. & School Phys. Popul- Happin. &
ior(n) Status(n) AoDear.(n) Anxietv(n) arity(n) Satis.(n)

Peers:
With-
drawal -0.11 (20) -0.22 (20) -0.18 (20) -0.50 (20)**-0.24 (20) -0.19 (20)

Aggr-
ession -0.41 (20) -0.18 (20) 0.24 (20) 0.07 (20) -0.21 (20) -0.32 (20)

Like-
ability 0.36 (20) 0.58 (20)** 0.35 (20) -0.04 (20) 0.37 (20) 0.25 (20)

Like
Most 0.01 (26) 0.22 (26) 0.24 (26) 0.13 (26) 0.14 (26) -0.03 (26)

Like
Least -0.38 (25)*-0.46 (25)*-0.30 (25) -0.27 (25) -0.48 (25)* -0.53 (25)*







*=p<.05
**=p<.01







71

social competence ratings. Table 11 shows that peer ratings of aggression, likeability, and

nominations of classmates liked the most were not significantly related to any of the parental

social competence ratings. Maternal ratings of school competence and total social competence

were significantly related to peer-rated social withdrawal (r=-0.50 and -0.54,

respectively). Father ratings of social competence and total competence were also

significantly related to peer-rated social withdrawal (r=-0.53 and -0.63, respectively).

Each of these correlations indicates an association between high levels of parent-rated social

competence and low levels of peer-rated social withdrawal. Additionally, father-rated social

competence and both mother and father-rated total social competence were significantly

related to peer nominations of classmates liked the least (r=-0.50, -0.43, and -0.49,

respectively). Each of these correlations indicates an association between high levels of

parent-rated social competence and low numbers of nominations as a classmate liked the least.


CorreDondence Between Ratings of Control Children
and Total Sample by Different Sources

Correlations were also obtained between ratings by the different sources for the control

group and for the sample as a whole. In this section only patterns of relationship different

from those found with the growth hormone deficient children alone will be discussed.

In terms of total behavior problems, correlations of the total sample contained three

more significant relationships. Child self-ratings of total behavior problems were

significantly related to both mother and father ratings of total behavior problems

(r=0.37,n=41 pairs,p<.05, and r=0.34,n=34 pairs,p<.05, respectively). Mother ratings of

total behavior problems also become significantly related to teacher ratings of total behavior

problems in the total sample (r=0.40,n=37 pairs,p<.05).

The internalization and externalization ratings are the same for the most part.










TABLE 11
Correlations between Parental Social Competence
Ratings and Peer Ratings
(Growth Hormone Deficient Group Only)

Peers
Aggr- With- Like- Liked Liked
ession(n) drawal(n) ability(n) Most(n) Least(n)
Mother:
Activities:
-0.34 (19) -0.01 (19) -0.17 (19) -0.26 (26) -0.38 (24)

Social:
-0.09 (19) -0.38 (19) 0.07 (19) 0.24 (26) -0.34 (24)

School:
-0.18 (17) -0.50 (17)* 0.13 (17) 0.27 (24) -0.07 (22)

Total:
-0.29 (17) -0.54 (17)* 0.01 (17) 0.20 (24) -0.43 (22)*

Father:
Activities:
-0.29 (15) -0.26 (15) -0.28 (15) -0.17 (21) -0.43 (19)

Social:
0.04 (15) -0.53 (15)* 0.06 (15) 0.28 (21) -0.50 (19)*

School:
-0.12 (14) -0.47 (14) 0.50 (14) 0.28 (20) -0.26 (18)

Total:
-0.30 (14) -0.63 (14)* 0.06 (14) 0.26 (20) -0.49 (18)*


*=p<.05







73

Several differences do emerge in the correlations between each ratings by each source

on the separate behavior problems:

Mother-Teacher Correspondence

Mothers and teachers agree on aggressiveness in the GHD children but not the control

children (GHD: r=0.49,n=25 pairs,p<.05; control: r=-0.15, n=11 pairs, p>,05).

Mother-Child Corresoondence

In the total sample there is a significant correlation between mother-rated withdrawal

and child-rated self-esteem (r=-0.26,n=64 pairs, p<.05).

Mother-Peer Correspondence

In both the control group alone and in the total sample there are significant

relationships between ratings of aggression by mothers and peers (controls: r=0.55,n=22

pairs, p<.01; total sample: r=0.44,n=41 pairs, p<.01).

Father-Teacher Correspondence

There is a significant correlation between father and teacher rated withdrawal in the

control group (r=0.82,n=10 pairs, p<.01).

Father-Child Correspondence

In both the control group and total sample there are significant relationships between

father-rated aggression and child-rated self-esteem (controls: r=-0.51,n=26 pairs, p<.05;

total sample: r=-0.34,n=52 pairs, p<.05).

