Title Page
 Table of Contents
 Maternal attachment
 The forgotten grief: The experience...
 Biographical sketch

Grief in response to prenatal loss
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00099507/00001
 Material Information
Title: Grief in response to prenatal loss : an argument for the earliest maternal attachment
Physical Description: vi, 116 leaves ; 28 cm.
Language: English
Creator: Best, Elizabeth Kirkley, 1954- ( Dissertant )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 1981
Copyright Date: 1981
Subjects / Keywords: Perinatal death -- Psychological aspects   ( lcsh )
Love, Maternal -- Psychological aspects   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Abstract: There has been relatively little research which has addressed grieving processes to loss in pregnancy. Many persons have assumed that no maternal attachment occurs in pregnancy and therefore that grief should not occur at loss. This study has been directed at observations of grieving reactions to losses occurring at all times in pregnancy and the variables affecting those reactions. Thirty-eight subjects experiencing prenatal loss were interviewed at one month postpartum regarding their experiences during the previous month. The structured interview was taped, edited, and blindly rated on a 10-scale rating system of grief symptomatology. Inter-rater reliability was established. As was predicted, when employing a stepwise regression model, gestational age accounted for approximately 33% of the variance in grief score, and the variance accounted for was increased to over 40% with the addition of number of children as a second variable. The number of previous pregnancy losses and the degree to which the pregnancy was wanted were not significantly associated with grief score. Demographic variables as predicted, were not associated with grief score, except for age which showed a slight negative correlation. The relationships of grief score with the variables described above are discussed. Implications for maternal attachment in pregnancy and directions for further research are offered.
Thesis: Thesis (Ph. D.)--University of Florida, 1981.
Bibliography: Includes bibliographic references (leaves 107-115).
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Elizabeth Kirkley Best.
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000295254
oclc - 07863729
notis - ABS1597
System ID: UF00099507:00001


This item has the following downloads:

PDF ( 4 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
        Page v
        Page vi
        Page 1
        Page 2
    Maternal attachment
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    The forgotten grief: The experience of parents of stillborn infants
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
    Biographical sketch
        Page 116
        Page 117
        Page 118
        Page 119
Full Text








I wish to thank my committee for their help in the design and

preparation of this study. The following people are included: Dr.

Hannelore Wass, Dr. Ted Landsman, Dr. Pat Miller, Dr. Harry Grater, and

Dr. Franz Epting, Chairman. For statistical and computer programming

assistance and advise, thanks are due to John Hallan, Terry LaDue, Rex

Walker and Jamie Algina. Dr. Ken Kellner and Valerie Hiers are noted

for their help with the actual mechanics of the study. The mothers who

participated in the study are sincerely thanked for sharing their time

and sorrow in an endeavor to help others who will be confronted with

perinatal death.

Special mention must be made of my daughter, Sarah Rose Best, who

demonstrated love and patience far beyond the call of a four year old.


ACKNOWLEDGEMENTS. ............ .. .... . . i

ABSTRACT.. .. . . .. .. . . . . . . . .

INTRODUCTION . . . . . . . . . . .

MATERNAL ATTACHMENT . . . ... ..... . . . 3

Maternal-Infant Bonding: An Argument for a Sensitive
Period of Maternal Attachment. . . . . . ... 5
Prenatal Maternal Attachment. . ... ... . . .. 15

GRIEF .. .............. . . . . . .. 21

Bereavement, Grief, and Mourning. . . . . . ... 22
Theories. . . . . . . . . . . . 23
Grief: Its Description, Processes, and Components .... 33

INFANTS .... . . . . . . . . . . . 57

Description and Processes . . . . . ..... 57
Phases of Grief . . ........... .. ... 58
Pathological Variants of Perinatal Grief. . . . ... 62
Conditions and Treatment in Stillbirth Bereavement. ... 65
Conclusions and Areas for Research. . . . . . . 71
Statement of the Problem. . .. . ........... 71

METHOD. .... ....... .. ............. 74

Subjects. ................... . 74
Procedures and Materials. .......... . 74
Training of the Raters and the Rating Scale . . ... 77

RESULTS . ... .. .. .. .... . . . . . . 80

DISCUSSION. . . . . . . . . .. . . ... 84

Conclusion. .... . . . . . . . .... 90
Summary . . . . . . . .... .... .. 92

APPENDICIES. ............. ... . . . . 94

INFORMED CONSENT FORM . . . . . . . ... 95

II TRAINER RATING. .... .. ..... . . . .. 96

III QUESTIONS . . . . . . . . . . . 99

IV INTERVIEW MATRIX. . . . . . . . . . 101

FORM .. . . . . . . . . . . . 102




BIBLIOGRAPHY . . . . . ... . . . . . 107

BIOGRAPHICAL SKETCH. ..... . . . . . . . . 116

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



Elizabeth Kirkley Best

June, 1981

Chairman: Dr. Franz Epting
Major Department: Psychology

There has been relatively little research which has addressed

grieving processes to loss in pregnancy. Many persons have assumed that

no maternal attachment occurs in pregnancy and therefore that grief should

notoccur at loss. This study has been directed at observations of griev-

ing reactions to losses occurring at all times in pregnancy and the

variables affecting those reactions.

Thirty-eight subjects experiencing prenatal loss were interviewed

at one month postpartum regarding their experiences during the previous

month. The structured interview was taped, edited, and blindly rated

on a 10-scale rating system of grief symptomatology. Inter-rater

reliability was established.

As was predicted, then employing a stepwise regression model,

gestational age accounted for approximately 33% of the variance in

grief score, and the variance accounted for was increased to over 40%

with the addition of number of children as a second variable. The

number of previous pregnancy losses and the degree to which the pregnancy

was wanted were not significantly associated with grief score.

Demographic variables as predicted, were not associated with grief

score, except for age which showed a slight negative correlation.

The relationshipsof grief score with the variables described above

are discussed. Implications for maternal attachment in pregnancy and

directions for further research areoffered.


In the last 20 years there has been a diverse and expanding

literature on maternal-infant attachment. Bowlby (1958) first used

the term attachment to imply an innate affectional bond between the

caretaker and the infant, and thereby established a researchable

construct, distinct from the behavioral/psychoanalytical notion of

dependency. The field of attachment and attachment behaviors has

served to produce prolific theoretical and empirical accounts.

At the same time, the last 10 years have been characterized by

a growing body of psychological literature dealing with the construct

of grief. The field of human grief and grieving processes has recently

led to theoretical systems, although empirical investigations of grief

have been limited. While grief may be described as the psycho-

physiological reaction to the severing of an attachment (Bowlby 1961;

Parkes 1972) the literature equating elements of attachment and grief

is almost nonexistent outside the works of Bowlby (1969, 1973, 1979)

and Parkes (1972). Persons who are researchers in the area of grief

tend to focus solely on the components, outcome, and treatment of

grief (e.g., Lindemann 1944; Parkes 1972; Sanders 1977). Persons

researching attachment are even more focused, with their empirical

efforts oriented towards the development of attachments primarily in

infancy (see e.g., Schaeffer and Emerson 1964; Ainsworth 1972; Rajecki,

Lamb, and Obmascher 1978). When grief is mentioned in the attachment

literature it is used primarily to demonstrate an element of separation

effects, when the infant is separated from his mother.

Grief and attachment, however, are not distinct nor separable

entities. They are reciprocally involved with one another without

regard to differential theoretical explanation. The processes of

attachment are intertwined with the processes of grief; an attachment

always risks the occurrence of grief and grief is as specific to the

bereaved individual as the attachment was.

The relationship between grief and attachment has not often been

studied in a scientific vein. The aim in this paper is to prepare an

argument in which the presence of grief will be maintained to be the

evidence of a lost attachment. The argument herein is aimed in a more

focused sense at prenatal maternal attachment to the unborn child as

evidenced by the existence of perinatal bereavement syndrome, or grief

at perinatal and fetal death. In order to adequately set forth a

thesis on the timing and elements of prenatal maternal attachment, it

will be necessary to focus on the literature of maternal attachment

and the literature on grief including the psychology of stillbirth.

The proposed order of this paper shall focus firstly on maternal

attachment, and secondly on the processes and elements of grief. A

third attempt will be made to tie together the constructs of grief

and attachment in a fashion which shall facilitate the argument for

the existence and timing of the mother's attachment to the unborn

child and the crossroads of grief and attachment. Those who are

interested in theories and research on infantile attachment are

referred to Kirkley-Best (1981).


Relatively little research has dealt with maternal attachment

until recently. There has been a curious twist to the research as well

for while infantile attachment literature has focused largely on the

child's 3rd and 4th quarter of life, the period immediately following

birth has spawned the most research interest in regard to maternal

behavior. This idea of a sensitive period for maternal attachment is

usually studied under the rubric of "bonding." Prior to the latter

half of the 1970's, however, only a few authors concerned themselves

with maternal attachment. Brody (1956) looked at patterns of maternal

behavior in a group of 32 mothers, but her work dealt to a large degree

with feeding behavior and was interpreted in a psychoanalytic context.

The focus of psychoanalytic theorists on feeding behaviors has not

fared well as being the sole determinant of mother-child attachment.

Newton (1951) focused on mothering styles as they related to personality

variables in the child, but again, the bond between the two was not the

central issue.

Robson and Moss (1970) were two of the first researchers to focus

attention on the mother's tie to the child. In a study of 54 prima-

parous mothers qualitative data were collected regarding the mother's

"level" of attachment. Women were interviewed shortly before birth and

several times after delivery until their children were three months

old. Attachment was defined as the "extent to which a mother feels

that her infant occupies an essential position in her life" (Robson

and Moss 1970, page 977). Based on their interviews, Robson and Moss

concluded that while in the hospital most women report "affection"

for their babies, they tend to see their infants as "non-social objects."

Further, many women in this study even reported negative feelings about

the new "intruder" in their lives in the first several weeks of the

child's life. Maternal attachments during this period are not true

attachments according to Robson and Moss and are in part attenuated

from the in-hospital stay.

In the second month, however, maternal attachments begin to

intensify. According to Robson and Moss, this is in large part because

they come to perceive their infant as responsive. The visual and gaze

smiling behaviors of the infant at this age are seen as "rewards' for

mothering. Mothers feel by this time that their infants recognize them

and this seems to solidify maternal attachment (Robson and Moss 1970).

Robson (1967) has remarked t ;at parents of blind infants often take

offense at misinterpreted gaze aversions by the infants.

Robson and Moss (1970) made a valuable contribution to the field

of attachment study by focusing research attention on maternal variables.

Their research, however, was lacking on many points. Their definition

of attachment was rather loosely defined. One of the qualifiers was

that a mother who was attached to the infant would experience a sense

of loss at the babies' absence. It is hard to see how an unattached

mother would grieve at a newborn's death (before 3 months), and yet in

studies of stillbirth mothers and mothers who lose newborns almost

every mother has been observed to grieve at the loss (Kennell, Slyter

and Klaus 1970). It is hard to imagine that the examples of the

mother's comments are in any way typical or representative of most

mothers on infants 1-2 months of age. One mother for example characterized

herself as feeling no connection with the child; another mother dreamed

of her baby being a monster in the crib. While frustrated new mothers

occasionally make similar comments under the fatigue and stress

characterizing the newborn period, these comments seem unwise as

characterizations of early attachments. The Robson and Moss study

was an uncontrolled, descriptive study and much more positive

characterizations of early maternal attachments are found elsewhere

in the literature (e.g., Rubin 1975).

One observation in the Robson and Moss study is of particular

interest. Over 75% of the mothers had traditional hospital care which

involves a great deal of separation in the in-hospital period. As

shall be seen in the following section, researchers have, since the

Robson and Moss study, taken a keen interest in the critical nature of

the post-partum period in maternal attachment. While Robson and Moss

felt that mothers who chose "rooming-in" (keeping the baby by the

mother's hospital bed instead of the newborn nursery) gained no

advantage in establishing attachment with their infants, this subjective

determination, as shall be seen, was premature.

Maternal-Infant Bonding: An Argument for a
Sensitive Period in the Development of
Maternal Attachment

In a variety of studies with animals, there appears to be a

critical period for maternal attachment to the infant shortly after

birth. If in this critical period the mother-infant pair are separated,

there can be a severe disturbance in maternal behavior towards the young

when the infant is returned to the mother. The effect has been noticed

in dogs, goats, sheep and cattle (Hersher, Moore and Richmond 1958;

Scott and Marston 1950; Arling and Harlow 1967; Klaus and Kennell 1975).

Evidence of this sort led some researchers to believe that there might

also exist in the human mother a sensitive period for attachment follow-

ing birth.

Klaus, Kennell, Plumb and Zuehlkke (1970) began this investigation

into the concept of a sensitive period in a study designed to observe

any consistencies in maternal behavior at first contact with her young.

Twelve mothers of normal infants and 9 mothers of preemies were observed.

Systematic progressions of maternal behaviors were observed in both

groups. In the words of Klaus et al.

. the mothers started with fingertip touch on the
infants extremities and proceeded in 4 to 8 minutes
to massaging, encompassing palm contact on the trunk.
(Klaus et al. 1970, page 187)

According to these researchers, most mothers began with a predominance

of fingertip touching (52%) and progressed in a number of minutes to a

predominance of palm contact. These authors like others (Robson 1967)

also noticed the central importance of eye-to-eye contact even in the

first few minutes of mother-child interaction. The remarkable similarity

in maternal behavior led the researchers to postulate a sensitive period

immediately following birth during which species-specific behaviors took

place (Klaus et al. 1970).

Since the Klaus et al. study there has been a growing body of

evidence emphasizing the importance of the first few post-partum hours.

This evidence has taken several forms. Several researchers have looked

longitudinally at the effects of separation usually comparing groups of

extended early-contact mothers with limited early-contact mothers.

Another area of contribution has dealt with physiological and

cognitive factors which predispose a mother to a sensitive period in

her attachment to her infant. These areas will be discussed separately.

Bonding and Separation Effects

Barnett, Leiderman, Grobstein and Klaus (1970) published one of

the first accounts of the effects of interactional deprivation in a

study of mothers of premature infants. While only descriptive data were

collected on the mother's emotional status, the conclusion drawn was

that increased contact with the premature infant resulted in more con-

fident and secure maternal attitudes. Perhaps the most significant

finding in this study was that there was no increased risk of infection

in the infants whose mothers were allowed to hold them and see them.

This was not a minor point. Klaus and Kennell (1972) in reviewing the

history of hospital care for full-term and premature infants had noted

that around the turn of the century mothers were not only allowed to

care for preemies but were strongly encouraged, and most full-term

births occurred at home. By the mid 1900's, however, it was not only

standard practice to separate mother and infant in hospital routine,

but also to exclude her from child care matters. This physical

separation of mother and newborn has been standard practice in most

American hospitals, due partly to the convenience of the staff, and

partly to the unsubstantiated "risk of infection" (Klaus and Kennell


With the knowledge that there was no increase in physical risk to

infants, researchers began to be willing to experiment with the effects

of allowing extended contact in the post-partum period. The first study

of this sort was conducted by Klaus, Jerauld, Kreger, McAlpine, Steffar

and Kennell (1972). Twenty-eight primaparas were divided into 2 groups

of 14 each. In the control group, mothers received limited physical

contact with their infants, typical of prevailing hospital policy. The

second group or extended contact (EC) group were afforded 16 extra

hours of infant contact during their hospital stay. When maternal

behavior was rated blindly one month later several statistically signif-

icant differences were found. During a routine infant exam, EC mothers

attended to the procedure most closely. In a structured interview, EC

mothers reported a reluctance to leave their infants in someone else's

care. During a filmed bottle feeding situation, EC mothers exhibited

significantly more soothing behaviors and a greater degree of eye-to-

eye contact.

