You are on page 1of 14

JOURNAL OF REPRODUCTIVE AND INFANT PSYCHOLOGY,

VOL. 22, NO. 2, MAY 2004, pp. 5769

Maternal foetal attachment and personality


during first pregnancy
B. SJOGREN, G. EDMAN, A. M. WIDSTROM, A. S. MATHIESEN &
K. UVNAS-MOBERG
Karolinska Hospital, Sweden

Abstract This study aims to apply and modify an inventory measuring maternal foetal
attachment (MFAS) for use in a Swedish population, and to relate these data to personality
traits measured by the Karolinska Scales of Personality (KSP). Previously, we found changes
in womens personality traits at childbirth. The attachment and personality profiles were
studied in women expecting their first child at the 13th (n~50, response rate 87%) and 36th
week of gestation (n~50, response rate 76%). The scores on the total MFAS were high,
significantly higher for women tested in the later phase of pregnancy (pv0.01). A factor
analysis revealed a four-factor structure (visualizing motherhood, wonder and worry,
nourishing self and foetus, and relation to foetus and own body), not replicating the five a
priori factors of Cranley. Linear regression analyses showed positive correlation between the
total attachment score and the tendency to feel guilt and the desire for social desirability as
reflected by the KSP. Attachment subscales showed a positive correlation with somatic
anxiety, guilt, psychasthenia, social desirability, inhibited aggression and phase of
pregnancy. Negative correlation with psychic anxiety and emotional detachment was
found. These findings are, to our knowledge, a fairly new contribution to the literature on
maternal foetal attachment.

Introduction
A womans first pregnancy has been described as an important period of mental
preparation for motherhood by many investigators. The psychic distress and challenge
is often emphasized as accompanying this process (Breen, 1975; Uddenberg, 1974).
The emotional bonding between the mother and the newborn has been well studied
(Bowlby, 1988; Klaus & Kennel, 1976). The development of an emotional attachment
to the foetus/future child during the pregnancy constitutes a fairly new field of
research. Psychoanalytic theory has in detail recognized and described the
development of the attachment using a small sample method in an intensive study
of each case, which is similar to the qualitative methods used today. The psychological
progression of the expectant mother from narcissism to object love, from being
Address for correspondence: Berit Sjogren, Department of Obstetrics & Gynecology, Karolinska Hospital, SE171 76, Stockholm, Sweden. Tel: z46 8 51773621; Fax: z46 8 318114; E-mail: berit.sjogren@ks.se

ISSN 0264-6838/print/ISSN 1469-672X/online/04/020057-13


# 2004 Society for Reproductive and Infant Psychology
DOI: 10.1080/0264683042000205936

58 B. SJOGREN ET AL.

absorbed by ones own perfection to emotional involvement in another being has been
described by Bibring and colleagues (1961). According to Bowlby, attachment
between a nurturing caregiver, usually the mother, and the infant is a basic
prerequisite for survival of the newborn (Bowlby, 1988). An empirical study of
maternal foetal and parent-infant attachment, using interviews and test instruments,
found that emotional bonds with foetus develops early in pregnancy (Leifer, 1977). In
this study, three different patterns of attachment processes were identified: (1) low
attachment persisting during pregnancy; (2) little initial nearness to the foetus but
increasing with time; and (3) much nearness from the beginning of the pregnancy.
Interesting approaches to the study of the maternal foetal attachment were
undertaken after the introduction of inventories for investigating different aspects of
the process. Cranley was one of the first scientists who presented a questionnaire for
measuring maternal foetal attachment (MFAS) (Cranley, 1981). The MFAS consisted
of five subscales, assumed to measure five different aspects of the maternal foetal
attachment: role-taking, differentiation of self from the foetus, giving of self,
attributing characteristics and intentions to the foetus, and interaction with the
foetus. The inventory has been used in many studies, which signals the need for
objective and quantitative measurements of maternal engagement in the growing
foetus and future infant.
The inventory has been criticized because the results of different studies were
found to be inconsistent or even conflicting (Muller, 1992). Muller has presented
another questionnaire (Muller, 1993). Some of the items of this scale had, however,
similarity with those of the MFAS and the scores showed correlation to the scores of
MFAS. Condon has criticized the MFAS for missing emotional aspects and measuring
qualities other than attachment (Condon, 1993).
Another problem of the MFAS is the inappropriateness of some items in early
pregnancy, e.g. I can almost guess what my babys personality will be from the way
s/he moves around, which reduces its usefulness in studies of the development of
attachment before and during pregnancy. There is great need for a modified version of
the MFAS, which is possible to answer before, in early as well as late pregnancy.
In spite of the criticism, several studies have been conducted using the MFAS. A
comparison of the results of the MFAS inventory with another scale for measurement
of maternal foetal attachment (Kumar et al., 1984) has shown good concordance
(Berryman & Windridge, 1996).
Personality and attachment
To our knowledge, the role of personality traits for the development of attachment has
not been studied. Our group has previously studied changes in personality during
pregnancy and lactation by using the Karolinska Scales of Personality (KSP). This
inventory has been used in more than 100 studies (Gustavsson, 1997). A previous
study showed that the mothers were significantly more tolerant to monotony after the
delivery than non-pregnant controls, reported less anxiety proneness (less muscular
tension), and had more social desirable behaviours than non-pregnant controls
(Uvnas-Moberg et al., 1990). In another study, the scores of primiparae and
multiparae showed higher preparedness for social interaction after delivery and,
moreover, the multiparae were calmer and more relaxed than controls (Nissen et al.,
1998). Recent findings have shown changes in personality during the first pregnancy
and lactation (Sjogren et al., 2000). Three subscales of the KSP inventory, muscular