Father-Peer Correspondence

Father and peer rated withdrawal are significantly related in both the control group and

total sample (controls: r=0.51,n=21 pairs, p<.05; total sample: r=0.46,n=36 pairs, p<.01).

Teacher-Child Correspondence

In the control group teacher-rated withdrawal is significantly related to child-rated

self-esteem (r=0.55,n=13 pairs, p<.05).









Child-Peer Correspondence

Both the control group and total sample show significant relationships between

child-rated self-esteem and peer-rated withdrawal (controls: r=-0.41,n=24 pairs, p<.05;

total sample: r=-0.37,n=44 pairs, p<.05), likeability (controls: r=0.59,n=24 pairs, p<.01;

total sample: r=0.47,n=44 pairs, p<.01), and nominations of classmates liked the most

(controls: r=0.41,n=31 pairs, p<.05; total sample: r=0.30,n=58 pairs, p<.05).

There were also some changes in the patterns of relationship between ratings by each

source and child ratings on the six Piers-Harris Self-Concept scale factors:

1. Mother-rated aggression was significantly related to self-rated inappropriate

behavior in both the control group (r=-0.60,n=28 pairs, p<.01) and the total sample

(r=0.30,n=61 pairs, p<.05).

2. Father rated withdrawal was significantly related to self-rated poor physical

appearance in the total sample (r=-0.29,n=49 pairs, p<.05). Father-rated withdrawal was

also significantly related to self-rated lack of popularity in the control group

(r=-0.40,n=25 pairs, p<.05).

3. Father-rated aggression was significantly related to both self-rated inappropriate

behavior and lack of popularity in the control group (behav: r=-0.63,n=25 pairs, p<.01;

pop: r=-0.50,n=25 pairs, p<.05) and total sample (behav: r=-0.35,n=49 pairs, p,>05; pop:

r=-0.28,n=49 pairs, p<.05).

4. Peer-rated withdrawal was not significantly related to self-rated anxiety in either

the control group or total sample. It was significantly related to self-rated inappropriate

behavior in both the control group (r=-0.56,n=23 pairs, p<.01) and total sample

(r=-0.49,n=43 pairs, p<.01).







75

5. Peer-rated aggression was significantly related to self-rated anxiety in the total

sample (r=-0.48,n=43 pairs, p<.01) and lack of popularity in both the control group

(r=-0.52,n=23 pairs, p<.05) and the total sample (r=-0.36,n=43 pairs, p<.05).

6. Peer-rated likeability was significantly related to self-rated appropriate behavior

in both the control group (r=0.46,n=23 pairs, p<.05) and the total sample (r=0.42,n=43

pairs, p<.01), physical appearance in the total sample (r=0.38,n=43 pairs, p<.05), and

popularity in both the control group (r=0.53,n=23 pairs, p<.01) and total sample

(r=0.47,n=43 pairs, p<.01).

7. Peer nomination as a classmate liked the most was significantly related to three

scales in both the control group and total sample: intellectual and school status (controls:

r=0.52,n=30 pairs, p<.01; total sample: r=0.41,n=56 pairs, p<.01), physical appearance

(controls: r=0.37,n=30 pairs, p<.05; total sample: r=0.34,n=56 pairs, p<.05), and

popularity (controls: r=0.50,n=30 pairs, p<.01; total sample: r=0.39,n=56 pairs, p<.01).

8. While self-rated unhappiness and dissatisfaction was significantly related to peer

nomination as a classmate liked the least in the GHD group, this was not true for either the

control group alone or the total sample.


Relationships Between Measures of Adjustment and Current Perceptions of Height in the
Growth Hormone Deficient Children

Thirty-five GHD children completed Silhouette Apperception Test-Revised (SAT-R)

ratings of their current height, expectation for adult height, and expectation for height in one

year relative to same age and sex peers. Ratings on each of these dimensions were also

obtained for each GHD child by the physician's assistant (PA) who works closely with them.

The PA's ratings were subtracted from the child's ratings to obtain a difference score on each

dimension. This section discusses analyses designed to see whether ratings of adjustment vary

between GHD children who are realistic about their present height and those who overestimate.







76

Four children rated themselves as shorter than they actually are according to PA ratings.

These children are not included in these analyses. Of the 31 remaining children 11 rated their

height at the same percentile as did the PA. The other 20 children overestimated their height.

The two perception groups were not significantly different on PPVT scores

(t(27)=1.74, p>.05). However, the group who overestimated their height tended to receive

higher PPVT scores (realistic: X=87.5, s.d.=14.1; overestimators: X=99.6, s.d.=20.0).

