Citation
The effects of maternal depression on infant language development

Material Information

Title:
The effects of maternal depression on infant language development
Creator:
Danko, Christina
Publication Date:
Language:
English
Physical Description:
ix, 59 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Postpartum depression ( lcsh )
Infants -- Language ( lcsh )
Mother and infant ( lcsh )
Infants -- Language ( fast )
Mother and infant ( fast )
Postpartum depression ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 51-59).
General Note:
Department of Psychology
Statement of Responsibility:
by Christina Danko.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
259747323 ( OCLC )
ocn259747323
Classification:
LD1193.L645 2008m D36 ( lcc )

Full Text
THE EFFECTS OF MATERNAL DEPRESSION ON
INFANT LANGUAGE DEVELOPMENT
by
Christina Danko
B.A., University of Virginia, 2005
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Clinical Psychology
2008


This thesis for the Master of Arts
degree by
Christina Danko
has been approved
Peter Kaplan
Elizabeth Allen
Hltsio^
Date


Danko, Christina M. (M.A., Clinical Psychology)
The Effects of Maternal Depression on Infant Language Development
Thesis directed by Senior Instructor Kevin Everhart
ABSTRACT
Postpartum depression occurs in 10-15% of mothers. Maternal depression has been
shown to affect child language development, cognitive development, and associative
learning. These effects are thought to occur through less responsive and stimulating
maternal interactions with the infant. These non-optimal interactions may have a
significant effect on the infants brain development, which could then later result in
language and cognitive delays. Previous research has looked at maternal depression
and child language development, but few studies have looked at the effects on
children as young as 12 months. In the current study, depressed and non-depressed
mothers, fathers, and their 11-13 month old infants participated in a study that looked
at maternal depression, infant language development, infant associative learning,
quality of mother-infant interactions, paternal involvement, and infant maternal and
paternal attachment. Infant language development was assessed using the Bayley
Scales of Infant and Toddler Development, Third Edition. It was found that an
episode of maternal depression in the first year of the infants life was associated with


slower infant expressive language development. Infant-paternal attachment was found
to be positively related to infant language development. There was a significant
interaction effect for elevated maternal Beck Depression Inventory-II (BDI-II) score
and infant-maternal attachment, where infants of mothers with elevated BDI-II scores
appeared to be protected from slower language development if the infants were
securely attached to the mother. However, there was no direct effect of elevated
maternal BDI-II scores on infant language development. Surprisingly, maternal
sensitivity was found to be negatively related to infant language development. In
addition, infant language development was not significantly related to amount of
paternal involvement. These findings and their implications are discussed.
This abstract accurately represents the content of the candidate^,
its publication.
recommend
Signed
Kevin Everhart


ACKNOWLEDGEMENT
I would like to thank Peter Kaplan, Kevin Everhart, and Elizabeth Allen for their
support over the past two years and help with this thesis. I would also like to thank
Amanda Moreno, Andres Diaz, and all the other members of the lab for their help and
support.


TABLE OF CONTENTS
Figures........................................................viii
Tables.........................................................ix
CHAPTER
1. INTRODUCTION...................................................1
Infant Development.........................................1
Infant Language Development..............................3
Maternal Depression and Infant-Directed Speech...........5
Maternal Depression and Infant Cognitive and Language
Development..............................................7
Maternal Depression and Mother-Infant Interactions.........8
Maternal Sensitivity.....................................9
Infant-Maternal Attachment..............................10
Types of Depressed Maternal Interactions................11
The Role of the Father....................................13
The Current Study.........................................14
2. METHOD........................................................16
Participants..............................................16
Measures..................................................17
Structured Clinical Interview for DSM-IV-TR.............17
vi


Beck Depression Inventory-II............................18
Bayley Scales of Infant and Toddler Development Third
Edition.................................................19
MacArthur Vocabulary Checklist..........................19
Associative Learning Paradigm...........................20
Emotional Availability Scales...........................22
Attachment Q-Set........................................23
The Child Care Activity Questionnaire...................24
Procedure.................................................24
3. RESULTS......................................................26
Demographic Data..........................................26
Power Analysis..........................................28
Language Development and Maternal Depression............28
Maternal Sensitivity and Structuring....................30
Infant Associative Learning Score.......................32
Attachment..............................................33
Father Involvement......................................39
MacArthur Vocabulary Checklist..........................39
4. DISCUSSION...................................................41
REFERENCES...........................................................51
vii