MATERNAL FOETAL ATTACHMENT

59

tension, somatic anxiety, and monotony avoidance, diminished significantly from


pregnancy to the test period after delivery. Based on these studies, a hypothesis was
presented that the attachment process might be related to personality.
Aim
The aim of the present study was to study the maternal foetal attachment process by
means of a modified version of Cranleys MFAS inventory (Cranley, 1981), adapted
for application in early as well as late pregnancy. The factor structure of the scale
should also be studied. Another aim was to investigate the relationship between the
maternal foetal attachment and the personality of the women.
Methods
Subjects
A consecutive series of 100 Swedish-speaking women, expecting their first child, was
recruited from all antenatal clinics in the catchment area of the Karolinska Hospital in
1990. These clinics cover all types of socio-economic populations. After obtaining
informed consent, the women were randomized into two subgroups for testing in the
13th week of gestation (Group 13 weeks) or to participate by filling in the attachment
questionnaire and the personality inventories during the 35th36th weeks of gestation
(Group 36 weeks).
Drop-outs. In Group 13 weeks, three of the 50 women had a spontaneous abortion
between the randomization moment and the time when they should fill in the
questionnaire, and were excluded by this reason. Thus, the response rate for the
test during early pregnancy was 41 out of 47 (87%). In Group 36 weeks, three
of the 50 women had a spontaneous abortion and one had experienced an
intrauterine foetal death in late pregnancy and had to be excluded for the same
reason as the three women in Group 13 weeks and 46 women remained. Three
women abstained immediately when informed, and the other eight women did not
return the questionnaire in spite of reminders. Thus, 35 of 46 participated, giving a
response rate of 76%. The drop-outs did not differ from the participants with
regard to age (mean 26.7, range 1943).
The women were examined in the 13th and 36th week of pregnancy, and named
Group 13 weeks and Group 36 weeks, respectively. The women were given the
questionnaire consisting of the KSP and the MFAS.
Personality inventory
The personality variables were measured by means of the Karolinska Scales of
Personality (KSP), a self-report inventory consisting of 15 subscales based on the
personality theory that assumes a relationship between personality and biological
characteristics (Gustavsson, 1997). The scales can be classified into three main
groups:

1. Anxiety proneness scales consist of five different subscales: the somatic anxiety
scale (autonomic disturbances, diffuse stress and panic attacks), the muscular
tension scale (muscular tenseness and aches, difficulties in relaxing), the psychic