Analyses used either ANOVA or MANOVA with the major factor being Perception at two

levels realistic (no different from PA) versus overestimated (rated self as taller than did

PA). The effects of age and sex were also investigated in each analysis and kept in the model

only where significant.

Child Self-Ratings

Social desirability was used as a covariate in these analyses. The first analysis compared

the two height perception groups on ratings of self-esteem. The two groups were significantly

different on this analysis (F(1,28)=4.9, p<.05). The children who had realistic perceptions

of their height had lower self-esteem as group than the children with overestimated

perceptions of their height. The mean self-esteem of the group who overestimated their height

was well above the mean self-esteem of the general population according to norms (Piers,

1969). Figure 19 illustrates this comparison.

The second analysis compared the two perception groups on separate scales of the

Piers-Harris Self-Concept scale. The overall MANOVA was significant

(F(6,21)=2.70,p<.05) taking social desirabilty into account. Separate ANOVAs indicated that

the groups rated themselves significantly differently on intellectual and school status

(F(1,26)=11.27,p<.01) and popularity (F(1,26)=11.08,p<.01). The children who

overestimated their height also rated themselves as having significantly higher intellectual



















Height Estimate=
Realistic


4W


Height Estimate=
Overestimated


(normative mean=655,.
Piers, 1969)


Self-Esteem








FIGURE 19. Mean self-esteem ratings by height perception.
(Actual means are depicted.)


Percent

5
Items

Endorsed







78

and school status and as being more popular than children who had more realistic perceptions

of their height. Figure 20 depicts these comparisons.

The second analysis compared the two perception groups on child ratings of behavior

problems on theYouth Self-Report Form of the Child Behavior Checklist. The two groups

were not significantly different on these ratings (F(1,18)=1.11, p>.05).

Ratings by Mothers

The first analysis compared the two height perception groups on maternal ratings of

withdrawal, aggression, activity competence, social competence, and school competence. The

overall MANOVA showed that the groups were not significantly different on these variables

(F(5,23)=0.84, p>.05).

On the second order factors of internalization and externalization the two perception

groups were not rated by their mothers as significantly different (F(2,28)=0.4, p>.05).

The final MANOVA compared mother ratings of the two perception groups on overall

behavior problems and overall social competence. The groups were not significantly different

in this analysis (F(2,25)=1.8, p>.05).

Ratings by Fathers

On the paternal ratings of withdrawal, aggression, activity competence, social

competence, and school competence the two height perception groups were not significantly

different (F(5,15)=0.53, p>.05).

The two perception groups were also not rated significantly different on the paternal

ratings of internalization and externalization (F(2,20)=0.4, p>.05).

The final analysis of father ratings compared the two perception groups on ratings of

overall behavior problems and overall social competence. Once again, the two groups were not

significantly different on these ratings (F(2,17)=0.1, p>.05).













Height Estimate=
Realistic


Height Estimate=
Overestimated


Behavior Intell &
School
Status


FIGURE 20.


Physical
Appear
& Attri-
butes


Anxiety


n 18 n= 18
ap at


5.




Popularity


Happiness .
Satisfact"Or


Mean ratings on each self-esteem factor
by height estimate.
(Actual means are depicted.)


Scores









Teacher Ratings

The first analysis of teacher ratings compared the two perception groups on ratings of

withdrawal and aggression. The two height perception groups were not rated as significantly

different on these variables by their teachers (F(2,19)=0.25,p>.05).

The second analysis of teacher ratings looked at the second-order factors internalization

and externalization. The two height perception groups were not rated as significantly different

on these variables by their teachers (F(2,20)=0.06,p>.05).

The analysis of teacher ratings of overall behavior problems indicated that the two

perception groups were not significantly different on these ratings (F(1,21)=0.0, p>.05).

The final analysis of teacher ratings compared the two perception groups on four

variables grades, effort, appropriateness of classroom behavior, amount of learning, and

happiness. The groups were not rated significantly different on these variables

(F(5,17)=0.12, p>.05). Ratings by teachers of a composite of these scores also did not

significantly differentiate the two height perception groups (F(1,21)=0.01,p>.05).

Peer Ratings

The first analysis of sociometric data compared the two height perception groups on the

PEI factors of withdrawal, aggression, and likeability. The groups were not rated significantly

different on these factors (F(3,13)=0.7, p>.05).

An analysis of classmates nominated as liked the most showed that the two perception

groups were not nominated at significantly different rates (F(1,21)=2.4, p>.05).

An analysis of classmates nominated as liked the least showed that the two perception

groups were not nominated at significantly different rates (F(1,19)=0.5, p>.05).










CHAPTER IV
DISCUSSION


The present investigation studied a group of 45 growth hormone deficient children who

were being treated with human growth hormone to increase the likelihood of linear growth.