LIST OF FIGURES
Figure
3.1 Infant language score means as a function of father attachment category ...34
3.2 Infant language score means as a function of mother attachment category and
maternal BDI-II category..............................................35
3.3 Infant language score means as a function of mother attachment category and
maternal sensitivity rating category..................................38
vm


LIST OF TABLES
Table
3.1 Demographic data...............................................27
IX


CHAPTER 1
INTRODUCTION
Postpartum depression affects 10-15% of mothers, and is defined as a non-
psychotic major depressive disorder that begins within a year of childbirth (Grace,
Evindar, & Sterwart, 2003). Major depressive disorder is characterized by a low or
sad mood, lack of interest in activities, disruption in sleeping or eating, loss of energy,
inability to concentrate, thoughts of worthlessness or guilt, psychomotor agitation or
retardation, and/or thoughts of suicide and death (American Psychiatric Association,
2000). Infants of depressed mothers have an increased risk of later language and
cognitive problems, but the mechanism is not clear (Sohr-Preston & Scaramella,
2006). The aim of the current study is to explore the relationship between maternal
depressive symptoms and child language development as a step in identifying
mechanisms and ultimately designing interventions.
Infant Develonment
The infants brain is in a vital period of accelerated growth from the third
trimester of pregnancy through the second year (Dobbing & Sands, 1973). Experience
influences the maturation of the infants brain, and for the infant some of these
experiences occur within the context of the attachment relationship (Schore, 2001a).
Stress and trauma impair brain development, whereas healthy attachment
1


promotes it (Schore, 2001b). Secure and healthy attachment is associated with
caregivers who are warm and responsive, as well as accessible to the infant
(Ainsworth, 1978). Caregivers normally help regulate infants internal states, and
caregivers who do not help regulate infants internal states may expose them to
prolonged periods of stress and dysregulation (Tronick & Weinberg, 1997). It has
been found that mothers with postpartum depression expose their infants to longer
periods of dysregulation that nondepressed mothers (Tronick & Weinberg, 1997).
One of the effects of these periods of stress and dysregulation may be that the infants
limbic system is exposed to high levels of excitoxic neurotransmitters (e.g. glutamate)
and high levels of cortisol. This can result in abnormal limbic circuitry and interfere
with the organization of the limbic system, which impairs the capacity to cope with
stress and novelty and organize new learning (Mesulam, 1998).
Infants are motivated to create a collaborative state with an adult caregiver to
help them regulate their internal state and also to make meaning of the world (Tronick
& Weinberg, 1997). If the infant becomes dysregulated with no adult caregiver to
help with regulation, all of the infants resources will be devoted to achieving a state
of equilibrium and will have no resources leftover to explore the world and learn
(Tronick & Weinberg, 1997). Thus, not only can dysregulation affect the infants
brain formation and organization through changing chemical levels, but it also
prevents the infant from experiencing the environment and learning from that
experience.
2


Right frontal EEG asymmetry has been found to be one of the major
psychophysiological markers of depression in adults (Schaefer, Davidson, & Saron,
1983). This finding extends to infants and children of depressed mothers as well
(Field, Fox, Pickens, & Nawrocki, 1995; Jones, Field, Fox, Lundy, & Davalos, 1997).
Jones, Field, Fox, Lundy, & Davalos (1997) reports that infants of depressed mothers
show right frontal EEG asymmetry due to hypoactivation in the left frontal lobe. They
found that these infants who had the right frontal EEG asymmetry had more negative
facial expressions and lower activity levels (Jones, Field, Fox, Lundy, & Davalos,
1997). This is important considering the stability of EEG patterns found in infants and
children of depressed mothers (Jones, Field, Davalos, & Pickens, 1997). Right frontal
EEG asymmetry may be an additional risk factor for infants of depressed mothers to
follow a trajectory of more negative developmental outcomes (Jones, Field, Fox,
Lundy, & Davalos, 1997).
Tnfant Language Development
Language can be divided into two types, expressive and receptive language
(Bowen, 1998). Expressive language is language that is uttered. In 12-month-old
infants, it can be thought of as preverbal communication, including cooing and
babbling, and starting to use words (Bowen, 1998). Receptive language is what is
understood by the child. Twelve-month-old infants can respond to sounds, tone of
3