60 B. SJOGREN ET AL.

anxiety scale (cognitive-social anxiety and insecurity, worrying and having slow
recovery after stress), the psychasthenia scale (low degree of mental energy and
difficulty in compensating for energy consumption) and, finally, the inhibition of
aggression scale (unexpressed anger and inability to speak up or assert oneself in
social situations).
2. There are five subscales in the group of extraversion scales: the impulsiveness,
monotony avoidance and detachment (need for social distance and coldness in
social relations) scales, the socialization scale (relations between the respondent
and his/her parents and other significant persons) and the social desirability scale
(measuring aspects of social adjustment).
3. The third group is the aggression-hostility scales, which have been adapted from
the Buss-Durkee inventory (Buss, 1961). Five scales are included in this group:
the indirect aggression scale (e.g. slamming doors), irritability (being easily
annoyed, quick to anger), verbal aggression (verbal expression of anger, e.g.
shouting, quarrelling, cursing), suspicion (e.g. being distrustful, projecting hostility to others) and guilt (feelings of remorse and shame).
The KSP inventory consists of 135 items with a four-point response format (Does not
apply at all to Applies completely). All scales have 10 items except the socialization
scale and the aggression-hostility scales, which have 20 and five items, respectively.
The subscales are transformed into t scores (mean 50 and SD 10) and are
standardized with regard to age and sex. The scales measure personality traits, which
have shown a high test-retest stability over a period of 10 years (Gustavsson et al.,
1997; Kampe et al., 1996). A normative group derived out of 200 women, randomly
sampled from a normal population in Sweden, was used for comparison with the
women in the study group. Thus the pregnant women were compared with women in
the same age span.
Maternal foetal attachment inventory
Maternal foetal attachment was measured by 17 items of the MFAS inventory
according to Cranley (1981). As the original MFAS seemed to be most suitable for
application in the last trimester of pregnancy, an adapted version of it was elaborated,
in which the items could be answered in early as well as late pregnancy. Seven items
were excluded. These items had the following numbers and wording in the original
scale: I enjoy watching my tummy jiggle as the baby kicks inside (no. 3); I can almost
guess what my babys personality will be from the way s/he moves around (no. 9); It
seems my baby kicks and moves to tell me its eating time (no. 16); I poke the baby
to get him/her to poke back (no. 17); I stroke my tummy to quiet the baby when
there is too much kicking (no. 20); I can tell that the baby has hiccoughs (no. 21); I
grasp my babys foot through my tummy to move it around (no. 24).
A four-choice response format, instead of the five-choice format, was used. The
reason for a four-choice format was to obtain congruency with the KSP and to avoid
the possibility of answering in a neutral way.
Statistical methods
All variables were summarized using descriptive statistics such as means, standard
deviations and frequencies. Differences between groups in attachment and personality

MATERNAL FOETAL ATTACHMENT

61

were analysed using Students t-test. Relationships were expressed as Pearsons


product-moment correlation coefficients. In case of skewed distributions, the nonparametric Mann-Whitney test was used.
The factor structure of the MFA scale was analysed by a principal component
analysis, Varimax rotation with Kaiser normalization (Winer, 1962). The number of
eigenvalues above 1.00 determined how the number factors that was extracted. Based
on the factor analysis, items were summed to a factor scale, i.e. the factor to which the
item had their highest correlation.
The relationship between attachment and personality was analysed by a stepwise
regression analysis in order to study the unique contribution of each personality
variable. The regression analyses were made on collapsed groups, i.e. Group 13 weeks
plus Group 36 weeks, and the grouping variable was entered as an independent
variable in the analyses. The inclusion criteria were set to 5% and the exclusion criteria
to 10%.
A significance level of 5% was applied, and all statistical analyses were two-tailed.
Results
Sociodemographic characteristics
Sociodemographic characteristics of the two groups are presented in Table 1. There
were no significant differences between the groups in age, education, number of
women living with the father of the foetus, employment or number of smokers. Thus,
the pregnant women in the total sample were in their mid 20s (mean age 26.8 years,
range 1939 years), mostly well educated (w50%), most often living with the father of
the foetus (w90%), working full-time (w90%) and primarily non-smokers (v25%
smoked).
Factor analysis of the attachment scale
The factor analysis generated four factors (Table 2, where the names of Cranleys a
priori subscales of the MFAS are also presented) explaining 65% of the total variance.
Factor I (visualizing motherhood1) included six items with factor loading above 0.40,

Table 1. Sociodemographic characteristics of two groups of women pregnant during 13 or 36 weeks.


Variable

Age
Education

Living with father of the foetus


Working condition
Smoker

Mean
Range
None
Low
Medium
High
Yes
Half-time
Full-time
Yes

Group 13 weeks
(n~41)

Group 36 weeks
(n~35)

26.2
1934
3 (8%)
16 (41%)
14 (36%)
6 (15%)
34 (90%)
4 (11%)
34 (90%)
7 (18%)

27.4
2139
3 (9%)
16 (49%)
12 (36%)
6 (6%)
32 (97%)
2 (6%)
31 (94%)
8 (24%)

62 B. SJOGREN ET AL.
Table 2. Rotated component matrix for the attachment scale (n~76); extraction method was principal component
analysis, and rotation method was Varimax with Kaiser normalization; only factor loadings over 0.40 are shown.
A priori factors

Factor
I

I have decided on a name for a girl baby.


I imagine myself taking care of the baby.
I have decided on a name for a baby boy.
I try to picture what the baby will look like.
I picture myself feeding the baby.
Im really looking forward to seeing what the
baby looks like.

DiffSelf
RoleTake
DiffSelf
RoleTake
RoleTake
DiffSelf

0.74
0.72
0.69
0.64
0.60
0.56

I wonder if the baby can hear inside of me.


I wonder if the baby thinks and feels inside
of me.
I wonder if the baby feels cramped in there.