Despite this treatment the height of the group as a whole was less than the fifth percentile for

age and sex. Overall the group was 8.15 inches below the mean for age and sex. Thus,

regardless of treatment these children were well below the average height expected for their

ages and sex.

An interesting issue involves how perceptions of the adjustment of these children vary

according to how tall the children perceive themselves to be. Twice as many of the growth

hormone deficient children overestimated their height relative to an objective standard than

endorsed accurate representations of their height. One might expect perceptions of the

children's adjustment to vary depending on how realistic they are about their own height. The

only sources who rated the children differently on various measures of adjustment were the

children themselves.

Only child self-esteem ratings were related to the children's perception of their height.

The children who over-estimated their height rated themselves as having significantly higher

intellectual and school status and as being more popular than children with more realistic

perceptions of their height (with social desirability accounted for). These children who

over-estimate their height could be over-estimating in all areas. They may be very skilled

at denying their limitations, whether physical, social, or in other areas. Conversely, the

growth hormone deficient children with realistic perceptions of their height may feel that

they are also unpopular and low in school status. It is inappropriate to draw a causal

conclusion. Nevertheless, one possible explanation for this relationship is that realistic








82

perceptions of short stature may result in poor self-esteem, particularly concerning

popularity and school status. Denial may be a healthy psychological response in this situation.

However, it is important to remember that only the children see themselves differently

depending on how tall they perceive themselves to be. It does not appear to make a difference

in how others see them.

Ratings of the self-esteem and behavioral adjustment of the growth hormone deficient

children were also compared to those of normal height children. These growth hormone

deficient children were rated as different than normal height children by each of the various

sources on a number of dimensions. The growth hormone deficient children rated themselves

as having lower self-esteem as a group than did the normal height children, even accounting

for a social desirability mindset. Means for both groups were above the normative mean.

In terms of behavior problems, growth hormone deficient and control adolescents with

higher Peabody Picture Vocabulary Test (PPVT) scores endorsed almost identical behavior

problem scores. At lower PPVT scores, however, the growth hormone deficient adolescents

indicated that they were experiencing twice the level of behavior problems (summed across

frequency and severity) than were the control adolescents. None of the other informants

(mothers, fathers, and teachers) indicated that the growth hormone deficient and control

groups were significantly different in overall behavior problems.

Social withdrawal and aggression are behavior problems that have been of particular

relevance in studying this group of children in the past. For example, Holmes, Hayford, and

Thompson (1982a) found their sample of growth hormone deficient boys to be more

withdrawn than normal height boys. Kusalic and Fortin (1975) showed that after 2 1/2

years of growth hormone replacement treatment the growth hormone deficient children they

followed had become verbally aggressive. Krims, in 1968, reported tht some short statured

children report intense unhappiness only after they begin to grow with growth hormone








83

treatment. A number of difficulties in accepting treatment have been discussed by Brust,

Ford, and Rimoin (1976). These difficulties include feelings of pressure to perform better.

Money and Pottitt (1966) termed this difficulty in accepting treatment a "readjustment

syndrome" caused by rapid change from disability toward normality.

In the present study, both mothers and peers rated the growth hormone deficient

children as significantly more withdrawn than the control children. However, fathers and

teachers did not. None of the sources rated the groups differently on aggressive behavior. A

more striking finding, however, is in the area of social competence. As assessed in this study

it involves the number of activities in which a child participates, the frequency of

participation and skill in the activity, the number of friends, frequency and type of contact

with those friends, and several school-related measures such as grades, type of class, and

problems in school.

The growth hormone deficient children in our sample were rated by both mothers and

fathers as being generally less socially competent than the normal height children. Mothers

indicated these differences in all three areas assessed competence in activities, social

interactions, and school functioning while fathers endorsed differences in two of the three

areas competence in activities and social interactions. In some cases the growth hormone

deficient children's average scores were so low they were close to those obtained by less than

2% of the same age and sex children in the normative population.

A number of the parental ratings of the growth hormone deficient children's social

competence were also significantly related to peer ratings of social withdrawal and

nominations as a peer liked the least. In those cases parental ratings of poor social competence

were related to either high levels of peer-rated social withdrawal or high numbers of

nominations as a classmate liked the least.








84

There are several possible explanations for these relationships. The growth hormone

deficient children may not be required to develop various age-appropriate social skills due to

their much younger physical appearance. A tendency toward juvenilization of the growth

hormone deficient child has been commented upon in several investigations (i.e., Rotnem et

al., 1977; Rotnem et al., 1979; Clopper et al., 1986). Another explanation is that the growth

hormone deficient child is excluded from age-appropriate activities because of his size. For

example, peers may not choose the growth hormone deficient child to be a teammate in sports

activities due to the child's short stature.

