The effect of attachment, attributions, maternal age, previous fetal loss and number of children on grief following spontaneous abortion

THE EFFECT OF ATTACHMENT, ATTRIBUTIONS, MATERNAL AGE, PREVIOUS FETAL LOSS, AND NUMBER OF CHILDREN ON GRIEF FOLLOWING SPONTANEOUS ABORTION A dissertation submitted to the Kent State Unive

IN FO R M A TIO N TO U SER S

This manuscript has been reproduced from the microfilm master UMI films the text directly from the original or copy submitted Thus, some thesis and dissertation copies are in typewriter face, while others may

be from any type of computer printer.

The quality of this reproduction is dependent upon the quality o f the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event th at the author did not send UMI a com plete manuscript and there are missing pages, these will be noted Also, if unauthorized copyright m aterial had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand com er and continuing from left to right in equal sections with small overlaps Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge Contact UMI directly

to order.

A Bell & Howell Information Com pany

THE EFFECT OF ATTACHMENT, ATTRIBUTIONS, MATERNAL AGE, PREVIOUS FETAL LOSS, AND NUMBER OF CHILDREN ON GRIEF

FOLLOWING SPONTANEOUS ABORTION

A dissertation submitted to the Kent State University Graduate School o f Education

in partial fulfillment o f the requirements for the degree of Doctor o f Philosophy

by

Rebecca Johnson Heikkinen

May, 1995

UMI Microform 9536635 Copyright 1995, by UMI Company All rights reserved.

This microform edition is protected against unauthorized

copying under Title 17, United States Code.

UMI

300 North Zeeb Road Ann Arbor, MI 48103

Dissertation written by

Rebecca Johnson Heikkinen

B.A., Wittenberg University, 1974

M.Ed., Kent State University, 1987

Ph.D., Kent State University, 1995

A -^ ^ 2 f*^D irector, Doctoral Dissertation

Dean, Graduate School of Education

ii

I wish to extend grateful acknowledgment to my dissertation committee,

including Dr Claire Draucker who offered both encouragement and guidance with

editing; Dr Dan Sanders who gave patient support; and especially Dr Tom Dowd for

his motivation and generous gift o f both time and mentorship All o f the above

individuals were instrumental in helping me to finish the long process leading up to

this final milestone in my doctoral studies

I am grateful to my friends and colleagues, but most especially to my family

for their acceptance o f the frequent disruptions caused by my extended tenure as a

student I am especially thankful for the on-going support afforded by my husband,

Carl, who has carried many additional responsibilities in the family during my seven

years o f graduate work I would also like to thank my two sons, Matthew and Eric for

their patience and good humor at times when my studies caused me to be unavailable

to them

Finally, I wish to thank my parents, Nils and Janet Johnson, whose support for

the pursuing of one’s dreams led to the vision o f the possibility o f obtaining a

doctorate mid-life

iii

TABLE OF CONTENTS

Page

ACKNOW LEDGEM ENTS iii

LIST OF T A B L E S vi

Chapter I THE PROBLEM 1

Introduction 1

Description of the P r o b le m 3

Psychological Repercussions of Spontaneous A bortion 3

A ttach m en t 7

Causal Attributions 10

Causal Attributions and Spontaneous Abortion 13

Previous Fetal L o s s 14

Statement of the Problem 15

Operational Definitions 16

Scope of the S tu d y 17

II REVIEW OF THE LITER A TU R E 19

Review of the Literature on the Problem Spontaneous A bo rtion 19

Definition of a Spontaneous Abortion 20

Psychological Repercussions of Spontaneous A b o rtio n 21

A tta c h m e n t 24

Attributional T h e o ry 30

Spontaneous Abortion and Attributions 34

Grief 41

III M ETH O D O LO G Y 45

Description of the S a m p le 45

In stru m en ts 48

Pregnancy Loss Attributional Questionnaire-Predictor Variable 48

Maternal-Fetal Attachment Scale-Predictor V ariable 49

Perinatal Grief Scale-Dependent V a ria b le 53

Design 56

Procedures 57

Statistical T rea tm e n t 59

iv

Results for Hypothesis I 65

Results for Hypothesis II 65

Results for Hypothesis I E 66

Results for Hypothesis I V 66

Results for Hypothesis V 66

V SUMMARY, CONCLUSIONS, AND RECOM M ENDATIONS 67

Summary 67

P urpose 67

M eth o d 69

Results and D isc u ssio n 70

Hypothesis I 70

Hypothesis II 74

Hypothesis I E 76

Hypothesis I V 79

Hypothesis V 81

Limitations 82

Conclusions and Recommendations for Future R e s e a rc h 84

APPENDDCES 88

REFERENCES 100

v

3 Sample Items and Alpha Coefficients for Grief Subscales 59

4 Descriptive Statistics for all Measures 65

7 Stepwise Multiple Regression Statistics 69

vi

THE PROBLEM

Introduction

When a pregnancy ends in a spontaneous abortion, the mother typically experiences

grief similar to the grief experienced after the loss o f a significant other (Kennel,

Slyter, & Klaus, 1970; Peppers & Knapp, 1980) The level o f grief following

spontaneous abortion is most likely influenced by a variety of variables, although few

have been identified in the literature By being able to isolate factors that may

influence the level o f grief experienced, individuals who are most at risk for

experiencing pathological grief reactions can be identified and given support soon after

the loss in an attempt to resolve grief expeditiously

The World Health Organization (1970) defined spontaneous abortion as a

spontaneous fetal loss occurring between conception and twenty-eight weeks o f

gestation The term "miscarriage" is synonymous with spontaneous abortion Buehler

(1983) stated the percentage o f pregnant women in the United States who realize they

are pregnant and then lose the fetus is about 10%-14%, while Cavanagh and Comas

(1982) placed the rate at 20%

"Attachment" is an emotional bond which is established between mother and fetus

1

and can be measured in terms o f the behaviors the mother engages in which represent

interaction and affiliation with the fetus (Kemp & Page, 1987) Although attachment

was originally believed to occur after birth, it is now described as one of the

developmental stages that occurs during pregnancy (Cranley, 1981)