Attrib
Attrib

0.85
0.82

0.74
0.73

Attrib

0.74

0.61

I eat meat and vegetables to be sure my baby


gets a good diet.
I do things to try to stay healthy that I would
not do if I were not pregnant.
I feel all the trouble of being pregnant is worth it.
I give up doing certain things because I want
to help my baby.
I can hardly wait to hold the baby.
I feel my body is ugly.
I refer to my baby by a nickname.
I talk to my unborn child.

II

III

IV

h2

according to Cranley

0.55
0.64
0.53
0.51
0.46
0.40

GivSelf

0.72

0.58

GivSelf

0.68

0.56

GivSelf
GivSelf

0.66
0.61

0.47
0.41

RoleTake

0.54

0.57

GivSelf
Interact
Interact

(0.50)

20.80 0.68
0.73 0.58
0.57 0.67

RoleTake~role-taking; DiffSelf~differentiation of self from the foetus; GivSelf ~giving of self;


Attrib~attributing characteristics and intentions to the foetus; and Interact~interaction with the foetus.

i.e. a correlation with thoughts about a baby, all concerning nurturing and contact
with the child-to-be. Factor II (wonder and worry) included three items concerning
the recognition of the foetus with special capacities. Factor III (nourishing self and
foetus) included five items concerning her own responsibility for appropriate nutrition
and generosity with her own resources in relation to the foetus. Factor IV (relation to
foetus and own body) included two items reflecting communication and the
relationship with the foetus and one item concerning the womans own body and
beauty.
Group differences in and characteristics of attachment
There was a significant difference between the groups in the total score (pv0.01),
women in a late pregnancy having the higher scores.
When the items were added to their respective factor scales, significant differences
between the groups were found in Factors I (visualizing motherhood) (t(74)~2.21,
pv0.05) and II (t(74)~3.14, pv0.01) with higher scores in the late pregnancy group.
Finally, there was almost a significant difference in the same direction in Factor IV
(relation to foetus and own body) (t(74)~1.82, p~0.073).
The women in Group 13 and 36 were also compared in all items (Table 3).

MATERNAL FOETAL ATTACHMENT

63

Women in Group 36 talked more often to the child (pv0.001), wondered if the child
was cramped (pv0.001), knew the name of the child if it was a girl (pv0.001),
wondered more often what the foetus thinks and feels (pv0.01). There were also two
non-significant tendencies (pv0.10) indicating that women in Group 36 thought more
often about how the child looked like and wondered if the child could hear.
Generally speaking, there was a high frequency of agreement in items assumed to
measure different aspects of attachment. A mean score of 2.0 indicates that the
respondents agree with the item statement. As can be seen from Table 3, the mean
score of both groups was close to 3.0, which translates to Applies pretty much. In
both groups, the highest mean agreement was found for the following items
(percentage of mothers agreeing in parenthesis): Pregnancy worth the trouble (95%);
Wonder how child looks (96%); Have a pet name (63%); Eat food good for child
(84%); Imagine how child looks (78%); and Abstain from something to help the
child (87%). Furthermore, most mothers disagreed with the item Think my body is
ugly (80% disagreed).
It is noteworthy that there was a significant difference between the mothers in
Group 13 and 36 in knowing the name, if the child would be a girl, but not for
knowing the childs name if a boy, indicating that during pregnancy the mothers in spe
become increasingly convinced of the girls name but not the boys.

Table 3. Mean (M) and standard deviations (SD) in the attachment scales completed by two groups of pregnant
women, 13 and 36 weeks.
Items of the attachment scale

I talk to my unborn child.


I feel all the trouble of being pregnant is
worth it.
I picture myself feeding the baby.
Im really looking forward to seeing what
the baby looks like.
I wonder if the baby feels cramped in there.
I refer to my baby by a nickname.
I imagine myself taking care of the baby.
I have decided on a name for a girl baby.
I do things to try to stay healthy that I would
not do if I were not pregnant.
I wonder if the baby can hear inside of me.
I have decided on a name for a baby boy.
I wonder if the baby thinks and feels inside
of me.
I eat meat and vegetables to be sure my baby
gets a good diet.
I can hardly wait to hold the baby.
I try to picture what the baby will look like.
I feel my body is ugly.
I give up doing certain things because I want
to help my baby.
Mean attachment score

Group 13 weeks
(n~41)

Group 36 weeks
(n~35)