These mothers were followed up again at 11 months post-partum, and

the EC mothers still showed significantly more soothing behaviors to their

child's cries and attentiveness during a well-baby exam (Kennell, Jerauld,

Wolfe, Chesler, Kreger, McAlpine, Steffa and Klaus 1974). At 2 years

of their infants age, a subsample of 5 mothersin each group was

studied in terms of mother-to-infant speech. When checked at age one,

the only difference in the EC and control group mothers in terms of

speech was that the EC group used fewer statements. When the child

was 2 years of age, EC mothers used fewer content words and more

adjectives in speech to their toddlers. They (EC) also used fewer

commands, asked more questions and used more words when making propo-

sitions to their infants than did mothers from the control group. From

the time their infants were 1 year old to the time they were 2, EC

mothers showed an actual decrease in use of verbs and commands (Ringler,

Kennell, Jarvella, Navojosky and Klaus 1975). Ringler et al. (1975)

concluded that the EC mothers were by all measures more attentive and

appropriately sensitive in verbal interactions with their children,

an observation which would most likely have great effect on the child's

future behavior.

Other researchers have found similar effects. Hales, Lozoff,

Sosa and Kennell (1977) observed 60 Guatemalan primiparas in three

groups. Group I was a control group receiving routine care, group II

received 45 minutes of extra contact with their infants immediately

after birth (EC-early) and group III received 45 minutes of additional

infant contact 12 hours after birth (EC-delayed). When observed for

maternal behavior differences at 36 hours post-partum EC-early mothers

showed significantly more kissing, smiling, fondling, talking and

"en face" behaviors than did control mothers. No differences were

found in regard to caretaking and proximity maintaining behaviors.

The one behavior that seemed to have the most differentiating effect

among groups was the "en face" behavior, or the holding of an infant

such as to maximize eye-to-eye contact. With regard to en face

behaviors, EC-early mothers received ratings significantly higher than

both the EC-delayed and the control group mothers. The importance of

eye-to-eye contact in attachments has already been discussed, and these

findings by Hales et al. again contribute to evidence for a sensitive

period in human mothering.

Two studies of Swedish primiparas have yielded additional infor-

mation regarding a maternal sensitive period (cited in Lozoff, Brittenham,

Trause, Kennell and Klaus 1977). In the first study by Carlsson, 62

middle class mothers were observed particularly in breastfeeding

situations. The EC mothers in this study were afforded an additional

hour at birth with their babies. These EC mothers were compared with a

control group unlike previous control groups in that Swedish hospitals

afford a great deal of mother-infant contact as standard practice

although usually not till several hours after birth. Nonetheless when

observed on the 2nd and 4th days of hospitalization, the EC mothers

showed significantly more physical affection towards their infants

while nursing. Even if these effects are temporary, Lozoff et al.

(1977) note that the hour right after birth can be a particularly

sensitive period for maternal attachment.

In another study of Swedish primiparas, de Chateau (see Lozoff

et al. 1977) compared the effects of a group treated with 15 minutes

of early physical mother-infant contact with a control group treated

according to hospital routine. (In this hospital, treatment was

similar to American hospitals; while mothers saw their wrapped infants

in a nearby crib for approximately 1 1/2 hours 30 minutes after birth,

during the next 3 days they only saw them at 4 hour feeding intervals.)

EC mothers received the same routine treatment as the control group

following their extra contact. Even under this minor manipulation, at

36 hours post partum the EC mothers showed more sitting up, holding,

close carrying and cradling behaviors when with their infants than did

control group mothers. As much as three months later EC mothers showed

more kissing and en face behaviors in an in-home 10 minute free-play

situation. Babies of EC mothers showed more smiling and laughing

behaviors and less crying than did control group babies. Since in

both Swedish studies mothers see their babies shortly after birth,

Lozoff et al. (1977) have noticed the critical role of physical contact

in the sensitive period.

Perhaps the most convincing evidence of the importance of the

first post-partum hours has come in an unpublished but widely-cited

study by O'Connor (in Lozoff et al. 1977; Campbell and Taylor 1979).

Two hundred seventy-seven American low-income primiparas were randomly

assigned to differing post-partum conditions. The control group received

routine care; the EC group were given up to 8 additional hours with

their infants. When these women were followed up one to two years

later only one case of child abuse was found out of 134 EC mothers.

On the other hand, 9 out of 143 late contact mothers exhibited a

parenting disturbance resulting in their child's hospitalization owing

to reasons like failure to thrive and child abuse. A study by Peterson

and Mehl (1978) looked at 46 mothers in the last trimester of pregnancy,

7 days after birth, and then again at I month, 2 months and six months

after birth. Peterson and Mehl found that the length of separation of

mother and infant post-partum was the greatest "predictor" of later

attachment scores. The results of this study can only be viewed as

tentative, though, due to many design deficiencies, unvalidated measure-

ments and risky interpretations.

While not arguing against early separation effects, Fields (1977)

in a complex study of mother-infant pairs involving pre-term, term, and

post-term babies demonstrated a somewhat more complicated interaction

between early separation and characteristics of the infant producing

worry in the mothers. According to Fields, extended post-partum contact

in and of itself does not insure healthy mother-infant attachment later.

Family situation, health of the infant and so on may obscure the

effects of early contact. Likewise, early separation does not guarantee

later aberrant attachment patterns. Nonetheless, a sensitive period does

seem to exist around the time of birth, and additional factors in this

early period lend themselves to this early, heightened emphasis on


Factors Influencing the Sensitive Period

It has been known for sometime that just the sight of an infant

elicits favorable responses from adults. This effect, termed "babyness"

by Lorenz, seems to hold for the young of most species in regard to

human preference (Brooks and Hochberg 1960; Fullard and Reiling 1976).

It seems reasonable to assume that a mother's first view of her infant

elicits even more intensified responses to the "cute" infant than more

generalized responses.

There is some evidence that mothers are able to differentiate

painful infant cries from hunger cries. Formby (1967) found that most

mothers were able to accurately distinguish their infants' cries from

those of other infants within 2 days after birth. The difficult dis-

criminative learning involved in the recognition of infant cries may

indicate a greater attentiveness than usual on the part of the mothers

in this period.

Physiological changes may contribute to an argument for a

sensitive period. In a recent paper Kimball (1979) noted an excess

of endorphins in the placenta at the time of birth. Endorphins have

been associated with "good" feeling states, and Kimball postulates that

in labor and delivery the endorphins may be released into the mother's

system resulting in a pleasant emotional state at birth. This period

of pleasurable feeling might well facilitate feelings of attachment

in the human mother.

When talking about a sensitive period for human attachment in the

post-partum period, most of the concentration has been on the mother's

sense of attachment. Current developmental research, however, has

demonstrated that the neonate is more competent than previously thought

and at the moment of birth possesses the behavioral capacity to interact

with the mother at some level. Brazelton (1963) and Campbell and Taylor

(1979) have noted research indicating that infants are alert and

wide-eyed for about one hour after birth although this length of this

period may be affected by maternal medication received in the course

of delivery. After this hour of attentiveness infants usually fall

asleep for long periods. Knowing the centrality of eye-to-eye contact

in establishing attachments, it follows that this infantile alertness

period which so closely coincides with a time when mother-infant separation

has lasting effects adds strong contributory evidence to support the

idea of a sensitive period in human attachment. Lozoff et al. (1977)

have noted that even a crying or drowsy neonate will become alert when

lifted to the mother's shoulder, arguing again for the critical role of

physical and not just visual contact. The heightened sensitivity of

both mother and child is also reinforced by the observation that at

birth the infant will differentially respond to sounds within the range

of the human voice, and at 1 week neonates will react with distress at

their "masked" mothers. Attachments for both mother and child, while

qualitatively different from attachments formed after the 3rd quarter

of life, seem nonetheless to take dramatic hold at the time of birth

and in the neonatal period in general.

Maternal Expectations

Before moving to a discussion of maternal attachment in pregnancy

it is fitting to mention a small body of literature which has dealt with

a mother's expectationof what her baby will be like. As Klaus and Kennell

(1976) have pointed out, a mother must come to reconcile her "inside"

baby to her actual infant. Most often the "inside" infant is idealized

to a degree that would put the Gerber baby to shame. Actual infant

outcome, however, is almost never in complete accord with maternal

expectations. This discrepancy may range from something as simple as

having pictured the brown haired infant as blond to dissatisfaction with

the baby's gender. Kaplan and Mason (1960) have pointed to maternal

reactions to premature birth as evidence for the sometimes difficult

adjustment to unconfirmed expectations. They have set a course of

tasks which the mother of a premature infant must work through in order

to deal with the outcome of a sick infant. Bidder, Crowe and Gray

(1974) have likewise noted that even the mother of a term infant sees

a discrepancy between this child and their ideal child. The most

beautiful child in the first minutes after birth, with elongated skull,

covered in vernix, cannot meet the mother's expectations (Brazelton

1963). The need to reconcile this discrepancy of the ideal to the real

child is used as a further argument toward keeping the mother-infant

pair together in the moments after birth so that this maternal cognitive

task may be successfully accomplished (Taylor and Hall 1979).


While most of the arguments presented here have supported the

notion of a sensitive period in human mother-infant attachment, it should

be noted that this research is not without its critics. Some of the

research which has been done has had methodological flaws, and many of

the studies have not studied attachment effects past the neonatal

period. Another criticism which faces bonding research is its emphasis

on "strength" of attachment rather than quality of attachment. As

Ainsworth (1973) has pointed out, a psychology of attachment will not

proceed when based on a quantifiable "amount" of attachment. Lastly,

much of the research on separation effects during the sensitive period

has been done by Klaus and Kennell and their associates. Their

arguments will be made more plausible when researchers in diverse

surroundings are able to replicate their findings. In the words of Taylor and

Hall (1979), "it is easy to become attached to bonding." (p. iv)

Prenatal Maternal Attachment

Maternal Attachment as an area of psychological study has yielded

only limited findings. Most systematic research has dealt only with

the post-partum period or the first few months after birth. By sins

perhaps of omission, there has been a curious lack of research attention

given to the maternal attachment to the unborn child in pregnancy.

This is not to say that pregnancy has been ignored in psychological

research. A survey of the literature indicates great interest in the

emotional aspects of pregnancy, particularly in how emotions figure

in the outcome of pregnancy. Similarly, researchers have been interested

in the role of emotions in wanted-unwanted pregnancies,and attitudes

towards labor and delivery (Grimm and Venet 1966; Colman 1969; Sugarman

1977; Standley, Soule and Coupans 1979). In a review of the pregnancy

literature over the last twenty years, it is noticeable that only a

handful of authors have addressed the problem of prenatal maternal

attachment. These accounts have rested largely within the context of

psychoanalytic theory, derived primarily from case history studies and

personal observations.

One of the first authors to address the mother's attachment to her un-

born child d was Deutsch (1945). Deutsch sought to interpret processes in

pregnancy in psychodynamic terms. She felt that pregnancy was a crisis

in personality development in need of resolution. Early in pregnancy,

according to Deutsch, women show an increase in narcissism and a general

turn inward though not directly toward the fetus. Upon quickening

(felt fetal movement) a woman begins to direct her psychic energy towards

the fetus and begins the task of owning the pregnancy with its threaten-

ing implications. Fetal movement is felt to be threatening because

subconsciously it represents the idea of a moving phallus. Successful

resolution of the crises in pregnancy serves also as resolution of

penis envy. Deutsch believed that pregnancy was characterized by a

general regression to the oral state resulting in an increased

dependency on the part of the mother.

Bibring and colleagues (1959, 1960, 1961) worked within a similar

framework, observing pregnancy as a crisis to be resolved. She again

pointed to quickening as a critical point in the mother's attachment to

her unborn child. She noted that at this point mothers often begin to

fantasize about the forthcoming child. Bibring and colleagues (1960) and

Klaus and Kennell (1975) have also remarked that after feeling fetal move-

ment, even unwanted pregnancies may seem more acceptable. Bibring agrees

with Deutsch in recognizing that a woman initially invests energy in

herself and that it is only later that she is able to focus that energy

in a relationship with the fetus. A woman must learn to individuate

herself from close identification with the fetus in the last trimester

of pregnancy, in a sort of "letting go" strategy.

Reva Rubin (1975) has attended particularly to attachment processes

in pregnancy. Rubin proposes that a woman has four major tasks in

attaching to her infant. These tasks are described as safe passage,

acceptance of the child by significant others, binding-in, and giving

of oneself.

Safe passage for both self and baby concerns the pregnant woman

in different ways throughout pregnancy. In the first trimester there

is mostly concern for the safety of self, but with the onset of fetal

movement, the woman focuses her attention on protection of the unborn

child. By the third trimester there is equal concern for self and child,

and environmental dangers are overemphasized. According to Rubin,

women tire of pregnancy in the eighth month and delivery is valued.

Concurrently, with safe passage a woman seeks to have "important

others" value her child just as she does. This is particularly

important early in pregnancy when the acceptance of the child by

others may be conditional upon certain characteristics of the child,

such as gender. A pregnant woman in Rubin's view is very vulnerable

to rejection.

"Binding in" refers to the task of a woman to bond herself to her

unborn child, and this bond is characterized at birth by a sense of

"we-ness." In the first trimester the binding-in is only to the

pregnancy and not to the fetus. In the second trimester quickening

brings on an intensified concentration on maternal tasks, assisted by

elevated levels of progesterone and estrogen which according to Rubin

contribute to a feeling of well-being. A woman's love for her child is

at a high point and then wanes in the third trimester. This may be

because while the mother loves the child, she is tired of pregnancy,

and this weariness may attenuate the binding-in process.

Finally, a woman learns to give cf herself. In the first trimester,

there is a cost analysis of the pregnancy. There is an appraisal of

what the child can give, compared to what he/she may take away. In

the second trimester a pregnant woman focuses on the quality of that

which she will receive. The third trimester is characterized by a

renewed appraisal of whether or not she will be able to give her child

that which she deems appropriate.

Bibring, Deutsch, and Rubin are not the only persons to have

focused on the attachment process in pregnancy, but when prenatal

maternal attachment is mentioned in the literature, the works of these

women are mentioned almost exclusively. While their work may be plauded

for the recognition of prenatal attachment, their propositions and

descriptions are lacking in many respects. Coming from a psychoanalytic

bias, their writings are often an effort to fit pregnancy in that

conceptual framework. Furthermore, their observations rest on a shakey

data base, relying on a few case histories or personal opinion. It is

a curious note that in a field so potentially important there have been

almost no controlled investigations of the phenomena involved in

prenatal maternal attachment.

The one observation that all observers seem to agree upon is the

importance of quickening in prenatal maternal attachment. This seems to

be when for many mothers the baby becomes a real entity. This "reali-

zation" of the fetus seems for some mothers to allow attachment to take

place, and for others to greatly intensify processes already begun.

Quickening provides the first tangible characteristics of infant

specifically, and not just the pregnancy in general. It is a point at

which the developing life is made real to the mother. An early, but

testable proposition is that quickening in and of itself may not be the

only attachment-producing phenomenon, but so will any situation which

produces a tangible characteristic of the infant. Monitors which

amplify fetal heart tones early in pregnancy or ultrasound "photographs"

of the infant in utero may likewise produce the evidence which forms

the groundwork of attachment. This point of realization has never been

explored empirically.