"For five months or longer (the mother) has had a physical and kinesthetic

awareness of the fetus, and for even longer she has had intellectual knowledge of her

child" (Cranley, 1981, p 281) Attachment has been found to be related to the

intensity o f emotion experienced after miscarriage (Madden, 1986)

The causal explanations that people place on life events have often been found to

influence their psychological adjustment and ability to cope following those life events

(Abramson, Seligman, & Teasdale, 1978; Bulman & Wortman, 1977; Janoff-Bulman,

1979; Seligman, 1975; Silver & Wortman, 1980; Wortman, 1976; Wortman & Brehm,

1975) O f particular interest are the explanations that people provide following a

negative life event, especially when this event is in reality beyond their control Even

in situations where it is evident that a change o f actions would not have resulted in a

different outcome, it is not unusual for people to attempt to find a causal link

Attribution theory suggests that our cognitions, expectations, and actions are based on

a mastery of the causal network of the environment A "causal attribution" is an

assignment o f perceived causation to one or more factors in an attempt to understand

events that one observes around oneself, including one’s own actions or the actions

taken by others (Harvey & Smith, 1977) Unlike a prediction, an attribution is made

following an event, although it may influence future events Causal attributions are

different causal attributions lead to different emotions (Weiner et al., 1971) One

negative life event that is usually beyond the control o f the persons involved is a

spontaneous abortion

When a woman plans a pregnancy, she may or may not think about the level o f

control she may have on its eventual outcome Once a woman experiences a fetal

loss, whether it is caused by a spontaneous abortion, a stillbirth, or a neonatal death,

she is apt to pay more attention to this issue As the number o f losses mount, it is

probable that a woman feels less able to control the outcome of any future pregnancy

In this study, the emotional aftermath o f spontaneous abortion will be viewed in

terms o f level o f grief "Grief' will be considered in terms of active grief, difficulty

coping and despair as measured by these factors on the Perinatal Grief Scale

Description o f the Problem

Psychological Repercussions o f Spontaneous Abortion

Until recent years, spontaneous abortion, which has been estimated to occur in

about one-fourth o f all pregnancies, has been almost neglected as a subject of social

science research Since 1969 a few investigations have been conducted, but for the

most part they have been lacking in methodological sophistication, have been filled

with statistical errors and design insufficiencies, and have been based on very little

data (Kirkley-Best & Kellner, 1982; Peppers & Knapp, 1980a) Reinharz (1988)

commented that "the most striking aspects o f the experiential literature on miscarriage

are its sparseness and its persistent complaints about the silence with which

miscarriage is surrounded" (p 85) The literature on spontaneous abortion is largely

unsystematic and is based for the most part on case studies or on very small-scale

studies usually involving married, middle-class couples (e.g., Drotar & Irevin, 1979;

Turco, 1981) Both types of literature have reported emotional trauma and often

intense depression on the part o f these women Many authors have stated that this

depression has not been supported by others, either in the lay public or by

professionals More complete information about the psychological repercussions o f this

type o f perinatal loss is needed in order to provide support for victims (Brody, 1980;

Holland, 1982; Jimenez, 1982; Pizer & Palinski, 1980)

The studies that venture beyond the case-study approach frequently use

unstructured interviews to assess the level o f grief (e.g., Rowe et al., 1978; Wolff,

Nielson, & Schiller, 1970) A few researchers have been more systematic and have

used the work of Kennell, Slyter, and Klaus (1970) to measure perinatal grief, while

including other variables o f interest In 1988, Toedter, Lasker, and Alhadeff

developed the Perinatal Grief Scale, providing the first reliable and well-validated

measure of this grief construct

The studies that have looked at perinatal grief indicate that maternal grief is very

similar to the reaction o f grief generally experienced with the death o f an older loved

one (Kennell et al., 1970; Peppers & Knapp, 1980) However, Leon (1987) listed six

reasons that perinatal loss is particularly difficult for women The first is related to

the narcissistic nature o f the loss experience Losing a fetus is like losing a part o f

(Furman, 1978) In order to mourn a perinatal loss, the fetus must be viewed as a

separate individual who has died The second reason is related to the self-blame that

is typically felt by the mother as she looks for a cause in her own behavior during the

pregnancy This will be discussed in more detail in the section dealing with

attributions

The third reason is that with perinatal loss there is usually no opportunity to

anticipate the death, which makes the shock much greater Parkes (1975) stated that

when advance warning is short and the death is sudden a much greater impact is felt

than with deaths that take place following a warning with the life terminating more

gradually Sudden death results in greater and more long-lasting disorganization in the

life o f the survivor The speed o f change has been shown to compound the effects

stemming from negative life events (Lauer, 1974) With perinatal loss a woman’s

status can often change from pregnant to no-longer pregnant in one day Pizer and

Palinski (1980) agreed that the stark contrast o f being in the process o f fulfilling a

dream one day and experiencing a devastating loss the next day is a reason

spontaneous abortion is so difficult to handle

The fourth reason Leon (1987) listed is a lack o f concrete memories or objects with

which to remember the child The fifth reason is the prospective nature of the grief;

that is, fantasies o f future interactions that might have taken place with the child

need to be mourned Finally, as already mentioned, there is a lack of social support