SD

SD

2.0
3.6

1.06
0.63

3.1
3.7

0.97
0.52

v0.001
NS

2.2
3.7

1.14
0.61

2.5
3.9

0.92
0.28

NS
v0.10

1.9
3.0
2.8
2.4
2.8

1.08
1.19
0.97
1.22
1.26

2.8
3.0
3.0
3.3
3.0

1.03
1.29
0.77
1.09
0.99

v0.001
NS
NS
v0.001
NS

2.5
2.7
2.5

1.10
1.22
0.98

3.0
2.9
3.1

1.12
1.14
0.97

v0.10
NS
v0.01

3.2

0.78

3.3

0.72

NS

2.7
3.1
1.8
3.3

1.09
1.05
0.85
0.87

2.9
3.4
1.8
3.3

1.01
0.84
0.93
0.64

NS
NS
NS
NS

2.8

0.51

3.1

0.43

v0.01

64 B. SJOGREN ET AL.

Differences in personality
The mean personality scale scores (t scores) are presented in Table 4. All the means
were close to the normal values (M~50), with no significant differences between the
groups.
Relationships between attachment and personality
In the linear regression analysis of the correlation between attachment (total scale) and
personality, only social desirability and guilt made significant contributions to the
regression equation except for the grouping variable (Table 5). The multiple correlation
was 0.50, explaining 22% of the total variance and indicating that high scores on
socially desirable attitude and a proneness to feelings of guilt were associated with
higher attachment, and being in the late phase of pregnancy.
There were several interesting relationships between the separate attachment
factors and personality. Except for the grouping variable, Factor I (visualizing
motherhood) was related to somatic anxiety, psychic anxiety, and guilt. The multiple
correlation was 0.42 (17% common variance). Thus, women in the later phase of
pregnancy, high in somatic anxiety and guilt, and low in psychic anxiety, had higher
levels of proneness of visualizing being a mother. A multiple correlation of 0.54 (29%
common variance) was observed between Factor II (wonder and worry) and the
grouping variable (later pregnancy was higher), psychasthenia (higher), social
desirability (higher), and irritation (higher). Factor III (nourishing self and foetus)
was positively correlated with social desirability, somatic anxiety, and inhibition of
aggression. The multiple correlation was 0.49 (24% common variance). Finally,
Factor IV (relation to foetus and own body) was negatively correlated with
detachment and psychic anxiety, and positively related to phase of pregnancy and
social desirability. The multiple correlation was 0.55 (30% common variance).

Table 4. Means (M) and standard deviations (SD) in Karolinska Scales of Personality (t scores) in two groups of
pregnant women, 13 and 36 weeks of pregnancy; none of the differences in personality were significant.
Personality scale

Somatic anxiety
Muscular tension
Psychic anxiety
Psychasthenia
Inhibition of aggression
Impulsivity
Monotony avoidance
Socialization
Detachment
Social desirability
Indirect aggression
Verbal aggression
Irritation
Suspicion
Guilt

Group 13 weeks (n~41)

Group 36 weeks (n~41)

SD

SD

47.67
46.91
45.35
48.36
46.92
51.44
53.75
52.54
46.93
48.76
49.61
48.97
50.82
49.28
46.42

8.33
10.97
9.11
13.37
11.03
7.92
7.00
8.58
8.00
8.12
11.17
9.02
7.92
12.11
11.73

48.12
47.53
45.26
45.72
49.31
50.24
51.18
53.95
48.48
49.75
49.05
49.15
49.43
48.18
45.53

8.43
8.71
9.56
9.16
10.10
9.48
6.92
8.02
9.51
8.63
8.14
6.67
8.74
9.35
10.28

MATERNAL FOETAL ATTACHMENT

65

Table 5. Regression analysis of relationships between attachment and personality.


Attachment variable

Variable

Beta
(unstandardized)

Attachment
Total scale

(Constant)
Group
Social desirability
Guilt, pregnancy

Factor I
Visualizing motherhood

Factor II
Wonder and worry

(standardized)

1.18
0.33
0.02
0.01

0.394
0.103
0.006
0.005

0.33
0.27
0.23

2.99
3.21
2.61
2.29

0.004
0.002
0.011
0.025

Constant
Group
Somatic anxiety
Psychic anxiety
Guilt, pregnancy

1.88
0.34
0.02
20.02
0.01

0.522
0.145
0.010
0.009
0.008

0.25
0.25
20.30
0.22

2.30
2.02
22.29
1.74

0.024
0.047
0.025
0.086

Constant
Group
Psychasthenia
Social desirability
Irritation

21.78
0.68
0.02
0.03
0.02

0.940
0.185
0.009
0.011
0.013

0.37
0.22
0.26
0.21

3.67
1.97
2.55
1.85

0.000
0.053
0.013
0.069

0.64
0.02
0.02
0.01

0.541
0.007
0.008
0.006

0.34
0.23
0.20

3.33 0.001
2.10 0.039
1.77 0.082

3.88
20.03
0.37
20.02
0.02

0.767
0.010
0.163
0.009
0.010

20.33
0.23
20.22
0.17

Factor III
Constant
Nourishing self and foetus Social desirability
Somatic anxiety
Inhibition of aggression
Factor IV
Relation to foetus and
own body