voice, can understand some words, and are starting to be able to respond to requests
(Bowen, 1998).
In adults, language processing areas are located in the left hemisphere. This is
not the case for infants and children, and lateralization of language is thought by
some to be a process that does not start until children are beginning to speak words
and finishes when children go into adolescence (Boysson-Bardies, 1999). For
instance, if a brain lesion in the left hemisphere occurs in childhood, the earlier it
occurs the less affected the childs language development will be (Boysson-Bardies,
1999). Although language is not fully lateralized in the left hemisphere in infancy,
there are indications that the infant is bom with the left hemisphere already
specialized for language. Researchers have shown that for 4 day old neonates as well
as infants that are a few months old, the left hemisphere does a better job of
differentiating between different speech sounds than the right hemisphere (Bertoncini,
Morais, & Bijeljac-Babic, 1989; Glanville, Best, & Levenson, 1977). Trevarthen
(1986) found that when infants had utterances of prespeech and/or cooing, it was
often accompanied with raised right hand gestures and extending the right hand.
Infants interaction with caregivers may directly affect language development.
Infants participate in daily routines with caregivers that have a beginning, middle, and
end (Tronick & Weinberg, 1997). The sequences of affective experiences and their
relation to external events act as a meaning system for the infant (Tronick &
Weinberg, 1997). For example, an infant may come to learn that when he cries, the
4


caregiver will pick him up and soothe him, and then he will feel better. These daily
routines serve as narratives that exist before a child has words and act as a
prerequisite for learning language (Tronick & Weinberg, 1997). This may provide a
foundational process that aids the infants language acquisition through the infant
developing the capacity to be sensitive to reciprocal interactions and making meaning
of the world with the aid of another (Bruner, 1983).
Maternal Depression and Infant-Directed Speech
Adults have a natural propensity to modify their speech and produce infant
directed (ID) speech when talking to infants (Jacobson, Boersma, Fields, & Olson,
1983). ID speech is characterized by higher pitch, more prosodic repetition, shorter
verbalizations, and often more expanded contours than adult-directed speech (Femald
& Simon, 1984). ID speech is thought to be the best way of gaining the infants
attention and facilitating communication with adults (Femald & Simon, 1984). ID
speech also facilitates learning in infants, especially learning language skills (Femald
& Simon, 1984). For instance, ID speech has been shown to promote word learning
and word segmentation (Thiessen, Hill, & Saffran, 2005).
Previous research in this laboratory has investigated the effects of ID speech
on infant learning with an associative learning paradigm. Association learning is a
basic form of learning present in infants where a conditioned stimulus (CS) comes
before and predicts a reinforcing unconditioned stimulus (UCS), and learning occurs
5


when the presentation of the CS creates an expectation that the UCS will come after
the CS (Kaplan, Fox, and Huckeby, 1992). Associative learning can be measured
through changes in responding to the CS as a result of CS-UCS pairings (Kaplan,
Fox, and Huckeby, 1992). Using the presentation of a smiling face as the UCS and
ID speech or adult directed (AD) speech as the CS, it was found that infant
association learning was facilitated more with ED speech than with AD speech
(Kaplan, Jung, Ryther, & Zarleng-Strouse, 1996).
ID speech produced by depressed mothers has been shown to have
significantly less fundamental frequency modulation than ID speech of nondepressed
mothers (Kaplan, Bachorowski, Smoski, & Zinser, 2001). Kaplan, Bachorowski,
Smoski, & Hudenko (2002) found that 4-month-old infants of depressed mothers did
not show associative learning in response to their own mothers ID speech, but did
show learning in response to an unfamiliar nondepressed mothers ID speech.
In a follow-up study, Kaplan, Dungan, & Zinser (2004) found that older
infants of chronically depressed mothers did not learn in response to their own
mothers or an unfamiliar nondepressed mothers ID speech, but did leam in response
to an unfamiliar nondepressed fathers ID speech. The infants of chronically
depressed mothers actually learned better from the unfamiliar nondepressed fathers
ID speech than infants of nondepressed mothers (Kaplan et al., 2004). ID speech is
thought to facilitate infant language development (Femald & Simon, 1984; Thiessen,
Hill, & Saffran, 2005), and therefore the lack of associative learning in infants of
6