by the medical profession and the community at large

Hutti (1992) stated that participants in her study o f miscarriage had no schema

model, or set o f expectations, for what a miscarriage is like because they had no prior

miscarriage experience personally or from anyone close to them Because the

experience in not generally openly discussed, the women did not have a framework to

view their own experience from, and therefore stated that they did not know what to

expect o f themselves or from the experience

Developmentally, Rubin (1975) described pregnancy as a time o f heightened

sensory perceptivity The pregnant woman turns inward with the realization that

others cannot share these sensory experiences This leads to a heightened sense of

uniqueness and estrangement, which can serve to isolate the pregnant woman This

isolation is exacerbated when the pregnancy is suddenly ended due to a spontaneous

abortion, often leading to feelings that no one can really understand Another

difficulty in dealing with a spontaneous abortion is that traditionally there has been no

ritual that accompanies the loss Burr, Leigh, Day, and Constantine (1979) talked

about the high levels o f family disorganization that can result when there are no

specific rituals used to help people make transitions Kohn and Moffitt (1992) and

Broner (1982) advocated the use o f rituals to symbolize the loss o f pregnancy

Beil (1992) pointed out that since so many pregnancies end in miscarriage, many

women who are in therapy will have been through this type o f loss Reproductive-

related events may well be related to depression in women, but the biopsychosocial

factors which may contribute to this depression are poorly understood by both

therapists and the women themselves The women themselves may be unaware of the

Few empirical studies have examined the psycho-social aspects o f spontaneous

abortion As Beil (1992) stated, "the study o f the internal experience o f miscarriage

has been largely neglected" (p 62) Variables that have been studied include the

relationship between grief following spontaneous abortion and future incidence of

spontaneous abortion (Dunbar, 1963; Peppers & Knapp, 1980), pregnancy used as a

grief resolution strategy (Horowitz, 1979), and the incidence o f pathological grief

following spontaneous abortion (Stack, 1980, 1984) More recently, Madden (1986)

has studied the emotional aftermath o f spontaneous abortion She found that

respondents remembered mainly negative reactions right after the spontaneous

abortion, but often felt more positive four months later

Attachment

"Attachment" is defined as the affectional tie that develops between a mother and

her child from conception on, causing the mother to maintain proximity to that child

following birth Attachment theory suggests that this response is programmed in the

species in order to ensure the survival o f the infant by establishing proximity of

mother and child in case o f threat or danger (Ainsworth, Blehar, Waters, & Wall,

1978; Bowlby, 1969)

Attachment that develops over the first two years of a child’s life has been studied

as it relates to various aspects of developmental functioning Because it has variously

correlated with the child’s exploration and problem-solving ability, curiosity,

sociability, and control, it has been seen as a critical developmental issue (Egeland &

Farber, 1984)

Recently, attachment has been viewed as something that may develop prior to the

birth o f the child Psychoanalysts have suggested that the mother-infant bond is

initiated with intrauterine movements (stated in Klaus & Kennel, 1982), but Cranley

(1981) suggested that this bond develops even earlier for the mother, sometimes from

the time she first becomes aware of her pregnancy Cranley believed this attachment is

the result o f both psychological and physiological events Kemp and Page (1987)

defined this prenatal attachment as "the extent to which the woman engages in

behaviors that represent affiliation and interaction with her unborn fetus" (p 179)

Developmental theory provides a basic framework for understanding attachment in

terms o f the tasks or stages o f pregnancy A developmental task is defined as "the

growth responsibility that arises at a certain time in the course of development,

successful achievement o f which leads to satisfaction and success with later tasks"

(Kemp & Page, 1987) Rubin (1975) described pregnancy as a period o f identity

reformulation and personality maturation with tasks to be completed at different levels

Three o f the major tasks o f pregnancy that a woman must accomplish include:

1 Seeking safe passage from conception through delivery for herself and her baby; 2

Attempting to secure the infant’s acceptance by significant family members; 3

Attaching or binding-in with her child

Rubin (1975) stated that during the first trimester, the woman’s method o f seeking

safe passage relates more to herself than the baby Her only early clue concerning her

pregnant status is a non-event, namely amenorrhea This can signify the presence of

represents a better alternative to the woman It is better to be pregnant than to have

cancer or another serious medical problem Rubin felt that attachment to the child

develops after the first trimester, when tactile, visual, and kinesthetic cues first appear

It is the attachment to and awareness o f the child that causes the woman to seek

prenatal care in an attempt to protect the child from being damaged This help can

take many forms including media (books, magazines, television), medical

professionals, or laywomen considered experts in childbearing

Along with seeking safe passage for herself and her baby, a woman has the task of

securing acceptance o f that child by significant others Especially during the first

trimester, the woman as childbearer must assure not only for physical accommodation

of the child into the family, but also for the psychosocial accommodation o f that child

Securing acceptance involves loosening relationship bonds o f intimacy and

exclusiveness and realigning these bonds in preparation for the addition o f a child

The acceptance process begins during pregnancy as a conceptual one If acceptance is

conditional on the child’s sex or health status, the acceptance is not considered

complete "Conditional acceptance involves implicit rejection" (Rubin, 1975 p 148)