Beta
SE

Constant
Detachment
Group
Psychic anxiety
Social desirability

23.04
2.25
22.02
1.68

0.003
0.027
0.047
0.097

Discussion
The results of the attachment measurements showed that the women in the study
groups were highly motivated for childbearing, as the scores were generally high. The
factor analysis of the modified MFAS gave subscales which had some, but not
complete, similarities with the subscales of the original MFAS. Moreover, the
attachment measurements showed that the relation to the future child was more
intense in late pregnancy.
A previous study found that, already in the 8th11th week of pregnancy, the
women had developed a considerable emotional involvement in the foetus (Sjogren &
Uddenberg, 1988).
The modified MAFS and factor scales
The factor analysis of the attachment scale revealed four factors, indicated different
and independent qualities of the mothers relation to the foetus and future infant. The
first factor was called visualizing motherhood. The content of these items concerns
fundamental aspects of motherhood: first, to feed, to hold and to care for the infant,
and, second, to look at the infant and to lay the foundation of a relationship with a
specific baby and body, a girl or a boy. These items are found in the subscales roletaking and differentiating of self from foetus of Cranleys inventory.

66 B. SJOGREN ET AL.

The second factor is associated with a somewhat uncertain perception of the foetus
as a separate living being that probably has special abilities: to hear, to think and feel
and to feel uncomfortable. This factor was called wonder and worry. Agreement with
the items of this factor reflects that the foetus is perceived as a growing living being
and a possible subject. It is very close to the Cranley subscale called attributing
characteristics and all items of wonder and worry are found in this scale.
The third factor consisted of items reflecting the womans responsibility for proper
nutrition of herself as well as of the foetus. The subscale reflected a behaviour
promoting physical health and generosity with the womans own resources. We call
this subscale nourishing self and foetus. Four of the five items on this subscale are found
on the subscale giving of self in Cranleys inventory.
The fourth and last factor reflected the expectant mothers process of communication, establishment of a relationship with the foetus, and her concerns
regarding her own body and beauty. This factor was named relation to foetus and own
body. The items are found in the Cranley scale giving of self and interacting with
foetus. High scores on the items reflecting communication with the future infant and
low score on the item concerning body image gave a high subscale value. The
psychological process behind this finding may be expressed by the words of a
psychoanalyst: to invest in the future infant and to give up ones own gratification of
needs in favour of the needs of this infant (Bibring et al., 1961).
Similarly, in a recent Swedish study, where the prenatal attachment was studied by
means of the Muller inventory, the multidimensional character phenomenon of the
attachment phenomenon was shown (Siddiqui & Hagglof, 1999). These dimensions
were named fantasy, interaction, affection, differentiation of self from foetus and
sharing with others.
Attachment during different phases of pregnancy
The item scores that were higher for women in late than in early pregnancy belonged
to Factor I, visualizing motherhood (two items of six), Factor II, wonder and worry
(three items of three), Factor IV, relation to foetus and own body (one item of three)
but none of the items of Factor III, nourishing self and foetus, differed significantly
between the groups. One interpretation is that the processes measured by this
scale mainly take place before the 13th week of gestation, such as ambivalence.
Ambivalence towards the pregnancy has been described as a conflict between the wish
of existential fulfilment and the fear of restriction of personal freedom (Wikman et al.,
1993). The decision to undergo a legal abortion or to continue the pregnancy has been
found to be taken in early pregnancy (Holmgren, 1988).
It is an interesting finding that thoughts about a name for a boy did not differ
between the groups, but the thinking about a female name did, a decision formed in
the end of the pregnancy. One interpretation might be that there are different
meanings attached to the possibility of having a girl or a boy baby, e.g. differences in
expectations and ambivalence between these two possibilities.
The concept of and the measurement of emotional attachment to a future infant
What do we mean by the concept of attachment of a woman to her foetus or to
anybody? Although efforts have been made to describe maternal foetal attachment, a
final definition has not been reached and differs between authors. The content of the