One last explanation for these relationships between social competence and withdrawal

is that the growth hormone deficient child may exclude himself from activities with peers. He

may not have developed age-appropriate competencies and so may stay to himself due to

embarrassment. For example, he may not know how to play ball and so may avoid

participating in activities involving this skill. Peers may interpret this avoidance as social

withdrawal. A different explanation is that the growth hormone deficient child is socially

competent, yet chooses not to interact with peers, possibly due to embarrassment because of

his short stature or having maintained friendships with only younger children.

In terms of the present findings, these social difficulties (withdrawal and lack of

competence in activities, social functioning, and school functioning) appear to be the major

problem areas faced by the growth hormone deficient children. These social problems were

reported by both mothers and fathers, as well as classmates. Therefore, we may assume that

these problems are evident both at home and at school.

Teacher ratings indicated that they did not view the growth hormone deficient children

as having more behavior problems, including social difficulties, than other children. The

structure of the usual classroom may provide a relatively safe social environment for the

growth hormone deficient child. The sample of students' behavior that teachers are exposed to








85

usually involves the students working quietly by themselves. Therefore, teachers may not

view the growth hormone deficient child's behavior as different than the norm based on this

rather limited sample of behavior. Quiet, withdrawn behavior is relatively adaptive in the

classroom. Social competence may not be a major issue there.

Classmates, however, have opportunities to interact with each other throughout the

school day. They have a broader sample of behavior from which to determine who is socially

withdrawn than do teachers. In the present study, not only did peers rate the growth hormone

deficient children as more withdrawn than normal height children, they also nominated them

significantly less often as a classmate like the most. While they were not particularly liked,

neither were they particularly unpopular. These findings suggest that the growth hormone

deficient child may be a socially neglected child.

Neglected children are usually rated as shy by their peers (Coie, Dodge, & Coppotelli,

1982). They do not necessarily exhibit more behavior problems than do average children

despite having few particular friends or enemies (French & Wass, 1985). Our results

indicate that the growth hormone deficient child fits this pattern. While he does not have more

behavior problems than the normal height child, he is less socially competent. His peers see

him as withdrawn and he is neither particularly liked nor disliked by them.

While teachers do not see growth hormone deficient children as particularly troubled

socially, they do see some difficulties with their classroom behavior. Specifically, teachers

rated them as having lower grades, putting forth less effort, learning less, and being less

happy than the normal height children. The literature to date does indicate that growth

hormone deficient children tend to do poorly academically (see Pollitt & Money, 1964; Siegel

& Hopwood, 1986; Holmes, Hayford, & Thompson, 1982b). Research has just begun to

investigate the particular types of academic problems these children evidence (Siegel &

Hopwood, 1986).








86
This study shows that different sources do not necessarily rate the behavior of the

growth hormone deficient child similarly. Of particular interest in this investigation were

mother, father, teacher, and peer ratings of withdrawal and aggression, problems noted in

growth hormone deficient children in the past (see Steinhausen & Stahnke, 1976, 1977;

Kusalic & Fortin, 1975). Mother and father ratings of withdrawal and aggression were

moderately and significantly related to each other as were mother and teacher ratings of

aggression and teacher and peer ratings of aggression. The growth hormone deficient children

were rated by their parents as generally poor in social competence, a condition which was

significantly related, as was previously stated, to peer ratings of withdrawal and nominations

as a peer liked the least.

It was expected that the agreement between sources would be different for the growth

hormone deficient and normal height children. One explanation for such differences would be a

tendency for all sources to see the behavior of the growth hormone deficient child in a

generally negative light, or"halo," due to his chronic medical condition. This study found that

there was generally more agreement between peer and parent ratings of the normal height

children than of the growth hormone deficient children. Sources were generally less in

agreement about the behavior of the growth hormone deficient children across settings than

they were of the normal height children.

It seems likely that the behavior of the growth hormone deficient child is relatively

different at home and at school. Teachers, in general, may not obtain a representative sample

of behavior other than that which is appropriate in the classroom. The growth hormone

deficient child may well be viewed as withdrawn by classmates, but parents and teachers may

not share this perception. Parents do, however, recognize their growth hormone deficient

children's deficits in social competence.








87

The present investigation does have several limitations that should be considered. The

sample sizes included are relatively small. This is largely due to the frequency with which

growth hormone deficiency occurrs. The psychosocial problems faced by these children are

nevertheless important and deserve investigation. A rather large number of statistical

analyses were conducted for this study which may increase the experiment-wise error rate.

Multivariate analyses were done where possible when logical associations between variables

existed. Nevertheless, statistical differences at the .05 probability level should be

interpreted with this issue in mind.

Another limitation involves the six factors of the Piers-Harris Self-Concept Scale.