depressed mothers with familiar and unfamiliar mothers ID speech may have a
negative impact on infant language development.
Maternal Depression and Infant Cognitive and Language Development
Many research studies have investigated the relationship between postpartum
depression and child cognitive and language development. Murray (1992) found that
infants of mothers with postpartum depression at 9 months were more likely to fail an
object concept task than infants of non-depressed mothers. The researchers found that
maternal depression did not account for performance of the infants on the Bayley
Scales of Mental Development, though, but that social class was a significant factor
(Murray, 1992). Murray, Fiori-Cowley, Hooper, and Cooper (1996) found that 18-
month old boys of mothers with postpartum depression had significantly lower scores
on the Bayley (cognitive) than boys of non-depressed mothers or girls. Whiffen and
Gotlib (1989) found that infants of depressed mothers had a poorer performance on
the cognitive scale of the Bayley than infants of depressed mothers, but the greater
amount of negative emotions expressed during the testing by the depressed infants
may have had an effect on this.
Sharp, Hay, Pawlby, Schmucker, Allen, and Kumar (1995) found that 4-year
old boys of mothers who had postpartum depression scored significantly lower on the
verbal subscale of the McCarthy Scales of Childrens Abilities than girls of mothers
with postpartum depression or children of non-depressed mothers. This effect was
7


significant even after controlling for variables such as maternal and paternal IQ and
mother-child interaction (Sharp et al, 1995). Cox, Puckering, Pound, and Mills
(1987) found that maternal depression affected language development in two-year
olds. The two-year-old children of depressed mothers were more delayed in
expressive language development than the children of non-depressed mothers (Cox et
al, 1987). Brennan et al. (2000) found that the Peabody Picture Vocabulary test score
of 5-year olds was not significantly related to timing of depression (e.g., postpartum),
but to severity and chronicity. Research carried out as part of the NICHD Early Child
Care Research Network (1999) showed that children of depressed mothers did not
perform as well on measures of expressive language skills and verbal comprehension
at 36 months old as children of non-depressed mothers. There have not been studies
published on the effects of maternal depression on language development in infancy,
and it is important to determine if effects of maternal depression on child language
development start in infancy or if it is something that occurs later on.
Maternal Depression and Mother-Infant Interactions
Mother-infant interactions have been found to be affected by maternal
depression (e.g. NICHD Early Child Care Research Network, 1999). Research has
shown that depression can affect maternal sensitivity (NICHD Early Child Care
Research Network, 1999), maternal attachment (Teti, Gelfland, Messinger, &
Isabella, 1995), and maternal interaction styles (Cohn, Matias, Tronick, Connell, &
8


Lyons-Ruth, 1986), which may then increase the risk for negative cognitive and
language outcomes for children (NICHD Early Child Care Research Network, 1999;
Jones et al., 1997).
Maternal Sensitivity
Maternal sensitivity may play a role in the relationship between maternal
depression and child language development. Maternal sensitivity includes many
components, including maternal responsiveness, affect, maternal acceptance, and
flexibility in attention and behavior (Biringen, Robinson, & Emde, 1993). Mothers
with depression have been found to be less responsive and sensitive to their infants
needs (NICHD Early Child Care Research Network, 1999). Sensitive, warm and
contingent maternal interactions have been associated positively with childrens
language competence (Sohr-Preston & Scaramella, 2006). Children at 36 months old
whose mothers had higher levels of sensitivity performed better on cognitive and
language tests, regardless of maternal level of depression (NICHD ECCRN, 1999).
Having a depressed mother with lower levels of maternal sensitivity partly explained
a child's lower levels of verbal comprehension and expressive language. This
suggests that maternal depression may lead to less maternal sensitivity, which
therefore may lead to poorer cognitive and language development in children
(NICHD ECCRN, 1999).
9