The rejection o f the child is tied to a rejection o f the self Cranley (1981) believed

that the development o f the mother-fetal relationship is integral to the consideration of

both the woman’s identity and the identity o f the developing fetus

Leifer (1977) stated that maternal involvement with the fetus can be seen in

prenatal attachment behaviors, such as talking to the fetus, calling it by a pet name or

maneuvering the fetus in an effort to allow the husband to observe the movement

These behaviors are evidence o f what Rubin (1975) termed binding-in or attaching to

the fetus as person, the third developmental task of pregnancy It is not known exactly

when these attachment behaviors begin or whether the level o f attachment effects the

level o f grief experienced following the severing of this attachment via spontaneous

abortion

Causal Attributions

Recently there has been much interest in understanding how people cope with

negative life events, and in particular, in determining what factors place individuals "at

risk" for negative psychological outcomes Seligman and his associates’ research

dealing with learned helplessness and depression (Abramson, Seligman, & Teasdale,

1978; Seligman, 1975), as well as the work o f Wortman and Janoff-Bulman on coping

with uncontrollable negative life events (Bulman & Wortman, 1977; Janoff-Bulman,

1979; Silver & Wortman, 1980; Wortman, 1976; Wortman & Brehm, 1975), have

been influential in this sphere

Peterson and Seligman (1984) stated that people have characteristic explanatory

styles If reality is ambiguous enough, a person will project and impose his habitual

style in the form o f an explanation or attribution for the event There are two things

that influence what particular explanation is chosen The first factor is the reality of

the bad event, which is the environmental stressor in a diathesis-stress model The

second factor is that o f explanatory style, which is sometimes seen as a risk factor, but

will only come into play in the event o f some precursor

and Abramson and Martin (1981) in terms of predicting depression Other theorists

have noted the tendency o f groups o f people to place blame on themselves for negative

life events that are in actuality beyond their control Examples of such groups have

included concentration camp prisoners (Bettelheim, 1943), parents o f children with

leukemia (Chodoff, Friedman, & Hamburg, 1964), rape victims (Burgess &

Holmstrom, 1974a, 1974b, 1976), and victims of freak accidents (Bulman & Wortman,

1977: see Wortman, 1976, for a review)

There has been a controversy among authors as to whether this tendency for self­

blame has played an adaptive or maladaptive role in terms of the coping process

Beck (1967) viewed self-blame as a maladaptive and self-deprecating symptom of

depression; Abrams and Finesinger (1953), stated that self blame was maladaptive;

Weisman (1976) also found self-blame to be a counter-productive response

From the other perspective, self-blame is seen as an adaptive coping response as

the assignment o f self-blame implies a sense o f personal control over outcomes

(Janoff-Bulman, 1979) Perceived control over one’s outcomes is thought to be related

to adjustment and health

Janoff-Bulman (1979), in an attempt to reconcile these different findings, proposed

two distinguishable types of self-blame; self-behavior blame, an attribution to a

controllable or modifiable aspect of the self; and self-character blame, an attribution to

uncontrollable aspects o f the self Because o f the lack of control postulated over

characterological blame, it is seen as a self-deprecating response Abramson and his

associates, in their reformulated model of learned helplessness, go beyond a simple

internal-external attributional distinction They add two additional dimensions: the

degree to which attributions are both stable versus unstable and specific versus global

(Abramson, Seligman, & Teasdale, 1978) Both models agree on the importance o f

distinguishing attributions that are characterological (internal, stable, and global) from

those that are behavioral (internal, unstable, and specific)

I f the characteristic explanatory style attributes internal, stable and global causes

(characterological attributions) the person tends to become depressed or to cope more

poorly when bad events occur because of the decrease in self-esteem and lack o f

control that exists with this attributional style Therefore, those attributions involving

characterological blame should be negatively related to coping but positively related to

depression Those that make behavioral attributions assume the ability to exercise

some control and make changes to prevent reoccurrence Therefore, a behavioral

attributional style should be positively related to coping but negatively related to

depression

External attributions imply causation due to an outside force, which may be stable

or unstable, and specific or global This type o f attribution results in the perception of

lack o f control However, what distinguishes between this type o f attributional style

and either type o f internal attributional style is the exemption o f personal

responsibility When an external source is found to be accountable for a negative life

event, one is personally exonerated from causation, and there is not a decrease in self­

esteem as is found with characterological attributions Therefore, making an external

attribution should be positively related to coping and negatively related to depression.

Causal Attributions and Spontaneous Abortion

Spontaneous abortion o r miscarriage can be viewed as a negative life event that is

normally out o f the control of the woman experiencing the event Cavanagh and

Comas (1982) stated one reason spontaneous abortions in particular can be so

traumatic is that a cause is not often identifiable Uterine infection or chromosomal

abnormalities can only rarely be isolated as the cause o f a spontaneous abortion This

ambiguity leaves the door open for the woman to attribute any cause she may choose

for the loss

It is not unusual for an attribution to take the form o f self-blame (internal

characterological or internal behavioral attributions), because the culture emphasizes

the need for pregnant women to be responsible for their unborn children’s health

Especially in recent years, both the medical profession and the culture at large have

emphasized that a woman’s behavior during pregnancy has the potential to impact her

baby’s health Warnings about smoking and drugs and admonitions to eat well and

exercise moderately are examples of this With the experience o f a spontaneous

abortion it becomes difficult to reconcile this idea o f responsibility for the health of

the fetus with the fact that some losses cannot be prevented with even the best prenatal

care (Reinharz, 1988)

There are two common myths that may affect the attribution made by the victim

(Stack, 1984) One is the idea that either physical exertion or an accidental injury are

common causes o f the loss of a fetus There is no medical evidence linking physical

activity or traum a and spontaneous abortions (Cavanagh & Comas, 1982) Although

most physicians impart this information to their patients (Stack, 1984), many women

still attribute causation to something they did (e.g., Pizer & Palinski, 1980)

The second myth is that anxiety can cause spontaneous abortion Although there

are psychoanalytic clinical reports that claim to identify psychological causes of

spontaneous abortion, there has not been such a link from larger, more systematic

studies Despite the reassurance of doctors, some women continue to think that the

fact that they are nervous or under a lot o f strain might have contributed to their loss