Analyses which included these factors should be interpreted cautiously as there is no proof of

the factors' reliability or validity. Those analyses are provided as interesting additional

information for the reader. One final limitation involves the abscence of self-ratings of

aggression and withdrawal by the child. As of this writing there were no behavior problem

scales available for the adolescent-completed Youth Self-Report Form of the Child Behavior

Checklist. Those should be available with norms and self-rated social competence measures in

the near future.










BIBLIOGRAPHY


Abbott, D., Rotnem, D., Genel, M., & Cohen, D.J. (1982). Cognitive and emotional functioning
in hypopituitary short-statured children. Schizophrenia Bulletin, 8, 310-319.

Achenbach, T.M. (1979). The Child Behavior Profile: An empirically based system for
assessing children's problems and competencies. International Journal of Mental
Health,L 7 24-42.

Achenbach, T.M., & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and
Revised Child Behavior Profile. Burlington, VT: Dept. of Psychiatry, University of
Vermont.

Ad Hoc Committee on Growth Hormone Usage. (1983). Growth hormone in the treatment of
children with short stature. Pediatrics, 72. 891-894.

Apter, A., Galatzer, A., Beth-Halachmi, N., & Laron, Z. (1981). Self-image in adolescents
with delayed puberty and growth retardation. Journal of Youth and Adolescence, 10.
501-505.

Brown, G.M., Seggie, J.A. Chambers, J.W., & Ettigi, P.G. (1978). Psychoendocrinology and
growth hormone: A review. Psvchoendocrinoloav, 3. 131-153.

Brust, J.S., Ford, C.V., & Rimoin, D.L. (1976). Psychiatric aspects of dwarfism. American
Journal of Psychiatry, 133. 160-164.

Cacciaguerra, F. (1978). Research on some aspects of mental levels and their developmental
process in chondrodystrophic and hypopituitary dwarfism. Acta Medica Auxoloaica,10.
103-111.

Clopper, R.R., Adelson, J.M., & Money, J. (1976). Postpubertal psychosexual function in
male hypopituitarism without hypogonadotropinism after growth hormone therapy. The
Journal of Sex Research, 12. 14-32.

Clopper, R.R., MacGillivray, M.H., Mazur, T., Voorhess, M.L., & Mills, B.J. (1986).
Post-treatment follow-up of growth hormone deficient patients: Psychosocial status. In
B. Stabler & L. E. Underwood (Eds.), Slow grows the child (p. 83-96). Hillsdale, NJ:
Lawrence Erlbaum Associates.

Coie, J.D., Dodge, K.A., & Coppotelli, H. (1982). Dimensions and types of social status: A
cross-age perspective. Developmental Psychology, 18, 557-570.

Cox, S.H. (1966). Family background effects on personality development and social
acceptance. Unpublished doctoral dissertation, Texas Christian University.

Crandall, V., Crandall, C., & Katkovsky, W. (1965). A children's social desirability
questionnaire. Journal of Consulting Psychology, 29, 27-36.










Crowne, D.P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology, 24. 349-354.

Daughaday, W.H. (1974). Hypopituitarism. In R.H. Williams (Ed.), Textbook of endocrinology
(p. 55-64). Philadelphia: W.B. Saunders.

Dean, H.J., McTaggart, T.L., Fish, D.G., & Friesen, H.G. (1986). Long-term social follow-up
of growth hormone deficient adults treated with growth hormone during childhood. In B.
Stabler & L.E. Underwood (Eds.), Slow grows the child (p. 73-82). Hillsdale, NJ:
Lawrence Erlbaum Associates.

Drash, P.W. (1969). Psychologic counseling: Dwarfism. In L.I. Gardner (Ed.), Endocrine and
genetic diseases of childhood (p. 1014-1022). Philadelphia: W.B. Saunders.

Drash, P.W., Greenberg, N.E., & Money, J. (1968). Intelligence and personality in four
syndromes of dwarfism. In D.B. Cheek (Ed.), Human growth: Body composition, cell
growth, energy. and intelligence (p. 568-581). Philadelphia: Lee & Feliger.

Drotar, D., Owens, R., & Gotthold, J. (1980). Personality adjustment of children and
adolsecents with hypopituitarism. Child Psychiatry and Human Development, 11,
59-66.

Dunn, L.M., & Dunn, L.M. (1981). Peabody Picture Vocabulary Test-Revised: Manual for
forms L and M. Circle Pines, MN: American Guidance Service.

Frankel, J.J., & Laron, Z. (1968). Psychological aspects of pituitary insufficiency in
children and adolescents with special reference to growth hormone. Israel Journal of
Medical Science, 4 953-961.

French, D.C., & Waas, G.A. (1985). Behavior problems of peer-neglected and peer-rejected
elementary-age children: Parent and teacher perspectives. Child Development, 5.
246-252.