Some research has examined the effects of postpartum depression on infant
language and cognitive development, but little research has examined factors that
may potentially mediate any relationship between those two variables. Burgess
(2006) found that maternal sensitivity was a better predictor of the infants associative
learning than maternal depression. Perhaps maternal sensitivity also mediates the
relationship between maternal depression and infant language and cognitive
development. This is an important area to study to gain an understanding of when
and how problems in language development start.
Tnfant-Matemal Attachment
Maternal depression has been shown to have an effect on infant attachment.
Teti, Gelfland, Messinger, and Isabella (1995) found that maternal depression was
significantly associated with insecure attachment in infants and preschool aged
children. It has also been found that lifetime maternal depression along with a
comorbid DSM-IV Axis I diagnosis was significantly associated with insecure
attachment with infants (Carter, Garrity-Rokous, Chazan-Cohen, Little, & Briggs-
Gowan, 2001). Murray et al. (1996) found that postpartum maternal depression
increased the risk for insecure attachment with the infant at 18 months.
It has been found that attachment security is related to language development.
Murray and Yingling (2000) found that secure attachment measured in 21-month-old
toddlers was positively related to toddler language development at 24 months.
10


Spieker, Nelson, Petras, Jolley, and Barnard (2003) found that children who were
securely attached at 19 months old scored significantly higher on the Peabody Picture
Vocabulary Test-Revised when they were 36 months old than the children who were
insecurely attached at 19 months, although the other language measures that they
used at 24 and 30 months were not found to be significantly different. Van
Ijzendoom, Dijkstra, and Bus (1995) conducted a meta-analysis of studies on infant
attachment and language development, and concluded that securely attached children
appeared to have significantly better language development than insecurely attached
children.
Gersten, Coster, Schneider-Rosen, Carlson, and Cicchetti (1986) found that
attachment had a protective effect for maltreated 24-month-olds. The maltreated
toddlers who were securely attached had better language development than the
maltreated toddlers who were insecurely attached, and the securely attached
maltreated toddlers had similar levels of language development as nonmaltreated
securely attached toddlers (Gersten et al., 1986). It is possible then that attachment
may also have a protective effect for language development of infants of depressed
mothers.
Types of Depressed Maternal Interactions
Mothers who have depression typically have one of two interacting styles with
their infant, withdrawn or intrusive (Cohn, Matias, Tronick, Connell, & Lyons-Ruth,
11


1986). Withdrawn depressed mothers are characterized by understimulating
interaction styles, and intrusive depressed mothers are characterized by
overstimulating interaction styles (Cohn et al., 1986). There are also some depressed
mothers whose interaction style closely resembles that of the nondepressed mothers
(Field, Healy, Goldstein, & Guthertz, 1990). Tronick and Weinberg (1997) found that
intrusive depressed mothers disrupted infants activities and infants reacted by getting
angry and pushing the mother away, which would occasionally work and repair the
mismatch. Withdrawn depressed mothers did not engage with their infant and
therefore left the infant in a dysregulated state that resulted in the infant self-
comforting and withdrawing (Tronick & Weinberg, 1997).
The withdrawn depressed mothers have been found to have relative right
frontal EEG activation as compared to the intrusive depressed mothers who have
relative left frontal EEG activation (Jones, Field, Fox, Davalos, Malphurs, &
Carraway, 1997). Right frontal EEG activation typically occurs with withdrawal
behaviors, and left frontal EEG activation typically occurs with approach behaviors.
Infants of these mothers have been found to have similar EEG patterns as their
mothers by 6 months (Jones, Field, Fox, Davalos, Malphurs, & Carraway, 1997).
Infants of the withdrawn depressed mothers have been found to explore less and have
lower scores on the Bayley Scales at one year than infants of intrusive depressed
mothers (Jones, Field, Fox, Davalos, Malphurs, & Carraway, 1997). Infants of
intrusive depressed mothers have also been shown to have better differential
12