Maternal attachment and all attachments are felt to be a corner-

stone in the personal and emotional adjustment of any individual as well

as playing a critical role in human species adaptation. However

unfortunate, attachments are often severed by death, separation and

even choice. When an enduring attachment has been formed and then


broken (especially by the finality of death) human beings universally

grieve, although grief may take a variety of forms. Attachment and

grief are inextricable from one another. In order to explore adequately

the relationship between attachment and grief it is necessary now to

turn to a discussion and review of the literature of grief.


To begin a discussion of grief is to enter a field encompassing

the most painful of human experiences, and also the most inevitable.

With the exception of the stillborn infant or the infant who dies

shortly after birth, all persons experience grief even if it be only

rudimentary responses to the absence of a caretaker. Unless one dies

as a child or a young adult, it is likely that he or she will experience

one or more major losses in his or her lifetime.

In the last 5 to 10 years, the subject of grief has received an

overwhelming amount of attention in psychology and medicine. Nonethe-

less, there are still comparatively few systematic investigations in

this field. The area of grief is prolific with popularized and uninformed

accounts of how to deal with the grieving person, very often stemming

from the "theories" of Elisabeth Kubler-Ross, a well-intentioned

physician who unfortunately produced a conceptually impoverished

framework of grief (Kubler-Ross 1969).

There have been a number of investigators who have slowly and

systematically pieced together controlled observations of grief which

have led to more useful conceptualizations, and these efforts are

noticed most completely in the works of Bowlby (1979) and Parkes (1972).

Drawing on their works and that of others, the present discussion will

proceed in the following manner. Theories of grief will be discussed,

as well as the processes, components and risks of grief. Atypical

patterns of grief and their determinants will also be examined along

with methods of treatment. While "grief may have the various cause" it

will be readily apparent that the concern in this paper is with bereave-

ment via death of a loved one. Much of the research and theoretical

focus has been on widows, so many of the studies considered deal with

widowhood which to a large degree produces a prototypical picture of

grief. In a discussion of grieving parents, this picture has been

slightly modified to accommodate the loss of a different sort of

relationship. A review of the literature on perinatal grief (especially

grief over stillbirth) will ensue and this "type" of grief will be used

as evidence for early maternal attachment.

Bereavement, Grief, and Mourning

Before beginning a discussion of the theories of grief, it is

necessary to distinguish between 3 terms found often in the literature.

These are bereavement, grief, and mourning. While some authors such as

Bowlby (1979) argue that definitions of grief and mourning are artificial

distinctions, a majority of authors use these terms specifically so it

is appropriate to provide definitions. Bereavement refers to the

actual state of loss; bereavee" comes from the Latin verb meaning to

rob. Grief may be defined as the psychophysiological reaction to the

loss, including the subjective understanding of the loss. Mourning

refers to culturally determined behaviors and practices for the expression

of grief (Dempsey 1975). Bowlby uses the term mourning as meaning the

psychological expression and process of dealing with the loss so that

most of his work includes the word "mourning" in place of grief (Bowlby

1979). It seems appropriate to use the term grief in this context in

an effort to distinguish the universally experienced emotions and

processes from cultural patterns of expression which may differ greatly

from one another. As can be surmized, while all three of these states

are highly associated, it is conceivable that a person might grieve and

not mourn (e.g., the stoic who refuses all ceremony but is internally

hurt by the loss) or to mourn but not grieve (e.g., the distant relative

who goes through the motions of the funeral but whose emotional reaction

is minor). Both grief and mourning occur in reaction to some type of

bereavement, most typically death.


Only a few theorists have concerned themselves centrally with the

issue of human grieving. The majority of these theorists have come

from the psychoalantyic point of view. Lately the ethological/cognitive

point of view has been formulated in some depth, and research on grief,

as on attachment, has not only emanated from this position but has

supported it consistently. There have been some behavioral accounts

of grief in recent years, but outside the work of Seligman (1975) there

have been no attempts to draw individual behavioristic models into a

system. Behavioral theories will be mentioned only briefly; psycho-

analytic accounts will receive somewhat more attention. The ethological/

cognitive theories arising from such authors as Bowlby (1979), Parkes

(1972), and Averill (1968) will be treated in some depth due to their

salience in the research literature on grief.

Behavioral Models

Perhaps the most well known behavioral model is that of Seligman--

the "learned helplessness" model. According to Seligman, bereavement is

a protypical circumstance in which a person learns that his behaviors

are ineffectual in bringing about the reinforcement he seeks. When a

person learns that his actions are ineffectual, he or she may give up

further efforts even when those future efforts would result in the

desired outcome. At some point the "helpless" person may decrease

his or her level of behavioral activity, attributing even successes

to fate or chance. Seligman uses his model essentially to replace the

concept of depression of which he believes grief is a special case

(Seligman 1975).

While helplessness is certainly apparent in the grieving process,

to hold it accountable for the whole process has been criticized. The

meaning a person attributes to their grief would be described only in

terms of behavioral contingencies although social learning theorists,

of which Seligman is one, are also likely to take into account some

cognitive processes as intervening variables. This model leads to a

view of grief in terms of behaviors which can be unlearned. If the

biological and cognitive processes of grief are ignored in favor of

behaviors alone, treatments can be composed which would be not only

ineffectual but detrimental, especially in terms of internal processes.

The idea of behavior versus process lies not only with Seligman's

theory. Ramsay (1977) has proposed a model in which grief is viewed as

a brand of depression resulting from insufficient reinforcement. The

grieving person is on an extinction schedule. Ramsay's model draws the

conclusion that the development of pathological grieving is analogous

to, if not the same as the development of phobias, and their treatment

is virtually the same. Ramsay proposes his model with intentional

disregard for research and theorizing in nonbehavioral approaches to

grief. The view of grief as resulting from the removal of positive

reinforcement is incomplete, failing to describe why the yearning and

pain of grief while becoming infrequent may last a lifetime in normal


Bugen (1977) has devised a behavioral model for the prediction of

the magnitude of grief. Bugen states that there are only two factors

which are necessary to take into account, centrality and preventability.

Centrality refers to whether the deceased play a central or peripheral

role in the life of the survivor. Preventability refers to whether or

not the death is perceived to have been preventable. According to

Bugen the most intense grief will result when a central figure dies a

death which might have been prevented.

Psychoanalytic Theories

Psychoanalysts have not been so eager to opt for easy answers

although they are as likely as behavioral theorists to present their

theories more in an effort to promulgate a model than to investigate

the phenomena of grief as they are observed. Valuable observations have

been made by several persons in this tradition especially Freud (1917),

Abraham and Klein (in Pollock 1961), and Siggins (1966).

Freud (1917), in his classic work "Mourning and Melancholia,"

laid the groundwork for psychoanalytic considerations of grief. Freud

saw the grieving process as a defensive one. According to Freud, when

a cathected object is lost, the reality of the situation demands that the

ego take notice. When the ego takes notice of the loss, an intensely

painful process is set in motion. In order that a person not be over-

whelmed with pain, defense mechanisms such as denial come into play.

The ego denies the reality of the death and may act as though it has

not occurred in order to escape the severe pain of grief. Eventually

however, reality again dictates and pangs of grief return. This piece-

meal progression of accepting and "mourning" the loss accounts for the

long duration of grief and the searching behavior which often appears.

"Grief work" as coined by Freud, sees the ego hypercathecting to the

artificially-prolonged object in order to eventually decathect that

lost object and invest that libidinal energy in new relationships

(Freud 1917).

The pain of grief perplexed Freud; it was not immediately evident

to him why the process should occur in so great an intensity, greater

even that that of separation anxiety. Says Freud,

S..why it is that this detachment of libido from its
object should be such a painful process is a mystery
to us. .. We only see that libido clings to its
objects and will not renounce those that are lost even
when a substitute lies ready at hand. Such then is
mourning. (Freud 1916, page 307)

Pollock (1961) notes that Freud's difficulty in addressing the

grieving process may easily have arisen out of Freud's own grieving

experiences including the loss of his father, the loss of Jung's support

and the traumatic loss of a young granddaughter. The key recognition

in Freud's position is that one never completely gets over a loss. In a

letter which Freud wrote to the psychiatrist Binswanger, he states,

Although we know that after such a loss the acute state
of mourning will subside, we also know we shall remain
inconsolable and will never find a substitute. No
matter what may fill the gap, even if it be filled
completely, it nevertheless remains something else.
(Freud 1960, Letter 239)

Abraham, working in the same tradition,went a step further in

suggesting that not only did a person work to decathect psychic energy

from the loved object but he also worked to introject aspects of the

object into himself. According to Abraham, a person first introjects

the lost object and then works to decathect the external object; the

ego being willing since the ego then carries a replica of the object

inside. If the former relationship with the lost person is ambivalent,

that ambivalence will be introjected with severe consequences since

the ambivalence will be turned toward the self, resulting in events

ranging from guilt to self-destruction and other forms of pathological

mourning. The writings of Fenichel confirm these observations and add

that in normal persons, decathexis is easier with introjected rather

than external objects.

While few psychoanalytic models differ, from this basic model, the

models of Klein and Bibring center on other salient aspects of grief.

To Klein, grief is a process of regression to an earlier state, a

throwback to the weaning process. Bibring sees helplessness and the

potential for a loss of self-esteem as the primary issue in grieving

(Siggins 1966). All psychoanalytic theories, though, have this in

common: that grief is the process of withdrawing the energy from a

lost relationship and establishing that energy elsewhere. This economic

system of psychic energy is seen as an adaptive process.

Ethological/Cognitive Model: The Bowlby-Parkes Formulations

The psychoanalysts interpreted grief as a process by which a person

withdraws energy from a lost object in order to maintain self-preservation.

Bowlby and Parkes have concluded that the face of the grieving process

is not inherently adaptive for the individual in a narcissistic attempt

to identify with the living. Bowlby instead sees the grieving process

as an outgrowth of an adaptive system which promotes attachment and

proximity to the attachment object; grief itself is not seen as a with-

drawal of energy but instead as an expression or process which usually

brings a separated individual back into contact with his group (Bowlby


There are two aspects of this theory of grief, apparent from its

title, those being ethological observations and observations stemming

from developments in cognitive psychology. Since both aspects are

vital to this theory, each shall be discussed in turn.

Ethological aspects of grief

The ethological bases for grief rest on the assumptions of Darwin,

that behaviors which most enhance the survival of the species will be

selected out over a long period of evolution. Human beings are a social

species and conditions which facilitate group cohesion will be promoted.

Attachment and proximity-seeking and maintaining behaviors have this

function. Grief in the ethological sense has as its base separation

anxiety. When an individual becomes separated from his attachment

figure or group, the behaviors and emotional pain brought on by this

separation serve the purpose of bringing the separated person back in

contact with the lost figures. Death, the permanent loss of an attachment

figure is statistically rare in the scheme of separations;and behaviors

such as weeping, searching, protesting, and so on which normally have

the function of reuniting the separated person with the group, occur

in response to the death even though they are clearly nonadaptive in

face of the permanence of the loss (Bowlby 1961, 1979). Says Parkes,

Only when searching is useless and reunion impossible
as in the statistically infrequent event of loss by
death does the involuntary expression of grief lose
its utility. (Parkes 1972, page 42)

Grieving occurs, then, not to decathect an object, but instead as an

anachronistic system which cannot accomplish its goal.

The ethological position receives support in studies of animal

"grief." One would expect that if the ethological position is correct

one would find grief-like behaviors in social animals high on the

phylogenetic scale. This has been observed. Bowlby (1961, 1979),

Pollock (1961) and Averill (1968) have reported on a variety of

observations of grief in animals, including geese, dogs, monkeys

and apes. Exemplifying these reactions in the case of Washoe, the

chimp raised with a human family and taught sign language as communi-

cation. At one point Washoe's infant became ill and required hospital-

ization. Each day when her caretaker came, Washoe signed questions

regarding her infant's welfare to which the appropriate reply was

made. After a time the infant died and when thus informed Washoe

retreated to a corner refusing to move, eat or communicate for days

(Omni 1979). Similar patterns have been noted in monkeys by all of the

above authors. Averill (1968) notes that grief may have an adaptive

function not to the individual as Pollock suggests but to to the

group. A bereaved individual may greatly decrease reproductive activity

in bereavement which in traditional ethological thought would have

resulted in the gradual decrease in these behaviors in human evolution.

Averill points out, however, that the behavior of the grief-stricken

tends to elicit the caretaking and proximity of others in the group.

The individual painfulness at separation leads each individual to favor

the group. Averill argues that natural selection promotes survival of

the species not necessarily of any one individual, and therefore the

bereaved, even if failing to reproduce, contribute to the social co-

hesion of the group.

Cognitive aspects of grief

While cognitive issues have always been addressed in Bowlby and

Parkes' theory, recently Bowlby (1979) has elaborated an information

processing approach to grief and the defensive processes therein. Bowlby

does not reject the psychoanalytic concept of defense but feels that

current cognitive theory better accounts for the observed phenomena.

Bowlby rests his approach to defense on information-processing

research. Bowlby notes that all persons initially process a great deal

more information than they retain. Only a small amount of this infor-

mation proceeds past the sensory level to higher order processing, and

this selective exclusion has adaptive value, protecting a person's

capacity to attend to stimuli from being overwhelmed and continually

distracted. Defensive exclusion is a type of exclusion of events which

prevents certain types of information from being processed, permanently

in some cases. Even though a person may not initially attend to a

piece of incoming information, that information may still influence his

behavior and mood without his awareness (as studies on dichotic listening

have demonstrated). Three things can happen to the incoming stimulus.

It can be excluded entirely, it can be kept momentarily in a "buffer

store" just long enough to affect judgement and mood, or it can enter

into complex higher-order processing, usually associated with conscious-

ness. This perceptual defense system seems especially to exclude infor-

mation which in the past has been associated with painful states.

Defensive exclusion can readily explain the "defense mechanisms" of

psychoanalysis in a cognitive view. For example "repression" results

from a consistent pattern of exclusion of information which would serve

as cues for a particular behavioral system and the thoughts and feelings

associated with it would remain totally out of the realm of awareness.

Most often though, the defensive exclusion is not complete, resulting

in "repressed" phenomena seeping through to consciousness. Defensive

exclusion can act in other ways such as diverting attention from one

line of stimuli to another, disconnecting or redirecting certain

cognitive events (as in "displacement"), "purposeful" misinterpretation

of a stimuli (such as in phobias--l'm not afraid to leave my family,

I'm afraid of wideopen spaces and so on). What seems to be critical in

the determination of "pathology" is how persistently certain cues are

excluded from awareness. When a person totally excludes information

necessary for the function of a cognitive-behavioral sequence (such as

in the case where the death of a loved one is so totally denied that no

grief takes place) the effects can be devastating. A system which is

constantly expending resources in dysfunctional exclusion processes

necessarily expends adaptive capacities in favor of the less adaptive

ones. Bowlby says,

There can be little doubt that those persons in
whom defensive exclusion plays a prominent part
are handicapped in their dealings with other human
beings when compared to those in whom it only plays
a minor part. (Bowlby 1979, page 72)

When, as in the case of grief we have to dismantle an entire

system (relationship) to which we are accustomed, the process is not

only difficult and slow, but painful as well. In this manner, defensive

exclusion allows for "bouts" with the painful tasks and moratoria from

the pain, a pattern which can well account for the initial numbness and

intermittent pangs of grief (Bowlby 1979).