(Madden, 1988; Pizer & Palinski, 1980)

Previous Fetal Loss

Control over future losses is one reason why it is important for individuals to make

a causal attribution If one can attribute the cause to something that is modifiable, the

individual believes that chances are the event will not be repeated once a specific

change is made Control is also important in terms of the number o f losses a woman

has experienced in the past and her confidence about being able to successfully

accomplish pregnancy and birth o f a healthy baby in the future With a second or

third loss, the woman begins to look for variables that were the same during every

pregnancy I f the first loss was thought to occur due to a modifiable factor and that

factor is changed, yet a subsequent pregnancy also ends in spontaneous abortion, it

should become more difficult to believe future losses will be controlled One attempts

to find commonalities between the losses and place a new cause o f blame that fits both

instances, but control seems less possible as the number o f losses mount For this

reason it is thought that level o f grief will be higher for those who have experienced

more than one fetal loss (spontaneous abortion, stillbirth or neonatal death)

Statement of the Problem

It is important to be able to identify who might be at risk for negative

psychological consequences following a spontaneous abortion so that interventions can

be implemented early in the grief process It is proposed that higher levels of

attachment to the fetus will increase the level of grief "Attachment" is defined as an

affectiona! tie that helps to bind individuals together Thus it is felt that the more tied

to the fetus one has become, the more difficult the loss of that "individual" will be

Further, it is proposed that an attribution for the spontaneous abortion that is assigned

as an aspect o f one’s character as opposed to being attributed to either one’s behavior

or some external cause, would have a negative effect on an individual’s level o f grief

following the loss of a pregnancy

This study will look at maternal attachment to the fetus during pregnancy,

attributions o f the cause o f a spontaneous abortion, number o f previous fetal losses,

number o f living children, and maternal age as predictors o f level o f grief following

spontaneous abortion for previously expectant females

From previous studies that have been done it is possible to theorize several factors

that might have an impact on level of grief following spontaneous abortion The

following hypotheses were addressed in this study:

1 Level o f grief will be significantly impacted by level o f attachment

2 Level o f grief will be significantly impacted by type o f attribution made.

3 Level o f grief will be significantly impacted by the number o f previous fetal

1 Maternal attachment: The mean score on Maternal Fetal Attachment scale

The greater the sum, the greater is the inferred attachment to the lost fetus

2 Attribution of cause o f the spontaneous abortion: The raw score total o f

items answered in the keyed direction on the Pregnancy Loss Attributional

Questionnaire subscales (External, Behavioral, and Characterological

attributions)

3 Number o f Fetal Losses: The total number o f losses from spontaneous

abortion, stillbirth, or neonatal death experienced previous to the current

miscarriage by the previously expectant mother

4 Number o f Living Children: The number o f surviving children who were

born to or adopted by the subject

5 Maternal Age: The self-reported age o f the subject

6 Perinatal grief: The total score on the Perinatal Grief Scale The larger the

score, the greater is the inferred level of grief

Scope o f the Study

This study is limited to an examination of maternal attachment to the fetus, the

attribution o f cause o f the spontaneous abortion, number of previous fetal losses,

number of living children, and maternal age as predictors of level of grief for females

who have spontaneously aborted

Attachment was previously seen as beginning following the birth o f a baby as an

inborn trait that served the purpose of keeping parent and child close to each other for

the sake of protection It has been shown that most women become attached to the

fetus prior to birth, at some time during the pregnancy Since this attachment occurs

at different stages during the pregnancy and develops at different levels, it is

hypothesized that this difference may affect level o f grief Similarly, since

people are known to spontaneously make attributions for the cause o f negative life

events, it is postulated that the type of attribution made for the spontaneous abortion

may affect level o f grief The issue of control over future losses is important in terms

o f the number o f losses one has experienced Women who have experienced more

than one loss are likely to experience greater levels of grief because they may have

attempted to change behaviors after an initial loss, yet experienced the same outcome

It is hypothesized that they will feel less control following subsequent losses, and

therefore should exhibit higher levels o f grief

Women who already have at least one child have a history o f successfully

completing a pregnancy in the past Besides knowing that they have been able to

conceive and carry a pregnancy to term, they also have the consolation o f knowing

that they already have a child This should impact the level o f grief that they

experience following spontaneous abortion

Older women should be aware that it is more difficult to become pregnant as they

age, since both men and women are most fertile in their mid-twenties Fertility for

women starts to decline slowly after the mid-twenties, until at about age thirty-five, it

begins to decline rapidly Younger women have statistically better chances o f

conceiving Therefore, it is felt that there will be a difference in level o f grief due to

maternal age

REVIEW OF THE LITERATURE

Review o f the Literature on the Problem

Spontaneous Abortion

Women who have suffered a spontaneous abortion have traditionally not been

supported psychologically through their time of grieving by either the medical or the

mental health community (Leon, 1987) Along with the lack of understanding and

concern on the part o f these professionals, the lay public has also tended to belittle the

concerns o f those who experience a spontaneous abortion There is a silence that

pervades the whole issue of perinatal loss - as if the mention o f the possibility may

bring about the occurrence (Reinharz, 1988) This silence begins very early in the

pregnancy cycle and in the case o f one experiencing a pregnancy loss, is most often

resumed immediately following the loss This unvoiced, superstitious belief serves to

contribute to the fact that couples are both uninformed about and unprepared for

pregnancy loss Literature addressing the nature of spontaneous abortion and the

factors that have been found to predispose individuals to pathological outcomes

following these losses is reviewed

One result o f the shroud of silence that has been found to surround the issue of