MATERNAL FOETAL ATTACHMENT

67

MFAS inventory has been criticized because of the small number of questions
regarding the mothers emotions (positive as well as negative) to the foetus and future
baby (Condon, 1993). Moreover, the MFAS does not include any questions about
detachment, which can only be expressed by a no agreement answer. This may be a
weakness of the scale. According to Condon, the attachment process, the subjective
experience of attachment or love, consists of five dispositions: the dispositions to
know, to be with, to avoid separation or loss, to protect and to gratify needs. Many of
the items of Cranleys MFAS, however, reflect the pregnant womans thoughts about
the need of the foetus and future infant for physical care and food. One important
question is whether imaginative actions to satisfy and help the baby to physical survival
are indications of attachment. Cranley described attachment as the expressions of the
emotions in imaginative or real actions. According to Websters Dictionary, attachment
is described as any passion that binds one person to another, but also as that bond
which attaches one object to another. The different characteristics of the two scales
may reflect the different cultures in which where they were created, the world of
midwifery and obstetrics where basic physical security is fundamental, on the one
hand, and the more abstract intellectual world of psychology, on the other. Thus, we
found that the questions about thoughts of food and basic care giving are relevant to
the description of maternal foetal attachment.
Attachment and personality
The results of the personality test did not differ between the groups, which may
confirm the nature of the personality variables as stable characteristics or traits that are
not significantly affected by pregnancy, although some changes were found to occur
during the period of delivery and lactation (Sjogren et al., 2000). The calculations of
the correlation between attachment to the foetus/future child, as expressed by the total
score of the attachment scale, on the one hand, and the personality characteristics, on
the other, mostly showed no correlation. However, positive connections were found
between the total attachment score and the personality scales social desirability and
guilt. Moreover, the phase of the pregnancy showed a more intense attachment in the
women tested in late pregnancy, as has been described above.
The relationship between the new subscales or factors of the attachment and
personality was analysed. The correlation between personality traits and the MFAS
subscales indicated a great tendency to do things to satisfy other human beings, since
three of the subscales indicated a connection to the personality trait social desirability,
the attachment subscales visualizing motherhood, wonder and worry and relation to
foetus and own body. This is in accordance with our hypothesis and previous findings
showing that the scores on the social desirability scale were higher in new mothers
than in controls (Uvnas-Moberg et al., 1990) and an increase during a first pregnancy
and lactation was also demonstrated (Sjogren et al., 2000).
The next important connection was between the attachment subscales and a
personality profile characterized by anxiety proneness in the KSP subscales somatic
and psychic anxiety or psychasthenia. These type of connections were found in three
of the four attachment subscales, visualizing motherhood, wonder and worry and
nourishing self and foetus. In the subscale relation to foetus and own body, however, a
negative correlation to the KSP subscale psychic anxiety was found. Moreover,
another important finding was the positive correlation between a high attachment to
the future infant, expressed in the visualizing motherhood and the KSP subscale guilt.

68 B. SJOGREN ET AL.

A tendency to guilt feelings may be related to strong feelings of responsibility and may,
in extreme cases, be negative and painful.
The finding that the KSP subscale irritation showed correlation to the MFAS
scale II, wonder and worry, indicated that some vulnerability to different kinds of
disturbance may be concomitant to nearness to and imaginative care of the growing
foetus. The finding of a negative correlation between the attachment factor relation to
foetus and own body and the KSP subscale detachment indicated that a low tendency
to need distance to other persons was present in cases of high attachment. However, it
must be kept in mind that most of the mothers in spe had high scores on attachment,
thus limiting the possibility of finding a significant correlation between attachment and
personality.
Important aspects of the attachment phenomenon were that it had a relation to
guilt and anxiety, the wish to satisfy others and thus may be accompanied by feelings
of guilt, tension and distress. An extremely strong feeling of obligation to others may,
however, turn into guilt and may, in some individuals, predispose to mental
difficulties, e.g. depressive and anxiety disorders (Gelder et al., 1996). Psychological
problems during childbirth are frequent, e.g. depression, which has been seen in 8
10% of women after childbirth (Lundh & Gyllang, 1993; Murray & Cox, 1990). One
tentative interpretation of these findings could be that extremely high attachment
scores do not necessarily indicate a healthy development of the mother-future child
relationship. Some of the 80 women in this group may run a risk of depressive or
anxiety disorders. Actually, three of the women had high scores on anxiety (more than
2 SD). This aspect of the present findings ought to be studied in future research.
Conclusions
The factor analysis of the modified version of the inventory for measuring maternal
foetal attachment according to Cranley showed a considerable, but not complete,
similarity with the original subscales. The women who were tested at the end of
pregnancy showed higher attachment to the foetus than those who were studied early
in the pregnancy. Attachment and the four subscales showed positive correlation with
personality, e.g. the tendency to feel guilt and anxiety and the experience of social
desirability, as well as a negative correlation with emotional detachment. To our
knowledge, these findings are a fairly new contribution to the literature on maternal
foetal attachment.
Acknowledgements
The authors are grateful to the Karolinska Institute and the Swedish Society of
Medicine for financial support from the funds of these institutions and to Birgitta
Froman, nurse midwife and organizer of the Antenatal Care of the Northwest Region
of the Stockholm County Council.
Note
1. Tentative factor name.
References
BERRYMAN, J.C. & WINDRIDGE, K.C. (1996). Pregnancy after 35 and attachment to the foetus. Journal of
Reproductive and Infant Psychology, 11, 133143.