Bowlby's arguments for cognitive and biological bases of grief fit

well with the observed data. Parkes (1972) has proposed that the

grieving individual is additionally in a state of alarm which was

appropriate and adaptive in evolutionary history. The world in grief

is a dangerous place and the physiological arousal of alarm fits well

also with the emotional observations of initial grief.

The one point which this theory lacks is the transition from the

building blocks of the theory (behavioral systems, defensive exclusion,

etc.) to the meaning of grief to the individual. This is not to say

that Bowlby and Parkes do not address the question of a loss of meaning

in grief. Both authors have expounded with great sensitivity on the

personal meanings of grief -in their research and case examples. This

theory does provide a suitable base of understanding into which the

"meaning of human grief" might fit. An attempt to integrate this

meaningfulness into a less mechanistic model would in the end provide a

complete understanding of human grief allowing all disciplines to con-

tribute to a comprehensive goal.

Grief: Its Description, Processes, and Components

In 1944 Lindemann published his now classic paper "Symptomatology

and Management of Acute Grief." Since his original propositions based

on the observation of grief in over 100 survivors of the Coconut Grove

Fire who had lost a relative in that disaster, there have been a number

of attempts to systematically describe and record the components of

grief. Most of these studies have focused on the acute phase of grief--

within the first few months of the loss, a time when the effects are

most observable. Most researchers now agree that grief is best viewed

as a process not a state (Parkes 1970, 1972; Glick, Weiss and Parkes

1974; Bowlby 1979). For this reason, it is best to consider "sympto-

matology"--both physiological and psychological--within the context of

the phases or processes of grief which persons encounter.

Grief to most workers is viewed as a disease process based on the

arguments of Engel (1961). Engel likens grief to any "natural or normal

response to trauma" as in the case of a wound. According to Engel the

course of the healing process, in the wound as well as in grief, may

have a normal progression ending in a healed state, or it may have a

bad outcome. Pathological grief in this manner of speaking may then

be analogous to a septic wound. Both Bowlby (1979) and Parkes (1972)

have accepted this analogy. It is not the intention in this paper to

belabor this point; it will suffice to say that arguments of this

nature can be countered. What is important to realize is that grief

is seen as a disease by most writers in the area.

Processes or Phases in Grief

Several long term studies of widows have been conducted. Fewer

studies have been conducted with parents of deceased children. A

diverse group of researchers have arrived at observations of the phases

an adult may go through in grieving a loved one. Studies of widows are

believed to be prototypical of the adult grieving process. Bowlby

(1979) and Parkes (1972) have outlined the following phases to account

for the progression.

Shock and numbness

When a death occurs, especially if it is without warning, a common

reaction is one of numbness. Parkes (1970) in a longitudinal study of

22 widows found that 16 out of 22 widows reported a feeling of numbness

or shock. An example of this reaction is "I just couldn't believe it;

it didn't seem real." Glick, Weiss and Parkes (1974) described similar

reactions. Bowlby (1979) feels that the reality of the tragedy is

impossible to take in all at once. Shock reactions may last only a few

minutes or hours or they may last as long as a week. Two of the 22 London

widows observed by Parkes (1970) reported a period of shock lasting as

long as a month before the pangs of grief set in. In another study of

68 widows and widowers, 2/3 reported an initial sense of numbness (Parkes

1975). When the painfulness of grief does set in, whether a few minutes

or weeks into bereavement, a further process becomes salient--the

process or phase of searching and yearning.

Searching and yearning

Of all the phases of the grieving process, perhaps the most

obvious is the phase which involves searching and yearning for the lost

object. This phase involves the most acutely observable behaviors and

affects. Many researchers when looking at grief are looking at this

phase which begins simultaneously with the phase of numbing but may be

characterized by differing initial intensity which develops into the

overwhelming affect so easily recognizable. Most of the persons in

Lindemann's study were probably most involved in this process, as were

those of Clayton, Desmarais and Winokur (1968).

The symptoms indicative of this phase are diversified. They

include sighing respiration, loss of appetite and resulting loss of

weight, sleep difficulties, digestive disturbances, heart palpitations,

headaches, and muscular aches. Psychological symptoms include

irritability (restless anxiety), lack of concentration, guilt, rumination

of events surrounding and leading to the loss, preoccupation with the

image and thoughts of the deceased, overwhelming somatic and subjective

distress characterized as despair, depersonalization/derealization,

disturbance in time sense and any number of other reactions (Lindemann

1944; Clayton et al. 1968; Parkes 1970, 1972; Glick, Weiss and Parkes

1974; Bowlby 1979). The best way to conceive of grief is probably not

in terms of the components mentioned above but instead in terms of the

cognitive/affective processes at play, for while an individual may or

may not experience each of the above symptoms, he is likely to be

characterized by repeated attempts to yearn for, search for and recover

his lost object. Bowlby (1979) and Parkes (1972) characterize this

process by the following:

1. restless activity and scanning

2. preoccupation with thought of the deceased

3. developing a perceptual set for the deceased so that all

perceptions related to the deceased are closely attended to

and other stimuli are attenuated

4. attending closely to places which were associated with finding

the deceased and

5. calling for the lost person.

Yearning for the deceased person involves great pain. A person

may or may not realize his tendency to search for the person but whether

he realizes it or not, every effort and thought tend to be drawn toward

the process of recovery. This is not to say that the bereaved denies

the reality of the death, it is more of a temporary, involuntary dis-

belief in the situation.

The grieving person develops a perceptual set for the dead person

seeking the deceased everywhere. As one woman remarked,

Everywhere I go I am searching for him. In crowds,
in church, in the supermarket. I keep scanning the
faces. People must think I'm odd. (Parkes 1972,
page 47)

This perceptual set to find and recover the object occasionally

leads to misperceptions in which the deceased is momentarily found.

This momentary recovery of the object results in a temporary mitigation

of the intense pain of fruitless yearning and searching. A widow may

report hearing the door creak at 5:00 and know it is her deceased

husband returning from work (Parkes 1972). A mother of a stillborn

infant may report hearing her baby cry when alone in the house or

occasionally still feeling fetal movement (Davidson 1979). Parkes

(1972) reports persons feeling drawn to the graves of the deceased, and

any objects which link themselves to memories of the deceased.

Rees (1971) in a study of 293 widows and widowers found that close

to half of the bereaved persons he interviewed had either "seen" the

deceased spouse or had a sense of the person's presence. Yamamoto,

Okonogi, Iwasaki, and Yoshimure (1969) found 90% of Japanese widows

interviewed reported hallucinations of the deceased especially in the

first few months of bereavement. This higher percentage among Japanese

may be accounted for culturally; the Japanese value the presence of

ancestors. Marris (1958) has also noted reports of presence of the

deceased in 50% of the London widows interviewed, a finding incidental

to his study on the socioeconomic status of the widows. This pre-

disposition to "see" the deceased may arise from the high motivation

to find the person in one's perceptions even if that person no longer


The searching and yearning of this phase tends to come in waves

lasting from approximately 20 minutes to an hour (Lindemann 1944;

Parkes 1972). Parkes notes that an individual could not withstand the

constant and severe pain of "pining" for the deceased, and therefore

must temporarily withdraw from the task of searching. This is accomplished

in "selective forgetting" and the withdrawal to numbness. This process

of searching and not finding, withdrawing and then searching again, is

seen by Parkes as a process of "realization"--it is only little by

little that a person comes to accept the reality and permanence of the

loss. When the loss becomes more and more "real" to the person and

he or she recognizes the irreversible nature of the change, the rest-

less activity of searching and pining begins gradually to drop out and

be replaced by a period of disorganization. The duration and intensity

of the searching phase differs from individual to individual, but the

presence of yearning and searching, accompanied by the preoccupation with

thoughts of the deceased, is pathognomic to this acute phase of grief

(Bowlby 1979).

One additional characteristic of this phase needs to be examined

and that is anger. Bowlby (1961, 1979), and Parkes (1970, 1972) have

remarked on the important role of anger in this early stage. The anger

in this phase is intermittent and may be directed against the deceased,

others, "fate," God, or the self, often resulting in guilt. When the

death is sudden and without reason there is an intense anger towards

those held responsible for the death, be they God, doctors or the

deceased themselves. There is often a reproach toward the dead

exemplified by the statement, "Why did you leave me?" A death shakes

a person's sense of justice and security, and makes the world a dangerous

place to live. The tendency to place blame is believed by Parkes (1972)

to bring the world into an understandable state--to produce a reason

for the death. When the reproach is directed against the self, guilt

may result. While a tolerable level of guilt is normal to grief,

exaggerated guilt may lead to severe problems (Bowlby 1965). What is

important to recognize is that anger is an appropriate component of

grief even if it is misdirected and seemingly irrational.


Searching and yearning which are initially done compulsively

eventually, drop out and recur only in association with persons,

places, things and events associated with the deceased. Cavenar,

Spaulding and Hammett (1976) have noted that the anniversary of a

death or other important events associated with the deceased may pre-

dispose an individual towards a recurrence of intense distress and

even depressive or psychotic reactions. Glick et al. (1974) and

others have taken precautions in designing longitudinal studies of

grief to avoid interviews around the anniversary of the death. Besides

these event-related episodes of "acute" grieving recurrences, however,

the picture of the grieving process changes from one of severe and

painful searching and yearning to a more general depression-like

orientation occurring typically 6 months to a year or more after the

death. A person who has been able to allow for his own inconsistencies

and anger in the previous phase comes eventually to realize the

redundancy of his behaviors and when realizing that nothing can be

recovered he comes to a position of apathy and depression. C. S. Lewis

in an exploration of his own grief over the loss of his wife (H.) wrote,

Thought after thought, feeling after feeling, action
after action, had H. for their object. Now their
target is gone. I keep on through habit, fitting an
arrow to the string; then I remember and have to lay
the bow down. So many roads lead through to H. I set
out on one of them. But now there's an impassable
frontier-post across it. So many roads once; now
so many culs-de-sac. (Lewis 1961, page 59)

Once depressed, a person must begin to redefine both him or herself and

his or her life. This process results in reports of loneliness,

especially at night and has been found in all the studies mentioned

in this paper (Marris 1958; Clayton et al. 1968; Parkes 1970; Glick

et al. 1974). This loneliness may be perpetuated especially in widows

where there is a reluctance to reestablish social contacts out of loyalty

to the dead husband (Bowlby 1979). Eventually, however, most bereaved

persons do begin to rebuild their lives although this is an arduous



When a person begins to rebuild and reorganize his life depends a

great deal on the personality of the persons involved and their relation-

ship, and other factors to be considered later such as age, socioeconomic

status and so on. The timing of this phase for most seems to take hold

at approximately 18 months to 2 years after the death. Efforts to

reorganize become apparent before this time but as with the preceding

phases, it is at the above mentioned time that reorganization becomes

fully apparent. Glick et al. (1974) and Parkes (1970, 1972) have noted

that certain events in the bereaved person's life may facilitate this

rebuilding. Many of the London and Boston widows in these studies

mentioned "marker events" such as redecorating the house or cleaning

out a spouse's closet, as significant events. Whatever the facilitating

event may be, it seems that reorganization eventually sees the return of

physical appetites, reestablishment of social patterns, and judgements

unaffected by the loss. The most important thing to keep in mind is

that grief is resolved only insofar as it no longer affects the everyday

life of the bereaved, and a tie is never completely broken and grieving

may recur at anniversaries and future losses and reminders. It is not

uncommon nor even inappropriate for a parent bereaved of a child to

respond with sadness to any reference to that child. One learns

eventually to live with the reality of the loss; it is unlikely one

learns not to love the deceased.

Patterns of Grief

While the above described "stages" of grief are nearly universal,

the timing and duration may vary among individuals. Lindemann (1944),

Parkes (1965) and Averill (1968) have described six patterns which

grief may follow. Each will be briefly described.

1. Normal grief--normal grief follows the 4-phase pattern of

processes described previously. Normal grief most often has

a favorable outcome, clearly discernable 2 years after the


2. Chronic or exaggerated grief--in chronic grief persons seem

to get "stuck" in the phase of yearning and searching, less

often in despair. The outcome is usually not favorable and

often manifests itself in the intensification of one symptom

such as guilt.

3. Abbreviated grief--a sincere but short grief reaction which

may result from a low degree of attachment or an immediate

replacement of the deceased's role.

4. Inhibited grief--a pattern in which any given symptom of

grieving is not exhibited but are instead "repressed."

Inhibited grief is often characterized by displaced symptoms

such as illness. Children often show this pattern of grief

in which sorrow is often replaced with aggressive acting-out


5. Anticipatory grief--anticipatory grief described in some detail

by Fulton and Fulton (1974). It is a form of grieving which

occurs before an actual death most often in the instance of a

prolonged illness. Anticipatory grief is similar to normal

grief in its symptoms and function, but when the death

actually occurs there is a more subdued reaction than there

is to sudden death. A discussion of the positive and

negative aspects of anticipatory grief may be found in Fulton

and Fulton (1974).

6. Delayed grief--delayed grief may be seen in a variety of

ways. Most often a short delay is noticed, followed by

either normal or exaggerated grief. In some cases the delay

is so long in terms of years that the pattern has been termed

"absence of grief" but a distorted form of grief--somatic

or psychological, invariably follows, even if it does not

become apparent until a future loss. Delayed grief can

have severe consequences.

Epidemological Studies of Grief: The Risks of a Broken Heart

Probably the most widely studied sub-area of grief has been the

study of the risks of the mortality and morbidity in bereavement. There

is so much evidence relating the stress of bereavement to risk factors

that grief has often been called "the hidden illness."

Mortality of grief

Dempsey (1975) remarks that the data from the National Office of

Vital Statistics in 1965 reports a higher mortality rate among white

widowers between 25 and 34 than among their married counterparts. The

death rate for these widowers is double that of their counterparts.

Widows in this age group, although having fewer actual deaths, have

a 2 1/2 times greater risk of dying within the first year of bereavement

under a variety of circumstances.

Other studies have confirmed this finding. Rees and Lutkins (1967)

in a study of bereaved relatives in Llanidloes, Wales,over a period of

six years, reported the following data. In the year following a death

in the family, close relatives had a mortality rate of 5%, compared to

.68% for a nonbereaved control group. If the survivor was a spouse, the

results were even more dramatic--12% versus 1.2% mortality rate for

married controls.

Parkes, Benjamin and Fitzgerald (1969) found an increased risk in

widowers of 40% over controls. Many of the deaths in this study (3/4)

were related to illness of the heart; and it was noted that the notion

of a "broken heart" may not be completely the product of fantasy. In

another vein, Bunch (1972) studying bereaved individuals who had lost

a parent or spouse found a 5 times greater risk of death through suicide

in these individuals than in a matched control group.

Only one study reviewed has not found this increased rate. Ward

(1976) in a study of 87 widowers and 279 widows found none of the

differences in mortality rates. Ward compared the number of actual

deaths in her bereaved group not with actual deaths in a control group

but with expected rates indicated by life tables for England and Wales

in 1970 to 1972. The studies already mentioned have used matched and

concurrently sampled control groups yielding a more efficient design and

more reliable results.