19

pregnancy loss is the idea that people assume that the whole reproductive process is a

simple linkage between desire and accomplishment, a fait accompli As Reinharz

(1988) stated

when conception is assumed to be the simple outcome o f the choice to conceive and when pregnancy and a baby are assumed to be the natural outcomes o f conception, an exaggerated notion of control has been introduced into our view of reproduction People speak o f the choice

to have or not to have children Unfortunately, the very notion of choice hides the fact that the only thing about which women have some choice is not to reproduce, (p 86)

The fact that most couples are led to believe that with modern medical care

pregnancy leads to a healthy baby has led to an assumption that they are more in

control o f the outcome than in actuality they are Therefore, a spontaneous abortion is

not only shocking since the silence surrounding loss has kept them unaware o f the

commonality o f the occurrence, but jolts them to realize that control o f reproduction is

not in their hands alone

Definition o f a Spontaneous Abortion

For the purposes o f this paper a spontaneous abortion (miscarriage) is classified as

a spontaneous (as opposed to induced) pregnancy loss occurring between conception

and twenty eight weeks gestation There are approximately 600,000 to 800,000

spontaneous abortions in the U.S alone annually (Beil, 1992) The collection of

epidemiological data is controversial because o f methodological problems Often

spontaneous abortions are not recognized by the women who have them, and some,

although recognized, require no medical intervention, and therefore go unreported

Some statisticians place the spontaneous abortion rate as high as one in four

conceptions Buehler (1983) stated the percentage o f pregnant women in the United

States who are cognizant o f their pregnancy and then abort the fetus is about 10%-

14%, while others place this rate at 20% (Cavanagh & Comas, 1982) When a study

o f 221 women who were attempting to conceive employed highly sensitive pregnancy

detection methods, 22% o f the 198 pregnancies ended prior to the pregnancy being

detected clinically Another 31% of the pregnancies in this study were lost after

clinical deteciion (Wilcox et al., 1988)

Psychological Repercussions of Spontaneous Abortion

There has not been much research conducted in the area o f psychological

repercussions o f spontaneous abortion Those studies that have been done have

consisted primarily of case studies with a very small number o f subjects The

literature has reported emotional trauma and often intense depression on the part of the

women experiencing spontaneous abortions Many women have stated that this

depression has not been supported by others, either in the lay public or by medical or

mental health professionals In order to provide emotional support for victims, more

complete information about the psychological repercussions o f perinatal loss is needed

(Brody, 1980; Holland, 1982; Jimenez, 1982; Pizer & Palinski, 1980)

One study that dealt with the emotions experienced following spontaneous abortion

was conducted by Madden (1986) Sixty-five women were asked to describe their

emotions and coping strategies following miscarriage Sadness following the

spontaneous abortion was by far the most common emotion reported by the women

This emotion was reported by 88.6 percent of the participants Following this were

frustration (35.4%), disappointment (35.5%), and anger towards themselves (28%)

All respondents reported experiencing a sense o f loss that was very intense, and all

reported crying afterward However, what was perceived as lost varied for the women

Immediately following the spontaneous abortion, emotions were more intense for those

women who had time to become attached, held themselves more responsible for the

loss and were unable to talk to anyone about the loss Madden (1986) concluded that

miscarriage is clearly a disturbing experience, but that there seems to be a wider range

o f reactions to it than has been indicated in case studies

Seibel and Graves (1980) administered a self-report questionnaire with both forced-

choice and open-ended questions to 93 patients who presented for a dilation and

curettage (D&C) following an incomplete spontaneous abortion A checklist o f 16

adjectives, including 4 positive and 12 representing depression, anxiety, and hostility,

was used Two thirds o f their subjects were either single, separated, divorced, or

widowed, and 72% of the pregnancies were unplanned Therefore, it was not surprising

that at least one positive adjective was checked by 29.3% o f the patients and 13.4%

checked two or more However, even with this population, negative feelings were

much more common, with 44% checking either unhappy or very unhappy Only 11%

checked no negative affect adjective, while 89% checked at least one and 23.1%

checked four or more

In a study looking at psychological outcome following spontaneous abortion,

Friedman and Gath (1989) interviewed sixty-seven women who were admitted to a

hospital for complete or threatened abortion treated by evacuation o f the uterus

Semi-structured interviews were completed within four weeks o f having had the spontaneous

abortion Along with the interviews, each woman’s mental state was assessed using

the Present State Examination (PSE; Wing, Cooper, & Sartorius, 1974) The women

also completed self-rating scales, including the Beck Depression Inventory (BDI; Beck,

1967); the Eysenck Personality Questionnaire (EPQ; Eysenck & Eysenck, 1975); the

Maudsley Marital Questionnaire (MMQ; Crowe, 1978); and the Modified Social

Adjustment Scale (SAS-M; Cooper, Osborn & Gath, 1982) Levels of emotional

distress were found to be high during the four week period following the spontaneous

abortion As determined by the PSE, 32 o f the 67 women were determined to have

major psychiatric disorders, which is four times higher than would be expected in the

normal population In every case, the diagnosis was depressive disorder Although

this diagnosis was confirmed by the scores on three depression rating scales, the

relatively mild level o f depression indicated that many of the women were already

beginning to recover Depressive symptoms were significantly associated with a

history o f previous spontaneous abortions Childlessness was also significantly

associated with depressive symptoms, although less so than history o f previous

abortions

In an attempt to predict short and long-term grief following perinatal loss (i.e.,

spontaneous abortion, ectopic pregnancy, fetal and neonatal death), Lasker and Toedter