MATERNAL FOETAL ATTACHMENT

69

BIBRING, G., DWYER, T., HUNTINGTON, D. & VALENSTEIN, A.F. (1961). A study of the psychological process
in pregnancy and the earliest mother-child relationship. In: The psychoanalytic study of the child. London:
Hogarth Press.
BOWLBY, J. (1988). A secure base. New York: Basic Books.
BREEN, D. (1975). Birth of a first child. London: Tavistock.
BUSS, A.H. (1961). The psychology of aggression. New York: Wiley.
CONDON, J.T. (1993). The assessment of emotional attachment: development of a questionnaire instrument.
British Journal of Medical Psychology, 66, 167183.
CRANLEY, M.S. (1981). Development of a tool for the measurement of maternal attachment during
pregnancy. Nursing Research, 30, 281284.
GELDER, M.G., GATH, D., MAYOU, R. & COWEN, P. (1996). Oxford textbook of psychiatry. Oxford: Oxford
University Press.
GUSTAVSSON, J.P. (1997). Stability and validity of self-report personality traits. Unpublished doctoral
dissertation, Karolinska Institute, Stockholm.
SBERG, M. (1997). Stability and
GUSTAVSSON, J.P., WEINRYB, R.M., GORANSSON, S., PEDERSEN, N.L. & A
predictive ability of personality traits across 9 years. Personality and Individual Differences, 22, 783791.
HOLMGREN, K. (1988). Time of decision to undergo a legal abortion. Gynecological and Obstetrical
Investigation, 26, 289295.
KAMPE, T., EDMAN, G. & HANNERZ, H. (1996). Ten-year follow-up study of personality traits in adults with
intact and restored dentitions. Journal of Oral Rehabilitation, 23, 443449.
KLAUS, M.H. & KENNEL, J.H. (1976). Maternal-infant bonding. Saint Louis: CV Mosby.
KUMAR, R., ROBSON, K.M. & SMITH, A.M.R. (1984). Development of a self-administered questionnaire to
measure maternal adjustment and maternal attitudes during pregnancy and after delivery. Journal of
Psychosomatic Research, 28, 4351.
LEIFER, M. (1977). Psychological changes accompanying pregnancy and motherhood. Genetic Psychology
Monographs, 95, 5596.
LUNDH, W. & GYLLANG, C. (1993). Use of the Edinburgh Depression Scale in some Swedish child health
care centres. Scandinavian Journal of Caring Science, 7, 149154.
MULLER, M.E. (1992). A critical review of prenatal attachment research. Scholarly Inquiry for Nursing
Practice: An International Journal, 6, 522.
MULLER, M.E. (1993). Development of the prenatal attachment inventory. Western Journal of Nursing
Research, 15, 199215.
MURRAY, D. & COX, J. (1990). Screening for depression during pregnancy with the Edinburgh Depression
Scale (EPDS). Journal of Reproductive and Infant Psychology, 8, 99107.
NISSEN, E., UVNAS-MOBERG, K. & WIDSTROM, A.M. (1998). Oxytocin and prolactin, milk production and
their relation with personality traits in women after vaginal delivery and Sectio Caesarea. Journal of
Psychosomatic Obstetrics and Gynecology, 19, 4958.
SIDDIQUI, A. & HAGGLOF, B. (1999). An exploration of prenatal attachment in Swedish expectant women.
Journal of Reproductive and Infant Psychology, 17, 369380.
SJOGREN, B. & UDDENBERG, N. (1988). Prenatal diagnosis and maternal attachment to the child-to-be.
Journal of Psychosomatic Obstetrics and Gynecology, 9, 7387.
SJOGREN, B., EDMAN, G. & UVNAS-MOBERG, K. (2000). Changes in personality pattern during first pregnancy
and lactation. Journal of Psychosomatic Obstetrics and Gynecology, 21, 3138.
UDDENBERG, N. (1974). Reproductive adaptation in mother and daughter: a study of personality
development and adaptation to motherhood. Acta Psychiatrica Scandinavica, 254, 2631.
UVNAS-MOBERG, K., WIDSTROM, A.M., NISSEN, E. & BJORVELL, H. (1990). Personality traits in women 4 days
after delivery compared with controls. Journal of Psychosomatic Obstetrics and Gynecology, 11, 261273.
WIKMAN, M., JACOBSSON, L. & JOELSSON, I. (1993). Ambivalence towards parenthood among pregnant
women and their men. Acta Obstetricia et Gynecologica Scandinavica, 72, 619626.
WINER, B.J. (1962). Statistical principles in experimental design. New York: McGraw-Hill.

You might also like