Epstein, Weitz, Roback and McKees (1975) have described five

hypotheses which have been used to account for the increased mortality

rates in bereavement. These include the selection hypothesis, the

the homogamy hypothesis, the joint unfavorable environment hypothesis,

the non-grief related behavior-change hypothesis, and the desolation

effects hypothesis.

The selection hypothesis suggests that healthy widows remarry and

so are notpicked up by the studies. This criticism is easily laid aside

when one considers the increased risk in the first 6 months when few

are likely to remarry at all, and that after 2 years, the mortality

risk returns to that of the general population. One would expect from

this hypothesis that a lessening in risk would not occur.

The homogamy or mutual choice of high risk mates hypothesis suggests

that the sick and dying seek out and marry the sick and dying. Again,

if this hypothesis were used to account for increased mortality one

would not expect to see a return to normal rates over a couple of

years. Likewise the joint unfavorable environment hypothesis suggests

that sharing the same unhealthy environment would predispose the

increased risk factor, but one would have to ignore the rare occurrence

of couples dying of the same illness.

The non-grief related behavior change hypothesis suggests that

survivors show high risk behaviors not because they grieve but because

they behave differently (such as not taking prescribed medicine, etc.).

The desolation effects hypothesis suggest that the severe strain of the

changes in bereavement as well as of grief itself, produces such stress

and hopelessness that physiological and situational changes detrimental

to the person's health may occur. It is this final hypothesis which

has been most firmly supported, especially by endocrine studies in

bereavement. Hofer, Wolff, Friedman and Mason (1972) and Fredrick (1977)

have noted elevated or at least changed levels in 17-Hydroxy-cortico-

steroid excretion in bereaved parents and Fredrick relates this finding

to the effect of stress on immunological factors predisposing persons

to illness. No matter what the interpretation, there is little doubt

that persons recently bereaved have significantly higher death rates

than the population at large.

Morbidity and grief

A similar area of concern is grief's effects on the general health

of individuals. Parkes (1964) using widows as their own controls noted

a tripling in the number of physician visits in the first six months

of bereavement as compared with the six months prior to the death. This

may be for grief related symptoms, or seemingly non-grief related


Lindemann (1944b) found that out of 45 patients who developed

ulcerative colitis, 26 had experienced a death just prior to the onset

of the illness. Onset or aggravation of a variety of illnesses in

bereavement have been reviewed by Schmale (1958) and Dempsey (1975).

These illnesses include cancer, asthma, tuberculosis, leukemia, arthritis

and heart and circulatory problems.

Some of the problems with most of these studies is that they are

done after the illness has been diagnosed and the researcher (usually

an M.D.) is not blind to the hypothesis. Thus, while malignant cancer

has often been linked with bereavement or "object loss" (Schmale 1958;

LeShaun, in Dempsey 1975), these terms have been loosely defined as a

"sense of hopelessness" or "utter sense of despair"--both of which are

interpreted by the researchers involved.

Schmale, using his definition of object loss and resulting hope-

lessness, studied a group of women all suspected of having cervical

cancer. Based on interviews with the women and their reported patterns

of loss, Schmale predicted which cases would turn out malignant and

which would turn out benign. Using the criteria of object loss, Schmale

correctly predicted 75% of the outcomes. Other researchers studying

leukemia and lymphoma in children have noted the separation, divorce

or death of their parents shortly preceding the onset of their illness

in as high as 80% of their cases (Dempsey 1975). Illnesses such as TB,

asthma and arthritis which may have been arrested may be exacerbated

in the early phases of bereavement (Parkes 1972).

It is not suggested that grief causes the above illnesses. Grief

as an overwhelming stress may predispose physiological factors to

"cause" an illness. Likewise, a person may see an already existing

condition intensified. While it is relevant to point out that these

studies place the bereaved at risk for both illness and death, it is

also important to note that many bereaved persons will not suffer these

consequences. The conditions affecting outcome of bereavement and the

possible pathological responses to bereavement will be considered later

in this paper.

Morbid Grief Reactions

The bereaved run the risk of increased ill health to be sure,

but perhaps an even grater risk in this culture is the risk of

psychological "illness" resulting from the grieving process. Parkes

(1965a, 1965b) attempted to differentiate abnormal grief reactions from

normal grief reactions by viewing the symptomatology of a group of

bereaved psychiatric patients. While he notices a preponderance of

affective disorders in bereaved psychiatric patients, the differences

between the control and study group were not as clear as one might

think. Parkes concluded that it was not so much as one might think.

Parkes concluded that it was not so much a matter of different symptoms

as it was a matter of differences in intensity and duration. Abnormal

grief reactions may be prolonged or exaggerated forms of normal grief.

Bowlby (1979) feels that abnormal grieving reactions,while

presenting themselves in a great variety of ways, fall into two broad

categories. These categories may be considered 1) chronic grieving or

2) prolonged absence of mourning. A third relatively rare category

is that of mania or euphoria which replaces grief.

Both the chronic grieving process and prolonged absence of grief

are felt by Bowlby to have certain features in common. Bowlby believes

that both patterns involve defensive exclusion of information relating

to the permanence of the loss. The chronic mourners become "fixated"

in the phase of yearning and searching and are continually unable to

lay aside the search. The anger inherent in this phase of grief may

result in reproaches against the self or others in a manner so

intensified as to lead in some cases to chronic depression and self

destruction, and in other cases to paranoia-like patterns. In both

chronic and absent grief, there is an absence of appropriate sorrow

over the loss.

Chronic morbid grief reaction is most typically seen in the de-

pressed individual who years after a death is still characterized by

the painful yearning usually associated with the acute phase of grief.

Persons in this condition may also exhibit suicidal behavior (Bunch

1972) and other maladaptive patterns such as alcoholism. Any one

component of grief may be magnified in a distorted fashion. It is

not terribly unusual to see a person guilt-ridden years after a death.

The hallmark of chronic grief is the degree to which it interfers with

day to day functioning (Bowlby 1979).

A pattern which is more difficult to detect but which may have

grave consequences is that of prolonged absence of conscious grieving.

While some degree of numbing is perfectly normal within the first few

hours or even weeks of bereavement, prospective studies such as that

of Parkes (1975) have noted that longer delays are usually associated

with detrimental outcomes of grief. While some may argue that no grief

is occurring in these cases, there are usually tell-tale signs that

the person is indeed affected by the loss; the exclusion of grief is

almost never totally successful.

Prolonged absence of grief may take a variety of forms. A

common reaction is when a person shows no grieving responses him/

herself but becomes greatly concerned with the grief or distress of

another. Greene (1958) notes several cases of persons who enter into

an overprotective relationship in order to partially act out their own

grief. Greene sees this form of behavior as a disguised cry for help.

Another more common pattern exhibited when grief is "absent" is

the appearance of symptoms, illnesses or characteristics of the

deceased. While to some degree identification symptoms are normal,

exaggerated identification can result in an inability to accept the

loss. Zisook and DeVaul (1977) have termed physical reactions

"grief-related facsimile illness" and they note that these reactions are

either misdiagnosed as hysterical reactions or are treated inappropriately

as medical illness. If diagnosed as distorted grief reactions, these

"illnesses" have a good prognosis but if misdiagnosed they may last


Gorer (1965) has noted a variation of unresolved grief which he

termed "mummification"--a form of preserving relics associated with the

deceased in a sense of great reverence. Queen Victoria, for years

after the death of her husband, continued to have his clothes laid out

and shaving water drawn daily. An even more unusual effort to maintain

the deceased has been reported by Gardner and Pritchard (1977). Six

cases were reported of individuals who kept the deceased's body, in

some instances for years, taking great care to preserve it in any

manner possible. When public authorities became aware of most of these

cases and questioned those who kept the bodies, the "bereaved"

individuals often reported no knowledge of the death. Four cases

involved the death of a mother in which the survivor was an adult

son; only, one case involved an individual diagnosed as psychotic.

Eventually, all persons who do not grieve initially, break down--

most often in a manner characterized as depression. Bowlby lists 4

events likely to precipitate full blown responses in individuals who

have previously avoided grieving. These include:

-the anniversary of the death

-another loss which may appear minor

-an event such as reaching the age of a parent who died

-and in the case of compulsive care for a vicarious object,

a loss which the object experiences.

In all cases of morbid reactions, some form of re-grief therapy proves

to be most effective. While a variety of treatments have been proposed

(e.g., Polak, Williams and Vollman 1972; Flesch 1975; Ramsay 1977;

Valkan 1975) the paradigm for treatment is one in which the grieving

individual is led through his grief as though his loss had just occurred,

supported by the therapist. Some treatments arising particularly in

behavioral approaches, use implosive or flooding techniques in which

grief is treated as a phobia. These latter approaches are based on

knowledge of behavioral techniques to the exclusion of knowledge of

the literature on grief.

Conditions Affecting the Outcome of Grief

What has been discussed is how the process.of grief may take

shape. The picture is a complex one though and researchers are just

beginning to tease apart the conditions and predictors of outcome.

Bowlby (1979) has reviewed a number of variables which are thought most

likely to predict outcome. These include:

1. the identity and role of the deceased

2. age and sex of the bereaved

3. causes and circumstances of loss

4. psychosocial circumstances at the time of death and after

5. the personality of the bereaved with emphasis on his/her

ability to form and maintain attachments and deal with


Each will be discussed in turn.

The most obvious condition which predisposes intense grief is

the centrality of the lost relationship. Parkes (1972), Bowlby (1979)

and others (e.g., Bugen 1977) have noted that almost all research and

reports regarding disordered variants of grief have arisen from the

loss of a parent, spouse or child, and less often, a grandparent or

sibling. While the nature of the relationship which lends itself to

disordered mourning may not be clear, one observation is consistently

made. Although any bereavement may lead to pathological responses,

most observers of grief have noted the particularly severe and devas-

tating reaction to the loss of a child. Gorer (1965), Lindemann (1944),

Parkes (1972), Bowlby (1979) and others have all noted the most severe

reactions of grieving in parents, particularly mothers who lose young

children. The ambivalence or security of the relationship with the

deceased seems likewise to have an effect, with the poorer outcomes

attributed to relationships with higher ambivalence (Vachon 1976;

Bowlby 1979).

Age, sex, and socioeconomic status may have an effect on the

outcome of bereavement. Age of the bereaved is not a clear cut issue.

In several early studies young bereaved widows were found to be at high

risk for grief-related psychiatric problems. However, more recent

studies have found no age difference in terms of poor outcomes. The

very young and very old are often excluded from general studies of

bereavement because they are thought to be prone to inhibited forms

of grief (Parkes 1972). Ball (1977) has painted a more complex picture.

It seems not that age per se affects the outcome of bereavement so much

as the interaction between age and the mode of death. Anticipatory

grief had a mitigating effect on grief in younger widows but no

differential effect was observed in older widows. Young sudden-death

widows were rated most highly on an overall grief reaction measure,

on severity of symptoms, and on total number of symptoms. While any

age group may be characterized by poor outcome, it appears that the

general picture is one of high risk factors in young (18 to 46) bereaved

persons (Jacobs and Ostfeld 1977; Ball 1977). Age as a factor warrants

further study.

There is a debate as to whether gender in and of itself predisposes

poor outcome. Jacobs and Ostfeld (1977), in a review of the literature

on grief, have noted that men are at higher risk for poor outcome than

women at all ages with particularly poor outcomes for elderly widowers.

Vachon (1976) in a similar review corroborates this conclusion. Bowlby

(1979) and Parkes (1975), however, find no conclusive evidence that

suggests that one sex is more predisposed to poor outcomes of grief.

Most studies have not focused on the grief of males per se, and widowers

and bereaved fathers in general represent very few of the grieving

persons studied. Bowlby points out that the clinical picture and

timing of poor outcomes may differ between the sexes, with males

presenting difficulties as late as 2 to 4 years after the loss.

Socioeconomic status is a mixed picture as well. Parkes (1975)

reports low socioeconomic status widows have greater difficulties

adjusting to major loss, but others have noted either no differences

or greater difficulties among professional level persons (Vachon

1976). All demographic variables which affect bereavement need further


Causes and circumstances surrounding the death may have great

influence on the outcome of grief although exactly how that influence

is exerted is debated. As Ball (1977) has noted, sudden death may add

to the detrimental outcome of the grief of young widows. Glick et al.

(1974) and a variety of studies reviewed by Bowlby (1979) have

corroborated the evidence that the sudden death of a loved one has

particularly devastating effects. One of the most explicit cases of

the effects of sudden death has been noted in the study of Harvard

widows by Glick et al. (1974). They found that 2 to 3 years after

bereavement none of the 22 widows who had lost a husband to sudden

death had any signs of remarrying. Thirteen of the 20 widows who

had had some warning, were either remarried or had made plans to do

so. According to Glick et al., it is probable that the suddenness

of death has more elements of catastrophe, and those widows who were

thus bereaved developed a fear of dealing with similar situations.

Bowlby (1979) likewise points out that especially in young age groups,

the deceased spouse was likely to have been young as well and this

factor characterizes the death as particularly untimely and unjust.

Feelings of guilt, anger, and blame may intensify and may in turn

predispose pathological reactions. Parkes (1972) has noted that in

sudden death, the life-circumstances of the survivor are greatly

shaken, and these widows find themselves with financial and social

upheaval to contend with as well as with their grief. Where there is

forewarning, preplanning may take place, and those widows need content

only with their grief.

Other circumstances may also predispose poor consequences in

grief. These include concurrent stressful life events (Vachon 1976),

how the bereaved is informed of the death, what kind of interaction had

occurred between the bereaved and the deceased, and where blame for the

death is directed (Bowlby 1979). Since research studies in these areas

are few and conjecture is large, these will not be considered


The two main categories of psychosocial elements exist which may

affect bereavement: living arrangements and beliefs and cultural

practices. Living alone or living with others is thought to influence

the course of normal bereavement. Bowlby (1979) has concluded that

the social isolation of living alone may predispose depressive symptoms.

Clayton (1975) has noted that in the first year of bereavement, there

are no differences between bereaved persons living alone or with others

in reports of depression. There is a highly significant difference

between the bereaved and matched control groups in reports of depression.

She concludes that it is the bereavement itself rather than the effects

of living alone which causes depressive symptoms. Living alone may

contribute to increased reports of aspects of depression. Living

with others, however, can have detrimental effects particularly if the

bereaved is responsible for the care of those with whom he/she lives

(Bowlby 1979).

Probably one of the most important variables affecting grief is

the cultural rituals surrounding grief. Fulton (1974) noted that little

by little cultural rituals surrounding death have become less and less

evident in the western tradition. These cultural practices which include

funerals, mourning practices (e.g., dress, behavior), etc. have become

poorly defined; it is not infrequent when a death occurs to hear persons

react by saying "I didn't know what to say or do" or reacting by trying

to hide all evidence of grief (Parkes 1972). This lack of tradition

may undermine what little structure is available to the bereaved and is

believed to risk poor outcomes. The relation of cultural practices

to individual processes of grief needs to be further explored.

Lastly, the personality of the bereaved person will affect how

he grieves and what outcome follows. Vachon (1976) has stated that

previous psychiatric history may precede detrimental grieving processes.

Clayton (1968) has observed that the best predictor of a depressive

reaction at one year is severe depression at one month. Parkes (1975)

has found that severe self-reproach, anger, depression and yearning

which will not subside even momentarily, best predicts maladaptive

patterns of grief. Observations of personality styles suggest that

persons who are prone to be defensively independent, have a high

degree of defensive exclusion (or unwillingness to deal with stress

honestly) or have a high level of death threat may be prone to poor

outcomes. The work (Bowlby 1979) relating grieving styles to

personality styles has been scant. Further research is necessary

to view the interactions of personality variables with the patterns

of grief.