responding to facial expressions than infants of withdrawn mothers (Diego, Field,
Hart, Hemandez-Reif, Jones, & Cullen 2002).
The Role of the Father
It has been found that paternal sensitivity and paternal involvement are
significantly related to toddler cognitive development (Easterbrooks & Goldberg,
1984; Shannon, Tamis-LeMonda, London, & Cabrera, 2002). Some have
hypothesized that fathers are less aware of their childs language abilities, which
results in fathers being more likely to use more complicated language (McLaughlin,
White, McDevitt, & Raskin, 1983). This may have an effect on child language
development by increasing the childs language ability.
There have been interesting findings about the infants relationship with the
father when the mother is depressed. Nondepressed fathers have significantly more
positive interactions with their infant than the infants depressed mother. (Hossain,
Field, Gonzalez, Malphurs, & Del Valle, 1994). Infants do not interact more
negatively when their fathers are depressed, in comparison to non-depressed fathers.
Hossain et al. (1994) proposed based on these findings that fathers may act as a buffer
for the effects of maternal depression on infants. Consistent with this, Kaplan et al.
(2004) found that infants of chronically depressed mothers did not learn from their
own mothers ID speech or from an unfamiliar nondepressed mothers ED speech, but
did learn from the unfamiliar nondepressed fathers ID speech. So it is possible that
13


fathers may act as a buffer or serve as a protective factor for the effects of maternal
depression on infant language development.
The Current Study
The objectives of this study were to examine the relationship between
postpartum depression and infant language development. The main hypothesis of this
study is that infants of depressed mothers will have less language development than
infants of non-depressed mothers, and this relationship will be mediated by maternal
sensitivity. Maternal sensitivity was also hypothesized to be positively related to
infant language development. Along with this, it was thought that the effects will be
more pronounced for male infants of depressed mothers than for female infants.
In addition, it was hypothesized that language development will be
significantly related to infants associative learning with ID speech. ID speech
appears to generally facilitate infant learning, and a growing body of evidence shows
that it also facilitates aspects of rudimentary language processing, including phoneme
discrimination, speech stream segmentation, and word learning (Thiessen, Hill, &
Saffian, 2005).
The effect of attachment security with mothers and fathers on infant language
development was also investigated. It is thought that infants with secure attachment
will have better language development. What effect father attachment and father
involvement has on infant language development will be explored. The effect of
14


intrusive versus withdrawn style of depressed parenting on infant language
development was examined as well.
15


CHAPTER 2
METHOD
Participants
Depressed and non-depressed participants were recruited through magazine
advertisements and flyers distributed at Early Head Start Centers. A total of 44
women and their infants participated in the study, and were paid for their
participation. The womens age ranged from 20 to 40 years, with a mean age of 30.84
(SD = 4.94). Infant age ranged from 319 to 429 days, with a mean age of 366 days
(SD = 31). Household income ranged between $0-$6,000 and more than $50,000,
with a median income of more than $50,000. Thirty-two (72.7%) of the mothers were
Caucasian, seven (15.9%) were Latina, one (2.3%) was African-American, three
(6.8%) were Asian, and one (2.3%) was Native American.
Of the 44 women who participated in the study, 30 of the women also had the
fathers of the infant involved in the study and each parent was paid for their
participation. These womens ages ranged from 20 to 40 years, with a mean age of
31.30 (SD = 5.25). Infant age ranged from 319 to 417 days, with a mean age of 369
days (SD = 25). The mens age ranged from 24 to 47 years, with a mean age of 32.30
years (SD = 5.90). Household income ranged between $0-$6,000 and more than
16