Gordon, MK., Crouthamel, C., Post, E.M., & Richman, R.A. (1982). Psychosocial aspects of
constitutional short stature: Social competence, behavior problems, self-esteem, and
family functioning. Journal of Pediatrics, 101. 477-480.

Grew, R.S., Stabler, B., Williams, R.W., & Underwood, L.E. (1983). Facilitating patient
understanding in the treatment of growth delay. Clinical Pediatrics, 22. 685-690.


Hollander, E.P. (1956). Interpersonal exposure time as a determinant of the predictive
utility of peer ratings. Psychological Reoorts, 2, 445-448.

Holmes, C.S., Hayford, J.T., & Thompson, R.G. (1982a). Personality and behavior differences
in groups of boys with short stature. Child Health Care, 11. 61-64.










Holmes, C.S., Hayford, J.T., & Thompson, R.G. (1982b). Parents' and teachers' differing
views of short children's behavior. Child: Care. Health and Development, L 327-336.

Holmes, C.S., Karlsson, J.A., & Thompson, R.G. (1986). Longitudinal evaluation of behavior
patterns in children with short stature. In B. Stabler & L.E. Underwood (Eds.), Slow
grows the child (1-12). Hillsdale, NJ: Lawrence Erlbaum Associates.

Holmes, C.S., Thompson, R.G., & Hayford, J.T. (1984). Factors related to grade retention in
children with short stature. Child: Care. Health and Development, 10. 199-210.

Kane, J.S., & Lawler, E.E. (1978). Methods of peer assessment. Psychological Bulletin, 85.
555-586.

Kaufman, A.S. (1979). Intelligence testing with the WISC-R. New York: Wiley.

Kaplan, S.A. (1975). Hypopituitarism. In L.I. Gardner (Ed.), Endocrine and genetic diseases of
childhood and adolescence (p. 106-126). Philadelphia: W.B. Saunders.

Krims, J.B. (1968). Observations of children who suffer from dwarfism. Psychiatric
Quarterly. 42, 430-443.

Kusalic, M.K., & Fortin, C. (1975). Growth hormone treatment in hypopituitary dwarfism:
Longitudinal psychological effects. Canadian Psychiatric Association Journal, 2.
325-331.

Lipsitt, L.P. (1958). A self-concept scale for children and its relation to the children's form
of the Manifest Anxiety Scale. Child Development, 29. 463-472.

Mayer, C.L. (1965). A study of the relationship of early special class placement and the
self-concepts of mentally handicapped children. Unpublished doctoral dissertation,
Syracuse University.

Meyer-Bahlburg, H.F.L., Feinman, J.A., MacGillvray, M.H., & Aceto, T.A. (1979). The
question of hormonal influences on intellectual development: Growth hormone.
Psychological Medicine, 187-189.

Millen, L. (1966). The relationship between self-concept, social desirability and anxiety in
children. Unpublished masters' thesis, the Pennsylvania State University.

Mitchell, C.M., Johanson, A.J., Joyce, S., Libber, S., Plotnick, L., Migeon, C.J., & Blizzard,
R.M. (1986). Psychosocial impact of long-term growth hormone therapy. In B.
Stabler & L.E. Underwood (Eds.), Slow grows the child (p. 97-110). Hillsdale, NJ:
Lawrence Erlbaum Associates.

Money, J., & Clopper, R.R. (1975). Postpubertal psychosexual function in post-surgical
male hypopituitarism. The Journal of Sex Research, 11. 25-38.










Money, J., Clopper, R., & Menefee, J. (1980). Psychosexual development in postpubertal
males with idiopathic panhypopituitarism. The Journal of Sex Research, 16. 212-225.

Money, J., Drash, A.W., & Lewis, V. (1967). Dwarfism and hypopituitarism: Statural
retardation without mental retardation. American Journal of Mental Deficiency, 74.
122-126.

Money, J., & Pollitt, E. (1966). Studies in the psychology of dwarfism. II.: Personality
maturation and response to growth hormone treatment in hypopituitary dwarfs. Journal
of Pediatrics, 68. 381-390.

Obuchowski, K., Zienkiewicz, A., & Graczkowska-Koczorowska, A. (1970). Psychological
studies in pituitary dwarfism. Polish Medical Journal, 1229-1235.

Parkin, M. (1974). Incidence of growth hormone deficiency. Archives of Disease in Childhood,
49,. 904-905.

Pekarik, E.G., Prinz, R.J., Liebert, D.E., Weintraub, S., & Neale, J.M. (1976). The Pupil
Evaluation Inventory: A sociometric technique for assessing children's social behavior.
Journal of Abnormal Child Psycholoy, 4. 83-97.