The processes, types, outcomes and conditions of grief have now

been described. It is now necessary to turn to the discussion of a

particular kind of grief, the grief in reaction to the perinatal

death, especially stillbirths. Since it is apparent that grief occurs

when an attachment has been severed, especially through death, the

issue of grieving the loss of a stillbirth becomes critical when out-

lining the earliest maternal "attachment" or love for a child. This

area has not been widely noted or researched but understanding grief

at stillbirth and fetal death may lead to discovery of emotional and


psychological factors surrounding both the birth and death of a child.

Before tying together the issues in this paper, the literature on

grief at the death of a stillborn will be reviewed.


The tragedy of stillbirth is a quiet tragedy. While each year

thousands of families in America experience the loss of an infant at

birth, these deaths, and the resultant grief, are seldom acknowledged in

our culture. When acknowledged, the stillbirth is not regarded with the

same respect as other deaths. The parent's grief and mourning are

often felt to be "abnormal" since others assume that no attachment

existed between the parent and child before birth. A growing body of

literature, however, has now documented the existence of grieving

processes in response to stillbirth. While very little research has

been done with these grieving parents, there can be no doubt that the

normal and appropriate response to a stillborn infant is one of intense

grief and mourning, just as with the death of any loved person.

Recognizing that grieving occurs in parents of stillborn infants,

attention may be turned to the aspects of grief, both similar to other

bereavements and particular to perinatal loss. This discussion focuses

on processes in perinatal bereavement, pathological variants of peri-

natal grief, and the conditions and interventions affecting the outcome

of grief at stillbirth. Suggestions are made for further research.

Description and Processes

When referring to stillbirth, two types of events may be delineated.

The first type of event, occurring infrequently, is that of fetal death


in utero (FDIU). In this case, the baby's death is diagnosed days or

weeks before delivery. In this case, where the mother has forewarning

of her child's stillbirth, one might classify the resultant grief as

anticipatory. Grubb (1976a, 1976b), however, has noted that in cases of

FDIU, even though a mother may verbalize her knowledge of the baby's

death, she is heavily involved in a process of denial, holcing on to

the most remote hope that her infant is alive. This observation has

been corroborated by the authors and others (Kish 1978). The second and

most common type of stillbirth is when there is no prior knowledge of the

death, i.e., the baby's death is diagnosed during labor and delivery.

This event can be characterized as a sudden death. While it is important

to recognize the differences in these events medically, the grieving

processes set in motion are the same as they both result from the

simultaneous birth and death of the child.

Phases of Grief

Bowlby (1979) and Parkes (1972) have outlined four processes which

occur in normal grieving. These are: 1) shock and numbness, in which

there is a lack of overt reaction; 2) searching and yearning, in which

a person physically and perceptually searches for the lost person either

consciously or "unconsciously" in a painful but futile attempt to recover

the lost person; 3) disorganization, characterized by a general depression

with increased affirmation of the loss; and 4) reorientation, in which

a bereaved person begins to restructure his life and return to a level

of functioning at least equal to the preloss period. While each process

best characterizes the grieving person at a particular time, all may be

present simultaneously.

It is important to remember that grief is a very individual

experience. While it is proposed that the above-mentioned processes

best describe human grief, it should also be stressed that each of

the processes of mourning will be experienced differently by different

persons. Still, the Bowlby-Parkes formulations serve as a useful

framework for understanding parental grief at stillbirth.

Shock and Numbness

Upon the death of the child, most mothers experience at least

some degree of disbelief. There is a tendency to believe that somehow

a mistake has been made and the infant is still revivable (Cullberg 1971;

Taylor and Hall 1977; Scupholme 19781 This phase is usually brief

whereupon the intense pain of yearning and searching begins.

Yearning and Searching

When most researchers discuss acute grief, they are focusing on the

painful, fruitless s arching and yearning for the dead person. As Klaus

and Kennell (1976), Jolly (1976), and Davidson (1977, 1979) have noted,

the yearning and searching is compounded in stillbirth mothers by the

nature of the situation. The expectations of mothers and fathers are

of an idealized infant, created from their hopes and dreams. While

all parents suffer some discrepancy between the ideal infant and their

real infant, these discrepancies are usually not too difficult to

reconcile in a normal outcome. Stillbirth parents, however, suffer the

worst discrepancy--not only is their real infant obviously different

from their ideal infant, but death, one of life's greatest sorrows has

occurred at precisely the moment in which the opposite, joy at birth,

was expected. Furthermore, pregnancy as a "life crisis" (Bibring et al.

1961) takes place over a relatively short time thereby focusing the

attention of family and friends on the event of the birth. The sudden

horror of the death is felt by all who attend it. At the time when a

mother may most want to hold her infant, there is no infant to hold,

and yearning of grief is painfully compounded. The untimeliness and

sense of utter injustice are as strong as with neonatal death.

All the signs and symptoms of this phase cf grief described in the

literature on adult mourning have been noted in the stillbirth mother.

These include preoccupation with the image of thoughts of the

deceased, anger and reproach, guilt, despair, sleep disorders,

appetite disorders, "pangs of grief," somatic distress, depressions,

hallucinations and illusions of the presence of the deceased (Kennell,

Slyter and Klaus 1970; Wolff, Nielson and Schiller 1970; Cullberg 1971;

Klaus Kennell 1976; Davidson 1977, 1979; Kennell and Trause 1978).

In the months that follow the baby's death, certain aspects of the

yearning-searching phase may be so pronounced as to cause problems in

later adjustment. Anger, reproach, and guilt play a particular role

in this type of loss. Wolff et al. (1970) noted that of 50 mothers,

17 blamed themselves for the death, 10 blamed God, and 9 blamed others--

doctors and husbands in particular (14 voiced no opinion). This anger

and blame is intensified by the fact that in approximately 70% of

stillbirths, no discrete cause of death can be discerned (Donnelly 1979).

The guilt involved can be overwhelming, as the mother goes meticulously

over the events of her pregnancy, trying to discover any etiological

factors. Failing this, she may turn her feeling outward. Anger and

guilt, then, while normal components of grieving, may be particularly

pronounced in parents of stillbirths.

Parents who did not see or hold their baby may have particular

difficulty in the phase of yearning and searching. When there is no

clear perception of the baby, the searching may go on endlessly, for

it cannot be mitigated by defensive processes. Women may report hear-

ing phantom crying when no infant is around (Davidson 1979). Mothers

have told the authors that they still feel the baby move even after

its birth. It is also normal for women to feel jealousy at babies of



While the processes of shock and numbness and searching and

yearning seem best to characterize the early months of mourning, the

latter half of the first year of bereavement may be characterized as

a period of disorganization. Most of the work with stillbirth mothers

has focused on the period when the woman is in the hospital and the first

two or three months when she returns to her doctor, so little has been

said about this later period of the grieving process. Davidson (1979)

has explicitly addressed this period of disorganization and depression

in stillbirth mothers. A mother's intense grieving gives way to feelings

of depression, devaluing of self worth, and apathy. These characteristics

have also been noted by Cullberg (1971).


Reorganization indicates that a mother has adequately resumed her

place in society with minimal discomfort to her. The first step towards

reorganization for many women seems to be the point at which they are

able to settle the affairs of the lost infant, to dismantle the nursery,

with tears perhaps, but not great anguish. No parent severs the tie

completely with the stillborn son or daughter, but a time arrives in

which attention may be focused on the living. A complete resolution

means she may decide to become pregnant again, not to replace her lost

child, but because her mourning has been fully expressed and she can

accept a new child in its own right. These statements are true for

most patterns of mourning a stillborn child. However, because of the

nature of the death, and lack of societal support for the grief of these

parents, severe problems may ensue leading not to reorganization but

pathologic variants of grief.

Pathological Variants of Perinatal Grief

It has been observed that families of stillbirths are at risk for

pathological (Helmrath and Steinitz 1978; Lewis 1979) outcomes. These

have been empirically studied only in mothers. Pathological outcomes

may be divided into two categories: chronic intense grieving, and

absent or delayed grieving.

Chronic Grief

The most typical presentation of chronic grief is depression.

Jensen and Zahourek (1972) found 6 out of 10 patients followed at

intervals throughout the first year of bereavement were significantly

depressed. Cullberg, in a study of 56 Swedish women experiencing

stillbirth, found a variety of serious psychological symptoms in 19

of the women one to two years after their baby's death. These symptoms

included 9 reports of anxiety attacks, 3 severe phobias (e.g., cancer

or death), 3 deep depressions, 2 cases of obsessive thought and 2 cases

involving psychotic reactions (Cullberg 1971). Other authors have

likewise noted an increased risk for serious psychological problems in

stillbirth mothers (Giles 1970; Lewis 1971; Davidson 1979).

Delayed or Absent Grief

Perhaps a more dangerous risk and also one that is more difficult

to detect is the effect of grief which is pushed aside because it is too

difficult with which to deal. This is the case of absent or delayed

grief. Cullberg (1971) found that women who demonstrated a suppression

of feelings about the stillbirth showed more prolonged psychological

symptomatology than those who expressed their feelings.

Perhaps the most common variant of delayed or absent grief which

occurs in stillbirth mothers is the rush into another pregnancy, usually

within a year of the loss. Cain and Cain (1962) described a "replace-

ment child syndrome" occurring in reaction to any bereavement a mother

might experience. Mothers frequently have been encouraged to become

pregnant in order to "forget the loss" by becoming busy with another

child. The problems in both mothers and children are painfully evident:

the mother never having worked out her original grief, searches for her

lost child and "finds" him or her in the replacement child. The

replacement child, however, is constantly compared with the idealized

deceased child and therefore lives in the shadow of the dead child,

often incapacitated by death phobias and fears of abandonment. Some-

times replacement children are held responsible for their sibling's

death and live in a hostile-dependent environment with their parents.

In any case, becoming pregnant to resolve a loss appears to be a pseudo-

resolution detrimental to all parties involved.

The replacement child strategy of coping with grief is seen in

stillbirth bereavement probably more frequently than in any other case.

Kennell and Trause (1978) basing their opinions on the psychoanalytic

(but non-empirical) work of Deutsch have claimed that a woman loses part

of herself in childbirth. While they do not recommend that women rush

into pregnancy, they claim that the reason so many women do is out of

a need for wish-fulfillment to have a baby. However, it is more

commonly observed that parents of stillbirths mourn for the particular

baby they lose, not just for the wish of a child. As for losing a part

of one's self, this is a feature of all bereavements and is not

peculiar to perinatal loss.

There are other problems involved when a family fails to mourn

a stillborn child. Emmanuel Lewis (1971, 1979) has described how

siblings of stillbirths often become involved in destructive fantasies,

especially when the mother acts with the irrational hostility of grief.

Lewis and Page (1978) describe a case in which a woman became pregnant

shortly after experiencing a stillbirth and subsequently was unable

to care for her new infant. Perhaps the best evidence of detrimental

effects of delaying grief by becoming pregnant comes from the work of

Rowe and associates (1978). They found that in a sample of 26 still-

birth mothers followed between 12 and 20 months, the only predictor of

morbid grief reactions was the presence of a surviving twin or sub-

sequent pregnancy within 5 months of the loss. Jolly (1976) has like-

wise warned against the attempt to replace the dead infant with another

child. Both Jolly and Lewis (1971) have also noted problems with

anxiety in subsequent pregnancies. While Wolff et al. (1970) have

suggested that the decision to become pregnant quickly was an individual

matter that the physician should not try to influence, the existing

evidence suggests replacement pregnancies may have severe consequences.

A mother and her family can develop any pathological variant of

grief which may develop as a result of any bereavement. What has been

described here are particularly evident patterns. Those who are interested

in additional information on the outcomes of mourning and their predictors

are referred to the works of Bowlby (1979) and Parkes (1975).

Conditions and Treatment in Stillbirth Bereavement

The conditions surrounding parents and treatment which parents

receive, especially in the hospital, are believed to have a significant

effect on how they will eventually resolve their grief. Cohen and

associates (1978) have described crisis intervention with stillbirth

parents as "assisting parental affirmation." This process of affirmation

is what Parkes (1972) has called "realization." The parents of a still-

born child probably have one of the hardest times of any bereaved

adult dealing with the reality of the death and the permanence of changed

expectations which it entails. The inability to fully realize death

and accept the consequences is believed to predispose bereaved persons

to pathological outcomes (Bowlby 1961, 1979). In the parents of a

stillbirth, there are several conditions which may effect this affir-

mation process. They may include the reaction of others to the still-

birth (especially doctors, nurses, and families), whether or not the

mother or father saw or held their infant, whether a funeral or memorial

service was held, whether autopsy results are received, whether siblings

or close subsequent pregnancies are involved (discussed previously),

previous losses, whether or not they receive appropriate crisis

intervention and information concerning the grieving process, and lastly

gestational age.

When a mother gives birth to a stillborn infant, the reaction of

others, especially doctors, nurses and family members may influence the

processes of grief. Until recently, society expected that a young

mother would not grieve for a stillbirth, and doctors and nurses shared

this attitude. It was not uncommon to hear the comment, "Don't worry,

you can have other children," immediately after the delivery. Doctors,

nurses, and families, being as uncomfortable with grief as anyone else,

would avoid all mention of the death making it into what has been called

a "nonevent" (Lancet 1977). Mothers were almost never allowed to see

their infants for fear they would be unduly upset, as if they were not

already. The attitutudes of hospital staff are changing, but very

slowly (Kowalski and Bowes 1976; Helmrath and Steinitz 1978; Rowe et al.


In 1971, Bourne conducted a study on the psychological effects

of stillbirth on the doctor. He sent out questionnaires to 100 randomly

selected doctors of mothers who had experienced stillbirth and 100

doctors of mothers who had live births in order to look at differences

in reactions between the groups and differing doctor-patient relation-

ships. This report was startling, for the doctors reported no

differences in the maternal reactions to stillbirths or live births

or in anxiety levels between the two groups of mothers in subsequent

pregnancies. Bourne concluded that doctors had either neglected real

differences or could not deal with the grief at stillbirth. Signif-

icantly more stillbirth doctors than live birth docotrs did not return

the questionnaire, and among those stillbirth doctors who did, signif-

icantly fewer questions were answered at all; most responses were

reported as "don't know." He concluded that doctors were subject to

inordinate stress and patients were in danger of neglect when a still-

birth occurred (Bourne 1971).

Corroborative evidence has been given by Wolff and his colleagues

(1970) in a longitudinal study of stillbirth mothers. They found that

over 50% of mothers perceived doctors as cold or indifferent to them

during their bereavement. Over 60% of the mothers rated nurses as cold

or indifferent. Rowe et al. (1978) had similar findings; 60% of

stillbirth mothers felt dissatisfied with the information offered

them about the death and the manner in which the information was


While staff members may have a detrimental influence on the mother's

well-being, they may also have a facilitating effect on the grieving

process with supportive intervention. Rowe et al. (1978) found that

stillbirth mothers who had followup were considerably more satisfied

with their treatment. Schreiner, Gresham, and Green (1979) found that

a simple, caring, phone call from a physician accounted for a reduction

in reports of major problems in the intervention group when compared

with stillbirth mothers who had received no phone call. The nursing

literature is likewise replete with examples of how nursing responses

may greatly facilitate the grieving process (Seitz and Warrick 1974;

Saylor 1977; O'Donohue 1978). Queenan (1978) has remarked that the help

one can be to the family of a stillborn child is often underestimated

and he encourages both doctors and nurses to play a central role in

the support of these parents.