$50,000, with a median income of more than $50,000. Twenty-four (80%) of the
mothers were Caucasian, four (13.3%) were Latina, and two (6.7%) were Asian.
Twenty-four (80%) of the fathers were Caucasian, four (13.3%) were Latino, and two
(6.7%) were Asian.
Measures
Structured Clinical Interview for DSM-IV-TR
The Structured Clinical Interview for DSM-IV-TR Axis I Disorders -
Research Version is a semi-structured interview used to make DSM-IV Axis I clinical
diagnoses (First, Gibbon, Spitzer, & Williams, 2001). The research version of the
SCID-I is longer than the clinician version and contains severity and course
specifiers, subtypes of disorders, and allows researchers to code details of past mood
episodes (First et al., 2001). It asks a combination of closed and open-ended
questions to ascertain whether or not the respondent meets diagnostic criteria for the
different clinical diagnoses (First et al., 2001). The SCID-I has been shown to have
test-retest reliability values of .70 to 1.00 (First et al., 2001). The SCID-I also has
been shown to be a valid instrument when compared to standard clinical interviews
(First et al., 2001).
Each mother was interviewed using the SCID-I by a trained M.A. level
graduate student. Interviews lasted about 30 to 60 minutes. There were a few mothers
17


who met DSM-IV criteria for Major Depressive Disorder (MDD; n = 2). There was
also a small number of mothers who were classified as being in partial remission from
a Major Depressive Episode (n = 6). In addition, there were five mothers who were in
full remission from a Major Depressive Episode that had occurred since the birth of
their child. In total, there were 13 mothers who experienced a Major Depressive
Episode since the birth of their child.
Beck Depression Inventory II
The Beck Depression Inventory II is a brief, self-report measure consisting of
21 questions about depressive symptomatology (BDI-II; Beck, 1996). Respondents
rate the intensity of the depressive symptoms over the past two weeks on a scale of 0
to 3, and then the responses are added up to come up with a total score (Groth-
Mamat, 2003). Scores of 13 and below are considered to indicate none or very mild
depression, and scores 14 and above are thought to indicate moderate depression to
severe depression as the scores get higher. The BDI-II has been found to have a test-
retest reliability over one week of .93 (Groth-Mamat, 2003). The BDI-II also has
concurrent validity due its moderate to high correlations of clinical ratings for
psychiatric patients (Groth-Mamat, 2003). The BDI-II has been found to be able to
discriminate between psychiatric from normal populations, suggesting a high level of
discriminant validity (Groth-Mamat, 2003).
18


Due to the small number of mothers in the study meeting criteria for a current
DSM-IV diagnosis of major depression, BDI-II scores were used to classify mothers
as being in the elevated BDI-II category (BDI-II >13) or in the non-elevated category
(BDI < 13). There were 12 mothers who were included the elevated BDI-II category
(M= 19.00, SD = 6.52) and 32 mothers who were included the non-elevated BDI-II
category (M= 5.91, SD = 4.37). Both mothers who met criteria for Major Depression
had BDI-II scores that were greater than 13.
Bavlev Scales of Infant and Toddler Development- Third Edition
The Bayley-III (2005) assesses cognitive, motor, language, social-emotional
and adaptive behavior development. The language subscale evaluates expressive and
receptive communication (Bayley, 2005). Items include assessing babbling,
responding to sounds, gesturing, and naming objects (Bayley, 2005). The Bayley-III
language subscales have been shown to be reliable measures, with high internal
consistency (reliability coefficients of .87 and .91; Bayley, 2005). The Bayley-III
language subscales have also been shown to be valid instruments; they have
convergent validity with assessments that examine similar constructs (Bayley, 2005).
M.A. level graduate students were trained to administer the Bayley-III.
Administrations of the language scales of the Bayley-III lasted from 15 to 40 minutes.
19