Piers, E.V. (1969). Manual for the Piers-Harris Children's Self-Concept Scale. Nashville,
TN: Counselor Recordings and Tests.

Pollitt, E., & Money, J. (1964). Studies in the psychology of dwarfism. I: Intelligence quotient
and school achievement. Journal of Pediatrics, 64. 415-421.

Powell, M. (1948). Comparisons of self-ratings, peer ratings, and expert ratings of
personality adjustment. Educational and Psychological Measurement, 8. 225-234.

Quay, H.C., & Peterson, D.R. (1979). Manual for the Behavior Problem Checklist. Miami, FL:
University of Miami.

Rosenbloom, A.L. Smith, D.W., & Loeb, D.G. (1966). Scholastic performance of
short-statured children with hypopituitarism. Journal of Pediatrics, 69. 1131-1133.

Rotnem, D., Cohen, D.H., Hintz, R., & Genel, M. (1979). Psychological sequelae of relative
"treatment failure" for children receiving growth hormone replacement. Journal of the
American Academy of Child Psychiatry, 18. 505-520.

Rotnem, D., Genel, M., Hintz, R.L., & Cohen, D.J. (1977). Personality development in
children with growth hormone deficiency. Journal of the American Academy of Child
Psychiatry, 16. 412-426.

Schaff-Blass, E., Burstein, S., & Rosenfield, R.L. (1984). Advances in diagnosis and
treatment of short stature with special reference to the role of growth hormone. Journal
of Pediatrics, 14.. 801-813.










Shizume, K. (1984). Long-term effects of human growth hormone on 1,959 patients with
pituitary dwarfism throughout Japan. Endocrinologia Japonica, 31. 201-206.

Shurka, E., Galatzer, A., & Baizerman, M. (1983). The self-concept of growth retarded
children, adolescents and youth: An exploratory study. International Journal of Eclectic
Psychotherapy, 2 21-35.

Siegel, P.T., & Hopwood, N.J. (1986). The relationship of academic achievement and the
intellectual functioning and affective conditions of hypopituitary children. In B. Stabler
& L.E. Underwood (Eds.), Slow grows the child (p. 57-72). Hillsdale, NJ: Lawrence
Erlbaum Associates.

Sines, J.O., Parker, J.D., & Sines, L.K. (1971). The Missouri Children's Picture Series Test
manual. Iowa City: J.O. Sines, University of Iowa.

Soyka, L.F., Bode, H.H., Crawford, J.D., & Flynn, F.J. (1970). Effectiveness of long-term
human growth hormone therapy for short stature in children with growth hormone
deficiency. Journal of Clinical Endocrinology, a. 1-14.

Spencer, R.F., & Raft, D.D. (1974). Adaptation and defenses in hupopituitary dwarfs.
Psvchosomatics, 15. 35-38.

Steinhausen, H., & Stahnke, N. (1976). Psycho-endocrinological studies in dwarfed children
and adolescents. Archives of Diseases in Childhood, 51, 778-783.

Steinhausen, H., & Stahnke, N. (1977). Negative impact of growth-hormone deficiency on
psychological functioning in dwarfed children and adolescents. European Journal of
Pediatrics, 126. 263-270.

Ullman, C.A. (1952). Identification of maladjusted school children. Washington: U.S. Public
Health Service.

Vimpani, G.V., Vimpani, A.F., Pocock, S.J., & Farquhar, J.W. (1981). Differences in physical
characteristics, perinatal histories and social backgrounds between children with
growth hormone deficiency and constitutional short stature. Archives of Diseases in
Childhood, 56 922-928.

Wechsler, D. (1974). Wechsler Intelligence Scale for Children-Revised. New York:
Psychological Corporation.

Wing, S.W. (1966). A study of children whose reported self-concept differs from classmates'
evaluation of them. Unpublished doctoral dissertation, University of Oregon.

Young-Hyman, D. (1986). Effects of short stature on social competence. In B. Stabler & L.E.
Underwood (Eds.), Slow grows the child (p.27-45). Hillsdale, NJ: Lawrence Erlbaum
Associates.











BIOGRAPHICAL SKETCH


Carol Lewis was born on December 25, 1959, in New Orleans, Lousiana, where she grew up.

In 1981 she received the Bachelor of Arts in psychology from Wake Forest University in

Winston-Salem, North Carolina. While there she was elected to Phi Beta Kappa and graduated

summa cum laude. She received the Master of Science degree in clinical psychology from the

University of Florida in 1983.








I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.



Suzanne Bennett Johnson, Chairman
Associate Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.




Nathan Perry
Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.



Sheila Eyberg
Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.



Jam s H. Johnson
Ass iate Professorl Clinical Psychology

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.




Carole Kimberlin
Associate Professor of Pharmacy