(1990) measured the levels o f grief experienced by one hundred ninety-four bereaved

parents at three intervals The subjects were referred from a variety of private

practices and hospital clinics, and grief scores were calculated at two months, one year,

and two years post-loss Variables fit into one of five categories: individual

characteristics, characteristics of the loss, coping resources, other stressful life

conditions, and expectancy for the success o f a subsequent pregnancy The research

team found that contrary to their hypothesis, those women who had no history of

fertility problems and who had expectations for a subsequent successful pregnancy

were likely to experience higher grief scores At two months post loss all variables in

the model were significant in predicting grief, except other stressful life conditions

The only variables found to predict grief throughout the two years post loss were

characteristics o f the loss, particularly length o f the pregnancy, and coping resources,

specifically mental health This was contrary to the hypothesis that individual

characteristics would be important for long-term grief

To summarize the literature, sadness, a sense o f loss, and depression tend to be the

hallmarks of the psychological repercussions following spontaneous abortion What is

perceived as loss evidently varies for individual women Levels of sadness and

depression appear to be at higher levels immediately following the loss, with fewer

characteristics having long-term predictability in terms of level o f grief Length o f the

pregnancy and coping resources have been found to predict grief levels two years after

the loss

Attachment

The relationship between the gestational age of the fetus at the time of the loss and

maternal grief has been investigated Kirkley-Best (1981) reported a strong positive

relationship between gestational age and grief (r=.39, p<.001) while Lasker and

Toedter (1990) found gestational age at the time o f loss to be the strongest predictor of

active grief

Other studies have looked at level o f grief compared to type o f perinatal loss

Type of pregnancy loss is highly correlated with length o f pregnancy Peppers and

Knapp (1980b), in their study of 75 women experiencing miscarriage, stillbirth, or

neonatal death, found no difference in the grief response according to length of

gestation The reaction to loss was as great in the case o f an early miscarriage as it

was for either a stillbirth or the loss o f a neonate Reactions to spontaneous abortion,

stillbirths, and neonatal deaths were not shown to be significantly different in intensity

o f maternal grief

Although Peppers and Knapp (1980a) did not find differences in grief between

types o f perinatal loss, the study was flawed in sampling, as the researchers neglected

to control for the length o f time since the loss In the Peppers and Knapp study, the

time since the event ranged from several weeks to thirty years Leppert and Pahlka

(1984) did not find a significant difference in the grief levels o f those experiencing

miscarriage versus stillbirth or neonatal death However, as the dependent measure

was the participants’ subjective estimate of their initial grief reactions and all

participants were being seen for counseling, the findings are limited The authors

pointed out that every subject in their study broke out in tears when speaking o f their

loss

In a related study, Kennell, Slyter, and Klaus (1970) conducted interviews with 20

women who had lost neonates and used six key signs to evaluate each mother in terms

of a mourning score These signs included sadness, loss o f appetite, inability to sleep,

irritability, preoccupation, and inability to return to normal activity The researchers

concluded that there was no apparent relation between the length o f the baby’s life and

the mourning score

Theut et al (1989) studied 25 pregnant women and their husbands who had

experienced a perinatal loss within the prior two years The study attempted to isolate

factors that might predict unresolved grief during a subsequent pregnancy They found

that the parents whose loss had taken place later in the pregnancy grieved more during

the subsequent pregnancy than those who had experienced a loss earlier in the

pregnancy They attributed this to increased attachment following quickening, or felt

fetal movement, which, along with Klaus and Kennell (1982) and Condon (1986), they

see as a "significant marker in the attachment process" (Theut et al., 1989, p 637)

These conflicting results suggest the possibility of a modifying variable for length

of pregnancy One possibility o f a modifier is attachment to the pregnancy

Attachment has been defined as "an enduring affectional tie that one person forms to

another specific individual" (Kemp & Page, 1987) Attachment to the pregnancy

pertains to the presence o f an affectional tie between the pregnant woman and the

Attachment theory suggests that this tie is an innate one, which serves the purpose

of assuring proximity between the mother and child following the birth as a safety

mechanism If the mother is affectively bound to the child she will strive to maintain

physical proximity, and will thereby be able to guard the child from threat of harm or

danger (Ainsworth et al., 1978; Bowlby, 1969)

The bond between a mother and her child that is so apparent immediately at the birth o f her child is developed and structured during pregnancy At birth there is already a sense of knowing the child, within the limitations o f not having had perceptions through the usual sensory modalities At birth there is already a sense

of shared experiences, shared history, and shared time on an intimate and exclusive plane There is a sense o f ‘We-ness’, a sense of ‘I-and-you.’ (Rubin, 1975, p 149)

Robson and Moss (1970) measured "level" o f attachment by 54 primiparous

mothers based on one prenatal interview and several interviews during the first three

months after birth They found that maternal attachments intensified beginning in the

second month postpartum Attachment was defined as the "extent to which a mother

feels that her infant occupies an essential position in her life".(p 977) They attributed

this increase at two months to the infant’s ability to interact responsively, especially

visually and via smiling at this age The recognition by the infants o f their mothers

which was reported to begin at this stage was felt to solidify attachment Kirkley-Best

(1981) pointed out that the Robson and Moss (1970) study, while making a

contribution by focusing on maternal variables, was flawed Their definition of

attachment was lacking and the descriptive nature o f the study provided few controls

Recent studies have attempted to identify a woman’s attachment in terms o f her

psychological view of pregnancy Leifer (1977) in a study o f 19 primigravidas,

identified affiliative behaviors and concluded that emotional attachment to the fetus

develops soon after conception and deepens with the progression o f the pregnancy He

found a relationship between prenatal affiliation and quickening, the beginning of felt

fetal movements in pregnancy Affiliation increased significantly following

quickening

Cranley (1981) administered her prenatal attachment tool to 71 subjects who were

between 35 and 40 weeks pregnant She also concluded that an emotional affiliation

with the fetus was present during pregnancy

Kemp and Page (1987) in their study o f 88 married women with normal versus

high risk pregnancies found no significant correlation between attachment scores on