One of the most controversial issues is whether or not the mother

should see or hold the baby. Yet, in the literature, there is almost

unanimous agreement that seeing and holding the infant is a helpful

factor in successful grief resolution (Kowalski andBowes 1976; Klaus

Kennell 1976; Tizard 1976; Cohen et al. 1978; Saylor 1977; Kennell and

Trause 1978; Scupholme 1978; Lewis 1979; Davidson 1979). The only

research widely known which has touched on the effects of seeing the

baby was by Kennell et al. (1970) in which seeing the infant was

associated with full expression of grief. Kellner and Kirkley-Best

(1981) have found that holding the infant was significantly more

frequently associated with having a funeral and burial rather than a

hospital disposal (cremation). This may indicate that mothers who

hold their infants are more willing to deal with the painful reality

of the death and treat it as they might the death of an older person,

or it may mean that holding the infant emphasizes the reality to

mothers, facilitating more beneficial grieving processes. Further

work is needed to define these affects. In general though, holding

and seeing the infant seem beneficial. When a mother cannot or has

not seen her stillborn infant, artifacts such as footprints, photo-

graphs, and apparel become critically important to the mother. These

have likewise been observed to have beneficial effects (Klaus and

Kennell 1976; Davidson 1979).

The same authors, cited previously, who encourage parents to see

their stillbirths also note the positive effect of an autopsy. According

toQueenan (1978), Cohen et al. (1978), and Kellner and associates (1981),

the autopsy seems to allay guilt and anxiety in parents of stillborns.

Even when no definite cause of death is found emphasis on the baby's

normality seems to alleviate a great deal of parental concern. A

caring explanation of the results serves to increase communication and

trust between parents and their doctors. These effects have yet to

be studied in depth.

Intervention Programs

In the last few years, a variety of intervention programs have

arisen to offer support and information to families experiencing

stillbirth. Most of these programs include in-hospital visits and

followup of parents at regular intervals. Some intervention programs

are run by parents of stillbirths themselves (such as PEND, Parents

Experiencing Neonatal Death, Klaus 1980) and some are individuals

(e.g., psychiatric nurses). Still others are bereavement "teams" (such

as the Perinatal Mortality Counseling Program at Shands Hospital in

Gainesville, Florida; Kellner, Kirkley-Best, Chesborough, Donnelly, and

Greene 1981). Most programs involve the same sort of treatment--a mother

is supported and encouraged to express her feelings, she is offered

full options in regard to her infant (seeing the baby, photographs,

etc.) and she is followed up throughout her bereavement period. (For

a full description of procedures, see Kellner et al. 1981). These

programs seem to have a beneficial effect on the grief of families

involved. Further research may lead us to the most appropriate methods

of crisis intervention.

The gestational age of the infant as a factor in perinatal

grieving has never been systematically explored. Almost all of the

research which has been conducted in the area of perinatal grief has

been with women losing an infant in the third trimester of pregnancy or

the first weeks of life (Kennell et al. 1970; Cullberg 1971; Rowe et

al. 1978). Jolly (1976) has indicated that parents losing a one pound

infant should receive the same regard as those losing full-term babies,

but has not specifically observed how early grieving responses occur.

One may get some idea that they occur very early based on scattered

reports from different disciplines which have found that grieving

responses occur towards early fetal death in adolescents (Horowitz

1978) and in an adult sample undergoing second trimester abortion (Pasnau

and Farash 1977). Exactly when women begin to respond with an intense

sense of loss, and the variables affecting the loss, has not been

studied. Pepper and Knapp (1980) explored grief in reaction to three

kinds of loss: miscarriage, stillbirth, and neonatal death. They con-

cluded that there was no difference in reactions to the three losses.

Their methodology was unfortunately too poor to draw this conclusion.

They divided the women into groups according to self-report about type

of loss which is often an unreliable procedure. Women do not always

distinguish among types of perinatal loss except at extremes, and

neither does the public (Kirkley-Best 1981). Further, grief scores

were determined by self-report questionnaires asking for retrospective

accounts of how they had felt at the time. The length of time since

the loss ranged from 6 months to 30 years. The authors chose to

interpret the lack of significant differences among groups as evidence

of the same reaction to all types of loss This strategy would entail

accepting the null hypothesis when no acceptance-of-the-null design was

offered. The lack of difference may be well argued to be due to the

methodological flaws previously mentioned.

Conclusions and Areas for Research

For reasons not entirely apparent, research in this field has been

surprisingly scant. Much of the research which has been published and

which is widely known is filled with statistical errors, design in-

sufficiencies, and rash conclusions drawn on the basis of very little

data. This is not to say that the conclusions drawn from the research

have been incorrect, but rather that more well controlled studies are

needed. Many of the studies on stillbirth have focused on grief only

in terms of Lindemann's (1944) symptomatology or Kubler-Ross's "stages"

(1969). The picture of grief is much more complex than this and persons

working with grieving mothers need to more fully educate themselves in

the literature of grief and loss. At present there is an overwhelming

need for careful research to enable an understanding of the particular

aspects of stillbirth bereavement which affect these mothers and

families. Without proper study, professionals are destined to follow

the fashions of the popularized literature on grief without appropriately

meeting the needs of families of stillbirths.

Statement of the Problem

When human attachments are in some way severed the result is grief

regardless of the form it may take. Grief is used definitionally to

describe the psychophysiological reaction to loss of an attachment

figure. Observations of grieving reactions must necessarily lead us to

the conclusion then, that there was a previously existing attachment.

Applying this principle to prenatal maternal attachment, if pre- or

perinatal grief is observed, we have evidence of an existing attachment

in pregnancy at least unidirectionally.

The central hypothesis in this study is that there is a maternal

attachment toward the child in pregnancy hwich increases in magnitude and

quality with the gestational age or length of pregnancy. Evidence for

this proposition should come from observations of grieving reactions to

prenatal loss which should also increase in quality and magnitude with

length of time in pregnancy.

Gestational age (GES) is proposed to affect grief in large proportion

if a developing attachment is occurring. Other variables which are felt to

affect grief in general are felt to affect perinatal grief, although not as

greatly as gestational age. These variables include demographic variables

(age, race, socioeconomic level [SES], education level [Educ], and marital

status), whether the pregnancy was wanted (Want), concurrent stressful

life circumstances (Life), number of previous losses in pregnancy (ABS),

number of children (Kids), size of the mother's family measured by number

of her siblings (FamSze), number of previous bereavements (Loss), sex of

the baby when determinable (Sex), and whether the baby was seen or held


Demographic differences were predicted to be nonsignificant. If

grief as a universal psychophysiological state is being measured, then

demographic variables should not come into play as determinants of the

level of grief. Whether the pregnancy was wanted should likewise be

nonsignificant if an attachment occurs normally in the course of pregnancy.

Life circumstances such as divorce, death and son on, has been predicted

to intensify reactions. Likewise, with progressively more losses in

pregnancy, a higher grief rating is predicted. The number of children

a mother has is thought to lessen certain aspects of the grieving process

such as caring about one's own life, so the scores are expected to be

attenuated somewhat from cases of first pregnancy losses. Size of the

mother's family based on siblings is predicted to be uncorrelated with

ratings of grief. Previous bereavements were thought to be associated with

grief score, with a slightly negative correlation. Sex of the baby is

predicted to be uncorrelated with ratings of grief. Seeing and holding

the baby, while confounded with gestational age, should be associated with

higher grief ratings, based on previous work (Kennell, Slyter and Klaus


Individual differences probably play a great role in perinatal

grief as in perinatal attachment. Measurement of individual differences

is not attempted in the present study because of the delicate nature of

grief, and the confounding of results in administering measures of

individual differences after the fact of the loss.



Thirty-eight subjects were recruited for the study. All subjects

were women experiencing spontaneous abortion (miscarriage), fetal death

and stillbirth, treated at Shands Teaching Hospital, Gainesville,

Florida. Solicitation was done in person. Approximately 70 persons

were contacted in the hospital within 24 hours of the loss. Of these

original contacts, 41 returned in four weeks to their post-partum appoint-

ment at which time their participation was requested. There was no

differential return rate of women at different gestational ages. While

one can only guess at reasons for poor return rates, the nature of the

population and emotional factors may have contributed. Shands Hospital

is a regional care center and many clients come from great distances.

Transportation difficulties are often responsible for non-returns to the

post-partum clinic, both in cases of fetal losses and healthy pregnancies

as well. Since mothers have usually attended the clinic throughout their

pregnancy it may be particularly difficult for some of them to return to

the painful reminders associated with the times in which the infant was

alive. Mothers who may have seen the event as a minor one may have

not returned because they did not feel it was critical, and may have

considered cost and/or time factors. It must be emphasized however

that all the reasons mentioned entail speculation, and may or may not


have influenced anyone's decision to return. Only one woman preferred

not to be interviewed. Two women were not interviewed because of past

histories of psychopathology. The age range was 16 to 42, with a mean of

23.76 and a standard deviation of 6.13. Fifty percent of the women were

primiparas. Of women who had children, the mean was 2 with a standard

deviation of 1.4. The mean education level (number of years) was 12.06

with a standard deviation of 1.89, and a range of 6 to 16 years. Most

subjects were of middle to low socioeconomic status. Sixty percent of

the sample was black. Twenty-nine percent had had one or more previous

losses in pregnancy. Of those who had had previous losses in pregnancy,

the mean was 1.5 with a standard deviation of .82. Forty-five percent

were married, 50% were single and 5% were separated, divorced or widowed.

The average family size from which the mother hailed was 5 siblings with

a standard deviation of 2.8. Forty-seven percent of the population had

experienced a previous death of a close family member with a mean and

standard deviation of 1.44 and .68 respectively. Thirty-two percent

of the women interviewed were in the first trimester, 39% were in their

second trimester and 29% were in the final trimester, with a mean of

20 weeks and a standard deviation of 10.7. All participants were inter-

viewed privately. Participants signed an informed consent statement

before participating in the study.

Procedures and Materials

Within 24 hours of hospitalization all women were contacted by

the principal investigator. Women received routine supportive care

from the Perinatal Mortality Counseling program during the in-hospital

stay. All women were given appointments before leaving the hospital

for a post-partum check-up in 3 to 4 weeks. While subjects waited to

see the doctors at their post partum exam, they were invited to

participate in a structured interview dealing with their experiences

over the past few weeks. With the participants' permission, the inter-

view was taped and identified only by subject number. The length of

the interviews ranged from 20 to 45 minutes. The principal investigator

conducted each interview, asking each person the same questions in the

same order. When the interview was concluded and the tape recorder

turned off, the participant was assured of the normality of all her

responses. The purpose of the study was briefly described. Each

participant then received further, routine care with members of the

Perinatal Mortality Counseling Program. Subjects were offered a copy

of the signed consent form and were informed that they might withdraw

their participation at any time. A copy of the consent form may be

found in the Appendix (Appendix I).

After the interviews were taped, they were edited to remove all

references which subjects made regarding their point in gestation at the

time of the loss and whether or not they had experienced fetal movement.

The tapes were then rated by a trained rater, blind to both gestational

age and the hypotheses in the study. The tapes were rated on a series

of 10 characteristics of grief (Kirkley-Best 1980), on a scale of 1 to

5, in which "1" indicates "symptoms not present" and "5" indicates

"marked presentation." An inter-rater reliability measure was used to

verify ratings. Ten tapes were randomly selected and rated by a second

rater, an obstetrician, experienced with perinatal bereavement. Inter-

rate reliability was r = .78 p < .005. Raters had no access to infor-

mation central to the hypotheses or ratings from the other person.

Those interviews for which two scores were obtained, received a rating

equal to the average of those two ratings.

Training of the Raters and the Rating Scale

Raters were not told of the actual hypotheses involved in thestudy.

Before any tape was actually rated, raters were met and methods of rat-

ing were discussed. Each category of ratings was gone over in detail.

Raters were given examples of possible responses which might arise.

Raters were told to take each category as a separate unit, and not to

let ratings from one category influence ratings from another category.

Raters were instructed to focus on the emotional content of the response

in addition to the verbal content. Problems in ratings were discussed,

including what constituted extreme scores. Raters were told to use a

.5 rating between ratings if they felt unsure of their response. The

tapes were to be listened to once thoroughly before the rating process

began. For the first ten interviews, each individual rating was

discussed and the trained rater was asked to explain why the rating

was chosen. Only two categories on different interviews were in question

in this initial 10. Rather than point out the category, the rater was

asked to re-rate the two complete interviews. The categories were changed

based on the general discussion in this case of the difference between

minor disturbances in sleep and major disturbances, and between anger

related to grief and anger which is connected to other tangible life

events. The high inter-rate reliability established supports the

consistency of observation. The specific scales used for rating may be

found in Appendix II.

The structured interview contained questions about general somatic

distress, appetite and digestive disturbance, sleep, concentration and

restlessness, preoccupation with thoughts and images of the deceased,

anger and reproach, guilt, depression-despair-depersonalization, and

caring about one's own life. Questions included in the interview may

be found in Appendix Ill. An intercorrelation matrix of scales of the

rating system with each other and with total grief score may be found

in Appendix IV. Data sheets were collected on each subject drawing

information from medical charts and self-reports of the subjects.

Information obtained included number of previous bereavements, previous

loss in pregnancy, size of themother's family, religion, felt fetal

movement, whether the baby was wanted and seen or held, and a variety

of demographic variables. A blank copy of the data form may be found

in Appendix V.

The mean rating and standard deviation (SD) for the somatic scale

was 2.31 and .978 respectively. The mean and standard deviation for

appetite disorder was 2.33 and 1.38. Sleep had a mean rating of 2.14

and a 50 of 1.25. For concentration the mean rating was 2.46 and 1.02;

preoccupation had a mean and SD of 3.53 and .91. Ratings of anger saw

a mean and SD of 2.3 and 1.23, while the guilt rating averaged 2.4 with

a SD of 1.0. Depression had a mean rating of 2.52 and a SD of 1.21.

Depersonalization and time sense had a rating of 1.85 with a SD of .83;


caring about one's life received a mean rating of 2.51 and .949 SD.

These figures, along with the range for each scale, may be found

summarized in Appendix Ill.


The scales of grief ratings were found to be consistent with the

total grief score. The mean total grief score was 24.4 with a standard

deviation of 7.9. The scores ranged from 10 to 42, out of a possible

50 points. A table of the range, means, and standard deviations of each

scale may be found in Appendix VI.

Because of the relatively small sample size for regression, only

4 variables were chosen for entry into a regression model. Choosing 4

variables for the model was based on the principle of using approximately

10 subjects per term in the model (Kleinbaum and Kupper 1978). While

regression procedures are seldom used with fewer than 60 subjects, the

technique is deemed appropriate here for several reasons. While 60

subjects is a standard group size, predictive models resulting from such

a size are often used to estimate responses of persons in general. In

other words a 60-subject sample is still a relatively small sample size

with most reference populations. An absolute sample of 38 is smaller

than an absolute sample of 60; however, 38 subjects out of all who

experience stillbirth or miscarriage in a year is proportionately a

much larger sample than, say, 60 general psychology students out of a

population of college-age persons. Of more concern than sample

representativeness is Type II errors and wide and therefore less precise

confidence intervals. These concerns are allayed though, by a demonstra-

tion of a significant and apparently stable model. Additionally, with

severe violations of assumptions such as normality and heteroscedasticity

standard results do not differ appreciably from the more conservative

non-parametric procedures which offer even further risk of Type II

errors (Glass, Peckham and Saunders 1972).