Cranley’s scale and educational level, age, race, whether the pregnancy was planned,

whether a sonogram was done, or the ordinal position o f the infant This would

support developmental theory that suggests maternal attachment is based on the

accomplishment o f a developmental task, demographic characteristics not withstanding

There were no significant differences in the scores o f normal versus high risk

pregnancy groups in terms of attachment in the Kemp and Page (1987) study Both

sets o f women developed feelings of attachment to their fetus during pregnancy One

reason presented as an explanation for not finding significant differences was the fact

that the instrument was completed during the third trimester, which may have been far

enough along in the pregnancy for the women in the high risk group to be optimistic

about infant survival It is unclear as to whether these variables may affect maternal-

fetal attachment to a greater extent in the earlier weeks o f pregnancy

Madden (1986) found that the 65 women who underwent a spontaneous abortion

in her study were very attached both to the idea o f the pregnancy and to an image of

the child The intensity o f the emotions experienced immediately after the miscarriage

was correlated with degree o f attachment and pregnancy length, but degree of

attachment and pregnancy length were not related However, planning o f the

pregnancy and wanting o f the baby were related to the level o f attachment Some

subjects became very attached to the baby within a very short pregnancy Others

reported less attachment and listed such reasons as prior miscarriages, other problems

with pregnancy, or ambivalence about the pregnancy

In a study by Hutti (1992) the degree of significance attached to the miscarriage

experience was related to whether or not the pregnancy and the baby were "real" to the

parents Women reported that even after medical confirmation, it took a while before

they began to "feel pregnant." This feeling was described as appearing when physical

symptoms o f pregnancy (i.e., breast tenderness and morning sickness) were first

noticed Males stated that the pregnancy became real to them when they could see

evidence of the baby’s growth (i.e., feeling it kick or seeing the baby during a

sonogram) Once the pregnancy became real for parents, they began to identify the

baby as a specific individual with characteristics and a personality, rather than as a

"generic" baby "When the pregnancy and baby within were real to the parent, an

apparently correlated grief response resulted after miscarriage occurred" (p 409) The

most intense grief responses tended to be experienced in parents for whom both the

pregnancy and baby were experienced as real For those parents perceiving only the

pregnancy as real but who had not yet started thinking about a particular baby with

specific characteristics, a less intense response o f shorter duration occurred When

parents perceived neither the pregnancy nor the baby as real, there was often little or

3 0

no response to the miscarriage Many o f these affiliative behaviors which indicate

awareness o f the baby’s unique characteristics are included on the measure o f

attachment

In summary, some studies show a relationship between gestational age and intensity

of emotion following loss, while others have not found such a relationship This

suggests the possibility of a modifying variable Madden’s (1986) study shows that

although both pregnancy length and attachment predicted intensity o f emotion

following loss, these two variables were not related This study will look at

attachment as a modifying variable which may predict level o f grief and help to

explain the disparity o f findings related to pregnancy length and level o f grief

Attributional Theory

Recently there has been a great deal of interest in looking at different individuals’

methods of coping with negative life events In particular factors that can help

determine those who are most at risk for negative psychological outcomes have been

studied The role of causal attributions for negative life events has been recognized as

an important one by researchers The relationship between learned-helplessness and

depression has been studied (Abramson, Seligman, & Teasdale, 1978; Seligman, 1975)

as has coping with uncontrollable negative life events (Bulman & Wortman, 1977;

Janoff-Bulman, 1979; Silver & Wortman, 1980; Wortman, 1976) Janoff-Bulman

(1979) looked at causal attributions in terms o f predicting coping and Abramson and

Martin (1981) in terms of depression Other theorists have noted the tendency of

groups o f people to blame themselves for negative life events that are in actuality

beyond their control Examples o f such groups have included concentration camp

prisoners (Bettelheim, 1943), parents o f children with leukemia (Chodoff, Friedman, &

Hamburg, 1964), rape victims (Burgess & Holmstrom, 1974a, 1974b, 1976) and

victims o f freak accidents (Bulman & Wortman, 1977)

There has been a controversy among authors as to whether this tendency for self­

blame has played an adaptive o r maladaptive role in the coping process Beck (1967)

viewed self-blame as a maladaptive symptom of depression; Abrams and Finesinger

(1953), in their work with cancer patients, stated that guilt was a maladaptive response;

Weisman (1976) stated that he found self-blame to be a counter-productive response in

his work with the bereaved Janoff-Bulman (1979), however, saw self-blame as an

adaptive coping response because it implies a sense of personal control over outcomes

There have been numerous studies in which perceived control over one’s outcomes

was related to improved adjustment and health (Glass & Singer, 1972; Langer &

Rodin, 1976; Thompson, 1981; Wortman, 1976) Studies where self-blame led to

better coping included work with accident victims (Bulman & Wortman, 1977), the

bereaved (Becker, 1962), and parents o f terminally ill children (Chodoff et al., 1964)

Janoff-Bulman (1979), in an attempt to reconcile these different findings, proposed

two distinguishable types o f self-blame The first, self-behavior blame, involves an

attribution to a controllable or modifiable aspect o f the self Self-character blame is an

attribution to uncontrollable aspects of the self, and is therefore seen as a self-

deprecating response Abramson and his associates, in their reformulated model o f

learned helplessness (1978), go beyond a simple internal-external attributional