The criteria for choosing the 4 variables of interest for the

regression model included the following considerations. Gestational

age, the critical independent variable, was retained without consideration

because of its centrality to the hypothesis of increased grief scores

with increased time in pregnancy. Other variables were chosen according

to their inclusion in the original hypothesis, and their independence of

gestational age and one another. Two highly correlated variables in a

regression model result in great difficulty in demonstrating the signifi-

cance of either variable. To achieve information about correlation among

variables, all variables of interest were placed in a correlation matrix.

The zero-order correlation matrix of all possible variables may be found

in Appendix VII. The variance which fit all criteria for inclusion were

gestational age, number of children, number of previous spontaneous

abortions or fetal deaths, and degree to which the baby was wanted.

The analysis used was the Statistical Analysis System's (SAS) Step-

wise Max/R regression technique. This procedure places each inputed

independent variable in the model according to the maximum amount of

variance explained, or the maximum amount of increase in variance explained.

With four inputed variable, Stepwise Max/R produces the best 1 term model,

2 term model, and so on until additional variables add less than .5

percent to the amount of explained variance.
In the present study, for the best 1 term model, R = .33, F =

12.36, p < .0012. Using Type II sums of the squares, term selected in

this model was gestational age, with S1 = .38, F(1) = 12.36, p < .0012.

In the best 2 term model, R2 = .39, F(2) = 9.44, p < .0005. The first

term in the model was again gestational age; the second variable was

number of children. For gestational age, l1 = .32, F(2) = 9.58, p <

.0039. For number of children, B2 = -1.8, F(2) = 5.08, p < .03, using

Type II sums of squares. Neither the number of previous pregnancy losses

nor the degree to which a baby was wanted contributed more than .5 percent

of additionally-explained variance, so the stepwise procedure concluded

with the 2 term model.

Two scores were suspected as outliers (an unusual score obscuring

trends in the data). When these outliers are removed, there is a

significant rise in the amount of variance accounted for, yet the

patterns of variables stay the same. With outliers removed, R2 for the

I term model (gestational age) is .37, F(1) = 19.5, p < .00. The 2 term

model, (gestational age and number of children) increases R2 to .41,

F(2) = 11.19, p < .00. A summary of the regression data and the models

may be found in Appendix VIII.

Variables which were predicted correctly include race, education

level, SES, and marital status which were not significantly associated

with grief score. Whether the pregnancy was wanted, number of the mother's

siblings and sex of the baby were likewise unassociated with grief score

as had been predicted. Those variables which were correctly predicted

to be significantly associated with grief score include gestational age

(positively correlated), number of kids (negatively correlated), seeing

the infant (positively correlated) and concurrent stressful life

circumstances (positively correlated). Incorrect predictions included

age (predicted as insignificant, but positively correlated with grief


score) number of previous losses in pregnancy (predicted to be positively

correlated but was non-significant) and number of previous bereavements

(predicted to be negatively correlated, but no significant relationship

arose). For the exact values, refer to Appendix VII.


Grief and attachment were proven to be two of the most difficult

of human processes to measure. In this study, as in others, a grief

score has been used as the dependent measure. Assigning a score to a

person's grief must be viewed in any study as only a very rough approxi-

mation of the amount or quality of the grief experienced. The use of a

trained independent, blind rater of interview responses has been used in

the present study to circumscribe unreliabel self reports which often are

based on the degree to which a grieving person trusts the person adminis-

tering a grief instrument. The flexibility of this method has allowed

ratings of not only verbal content, but of emotional content as well.

The interview which was used in the study appears to be consistent

and fairly reliable in its measurement of grief. Based on the positive

correlations between each scale and the total grief score, inter-scale

reliability is at least preliminarily established. Most of the scales

produced an average rating of about 2.5 with a standard deviation of

roughly I, with two notable exceptions. These exceptions were pre-

occupation with thoughts and images of the deceased infant and deperson-

alization/time sense.

Preoccupation tended to receive higher ratings than all scales of

the interview. Women in the interview spent by in large a great deal of

time thinking about the loss whether or not all of the emotional

components of grief were present. Women who lost the baby after quick-

ening mentioned very frequently that they still felt the baby move,


although they universally recognized the movement as a disturbing

distortion of reality. Even mothers early in pregnancy were likely to

spend a lot of time thinking about the miscarriage, the pregnancy, plans

for the future and what the infant would have been like. The higher

mean score on preoccupation scale probably indicates the importance

of fetal loss (and pregnancy as well) as a life event, whether or not

the loss evokes full fledged grief reactions.

Depersonalization/time sense received a much lower mean score than

the other scales. One may attribute this score to a somewhat less

reported phenomenon (although it is correlated with grieving reaction)

or to the difficulty which arises in articulating the experience. Both

reasons may exert an influence although persons struggling with more

intense grief reactions, seemed immediately to understand and identify

with the phenomena of depersonalization when described. No one received

a score indicating chronic depersonalization.

No factor analysis of the scores was attempted due to so few

subjects in the sample. Clinical impressions indicate that the classifi-

cations seem appropriate, although a statistical investigation seems to

be warranted in future research.

Of the two variables which have accounted for significant portions

of the variance in the model, gestational age, as had been predicted,

accounted for the most variance, or in other words, showed the greatest

degree of relationship with the grief score. The pattern of relationship

is one that would suggest a steadily increasing grief with losses later

in pregnancy. There is not a leveling off of scores after quickening

(about 20-25 weeks) as had been hypothesized by other investigators (e.g.

Taylor and Hall 1979). Quickening itself may be only one aspect or

milestone of attachment of which there are many. This increasing grief

intensity with later fetal death suggests conversely, a growing attach-

ment process in pregnancy which does not, as some have suggested (Rubin

1975, Bibring 1961, Campbell and Taylor 1979) level off in the third

trimester. This is not to discount intense grief reactions to early

loss which of course can and do occur. Early intense grief may be the

product not only of individual differences but of other circumstances

as well which are very difficult and/or unethical to measure in early

grief. The impression which emerges based on interview responses is

one in which grief responses to loss in the first trimester are indicative

of grief at the loss of future plans and of the "pregnancy," whereas

grief at later loss is grief over the severing of an attachment to a

particular child, an attachment already well-formed at birth. Bonding

at birth is proposed to be another milestone in the realization of the

child as an individual son or daughter, but bonding is not the point at

which parents become attached to the child or not, attachment being a

process already having long begun.

The number of children a mother has is negatively correlated with

grief score accounting for additional, but less variance than gestational

age. The number of children a mother has made a difference especially on

some aspects of the grief scale. Caring about one's own life was particu-

larly evident as a distinguishing scale, in which mothers with other

living children were more prone to want to put their lives back together.

Several mothers reported that if it were not for their child they would

feel they could not go on. This pattern suggests that first losses may

particularly difficult. Attachment processes in the first pregnancy may

also be of critical importance and needs to be further researched.

The two variables in the regression model which contributed almost

no additional explanation were the number of previous pregnancy losses,

and whether or not the baby was wanted. The lack of findings on previous

pregnancy losses may be an artifact of the difficulty of quantifying the

meaning of events to the individual mother. Based on interview responses,

it appears that for some women, a first loss in pregnancy may be over-

whelming, while for others, the first loss is taken as "something that

just happened." This is especially evident in early loss where for some

women the first or second spontaneous abortion may be taken in stride,

but the third, for example may constitute a loss of not only a pregnancy

but also of the ability to be a mother. Without adequate measures of

individual differences, these relationships may only be estimated. The

degree to which the child was reported as wanted was predicted to have

been non-significant. The matrix of zero-order correlations among

variables suggests that while marital status is associated with whether

the baby was wanted, the grief score was not significantly associated with

the degree to which a baby was wanted. While a sample size of 38 does

not permit the ability to unalterably accept the null hypothesis and

conclude no sort of relationship between these two variables at all,

the prediction of the non-significance of the "wanted" variable permits

more leeway in interpretation. One must be extremely cautious however

in interpretations based on this sample size. Oddly, if there is any

trend at all in the degree to which the infant was wanted and the grief

score, there was a slightly negative one. Based on interview responses

and the pattern in the matrix a conclusion may be drawn. While many

women find themselves in circumstances which contribute to not wanting

the pregnancy, very few if any report not wanting the baby. The

ambivalence which arises in these mothers is clearly evident. Caretakers

would do well to recognize this dilemma in mothers contemplating

alternatives such as elective abortion and adoption, and weigh carefully

the emotional cost to the mothers. While many factors may arise which

make a pregnancy undesireable, the development of a relationship with the

unborn child progresses in spite of poor circumstances. The intimate and

inextricable physical relationship between mother and developing fetus

promotes a steadily progressing attachment regardless whether a mother

prefers the pregnancy. The investment of self in one's offspring (especially

first offspring) may come into play as well, although until more infor-

mation is obtained little can be surmized.

Evidence for grief (and thereby attachment) as a universal, human

process, is suggested by the predicted lack of findings on demographic

variables. Race, socioeconomic status, and education level all show no

relationship with grief score according to the matrix of zero-order

correlations. Only age showed a significant negative correlation with

older mothers showing scores lower than those of younger mothers. Since

age is also positively correlated with the number of children a mother

has, the relationship cannot be attributed to an aging process alone.

The exact effects of age need further study. Demographic variables are

known to have affects on the outcome of the grieving process (Kirkley-

Best 1980); however, the actual psychophysiological process of grieving

is thought to be universal. The lack of findings with so small a sample

size must be approached with caution, although this lack of race,

education, and socioeconomic differences is in line with predictions and

is consistent with the view of grief as a universal phenomenon.

The regression model (with outliers removed) including the predictors

of gestational age and number of children accounts for over 40 percent pf

the score variables. When the regression model is used to aid prediction,

an R2 of .41 is not considered appropriate. (For example, if college

grade point average was necessary to predict from SAT scores and high

school average, one would want even greater precision, or an ability to

explain 80 or 90% of the variance.) The model here was employed primarily

for description of the variables affecting grief in pregnancy. For

descriptive purposes, we have accounted for much variance by employing

only two factors, time in pregnancy and parity. The remaining bariance

which has not been accounted for would be unwisely attributed to pure

error varaince. Personality and other individual differences probably

account for the largest amount of variance. These individual differences

are not easily measured as personality psychologists will attest, and

to attempt to measure these characteristics in the immediate post-loss

period is of questionable ethics. Grief can only be fully understood

in terms of the context in which it occurs, and its meaning to the

individual. The meaning of grief to an individual may be impossible to

assess precisely. While one can count the number of previous bereavement

experiences a person has had, this number does not yield us with infor-

mation about how those losses were dealth with and resolved or with

personality differences that affected both the attachments and the losses.

Having accounted for so large a proportion of the variance with only two

variables, not directly related to personality variables, suggests a

considerable importance which these factors may possess.


The most central issue to keep in mind when dealing with parents

experiencing a perinatal loss is that no single variable is appropriate

for predicting or understanding an individual's grief. While it has been

demonstrated that the time in pregnancy that a baby is lost is associated

with grief ratings, there is little value in assuming, for example that

early miscarriage will not lead to grief. There are individuals who are

deeply grieved at even two month losses. All the aspects of a person's

life which may affect any grief, affect grief at stillbirth and miscarriage.

The exact pattern of variables associated with grief ratings is more

relevant with regard to theoretical considerations. A steadily increasing

incidence of intense grief at pregnancy loss suggests a fruitful area of

research in prenatal maternal attachment. Attachment factors in normal

outcome pregnancy needs further exploration.

Probably the greatest information gained from this study is that

women grieve at all points to loss in pregnancy even though there is a

discernable rise in intensity of reactions across time of pregnancy.

Based on previous literature and on the results of this study, it is

suggested strongly that further research be conducted to aid these parents

in their grief. Effective counseling and support for these parents will

only come about with careful human research into the variables affecting

perinatal grief.

One area of research which has been almost entirely ignored is the

reaction of the father of the baby to the stillbirth. No one to date has

dealt exclusively with any variables pertaining only to fathers of still-

born infants. Preliminary observations lead to the suggestion that while

reactions in fathers differ widely, they are often just as intense as

maternal reactions and in some cases are even more intense. Anger seems

to be particularly evident in fathers, as does an inability to discuss

emotional responses to the loss. An exploratory study of fathers is a

promising area of future research.

The way a family copes with stillbirth and other fetal death needs

to be explored as well. Children's reactions to stillbirth have been

almost universally ignored except in clinical reports in which destructive

fantasies ensued (Lewis 1977). The family structure in this crisis seems

likewise to be a fruitful area of research, based on previously discussed

risks of divorce and separation in these parents.

Attitudes which persons hold towards stillbirth has begun to be

explored by the author. Findings suggest that there exist a relative

lack of understanding among friends, relatives, and the medical community

about the intensity or even presence of grief at stillbirth. What the

factors are in changing these attitudes needs further research.

Other variables need to be carefully teased apart to view the

significant role each may play in stillbirth bereavement. These variables

have been noted to have affected perinatal grief but have never been

empirically explored. These variables mentioned previously, include

the role of the autopsy, seeing the infant (at all points in pregnancy),

follow-up care, artifacts and moments of the baby, the bereavement

history of families and so forth. The major conclusion which has been

arrived at in this study though, is the importance of taking the intense

grief of these parents seriously, allowing them to mourn their infants,

and recognizing that infant as briefly encountered as a stillborn infant

or miscarried fetus still maintains an important role in the lives of

mothers, for whom the child was a real son or daughter however short the



There has been relatively little research which has addressed

grieving processes to loss in pregnancy. Many persons have assumed that

no maternal attachment occurs in pregnancy and therefore that grief should

not occur at loss. This study has been directed at observations of griev-

ing reactions to losses occurring at all times in pregnancy, and the

variables affecting those reactions. The central hypothesis in the

study was that there is a maternal attachment towards the child in

pregnancy which increases in magnitude and quality with the gestational

age or length of pregnancy. Other variables predicted to affect grief

or not to were concurrent life circumstances, number of previous losses

in pregnancy, number of children, number of previous bereavements and

demographic variables.

Thirty-eight subjects experiencing prenatal loss were interviewed

at one month post-partum regarding their experiences during the previous

month. The structured interview was taped, edited, and blindly rated

on a 10-scale rating system of grief characteristics and symptomatology.

A random sample of 10 tapes were blindly rated by a second rater. An

inter-rater reliability measure of .78, p < .00 was established.

As was predicted when employing SAS Stepwise/MaxR regression

procedure, gestational age accounted for approximately 37 percent of the

variance in the grief score, increasing to over 40 percent with the

addition of number of as the second variable.

The number of previous pregnancy losses and the degree to which

the pregnancy was wanted were not significantly associated with grief

score. Demographic variables as predicted were not associated with


grief score, however stressful life circumstances was significantly

and positively associated with grief score, with r = .45, < .01.

The relationship of grief score with the variables described above

are discussed. Implications for maternal attachment in pregnancy and

directions for further research is offered.


University of Florida Home Page
© 2004 - 2011 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated May 24, 2011 - - mvs