Mac Arthur Vocabulary Checklist
The MacArthur Short Form Vocabulary Checklist: Level I is a parent report of
the childs productive and receptive vocabulary. This measure is a checklist of 89
words for the parents to rate whether their child understands or understands and says
each word. The MacArthur Short Form Vocabulary Checklist: Level I has been
shown to have an internal consistency reliability coefficient of .97 (Fenson et al.,
2000). This measure also had concurrent validity coefficients of .90 for productive
vocabulary and .88 for receptive vocabulary with the long form of the MacArthur
Vocabulary Checklist: Level I (Fenson et al., 2000).
Associative Learning Paradigm
The associative learning paradigm was administered to assess infant
associative learning in response to infant-directed speech. Infant-directed speech
samples were recorded separately for both the mother and father. A hand-held Sony
microphone and a Sony TCM 5000EV tape recorder/player were used to record the
speech samples. The parent was asked to interest the infant in a stuffed toy gorilla by
using the phrase pet the gorilla. The speech samples were then edited so that the
sample contained two questions and then one statement (e.g. Can you pet the
gorilla? Will you pet the gorilla? Pet the gorilla.) and repeated once to make a 10
second speech sample.
20


The associative learning paradigm was conducted with two sessions, each
lasting 5 minutes. For the first visit, the infant heard an unfamiliar nondepressed
mothers speech sample for one session and an unfamiliar nondepressed fathers
speech sample in the other session with the order being counterbalanced. During the
second visit, the infant heard their own mothers and fathers speech samples in the two
sessions with the order counterbalanced.
During each session, the infant was seated on the mothers lap and looked at a
translucent Plexiglass screen mounted inside of a board. For the first six conditioning
trials, the infant heard a 10 second speech segment and would then immediately see a
picture of a smiling womans face projected onto the screen for 10 seconds. After
each conditioning train there was a 10 second break with no speech segment being
played and the screen being illuminated by light. Then at the end of the sixth trial,
there were four 10 second postconditioning trials using a checkerboard stimulus that
was presented for 10 seconds with a 10 second break after each presentation. For the
first and fourth presentations of the checkerboard stimulus, the speech segment was
playing simultaneously. For the second and third presentation of the checkerboard,
there was no speech segment. One to two observers recorded the amount of looking at
the stimulus by the infant (Kaplan, 2004).
21


Emotional Availability Scales
This study also used the Emotional Availability Scales (EAS) to rate maternal
sensitivity and maternal structuring. The EAS consists of 5 scales measuring the
emotional availability of mother toward the child and child toward the mother
(Biringen et al., 1993). These scales are coded from a 10-15 minute play interaction.
Only the Maternal Sensitivity and Maternal Structuring scale were used in the current
study.
The Maternal Sensitivity scale looks at aspects such as affect, responsiveness,
maternal acceptance, variety and creativity in play, flexibility in attention and
behavior, and how conflict situations are handled during the interaction (Biringen et
al., 1993). This scale is not directly linear. A rating of 10 refers to the mother being
hyper-sensitive and anxious. The optimal rating is 9, where the mother is highly
sensitive and shows genuine pleasure and interest in the infant. For ratings below 9
and down to 1, the mother is rated as more insensitive and for the lower ratings is
either affectively flat or harsh (Biringen et al., 1993).
The Maternal Structuring/Intrusiveness scale looks at aspects of the
interaction such as structuring play, letting the child lead in the interaction, and
setting limits for the childs behavior. This scale is not directly linear either. Values
range from 9, where the mother is overly high in structuring and is too controlling, to
1, no structuring where the mother does not set limits for child behavior and does not
22


provide the child with support. The optimal rating is 5, where the mother provides
suitable levels of structuring and support while letting the child lead and explore.
Interrater reliability for this measure has been found to be greater than .80 in
research studies (Biringen, 2000). The EAS scales have also been shown to have
construct validity due to findings that .. .parental and child EA are related to
attachment, as well as to other meaningful aspects of the parent-child relationship.
(Biringen, 2000, p. 107). The videotaped play interactions were coded by trained
observers who were graduate students in the M.A. program and were blind to the
SCED diagnosis and BDI-II score of the parent.
Attachment O-Set
The Attachment Q-Set (AQS, Waters, 1987) was used to assess attachment
security for the mother and for the father. The AQS consists of 90 items measuring
secure base behavior and security of attachment, as well as filler items. A home visit
was conducted by two trained independent observers for a period of about 1.5 hours.
The primary focus of the home visit was to observe, but the parent was asked a
couple of questions and a toy was brought out for the last 10-15 minutes of the visit
for the parent and child to play with. After the visit, the visitors sorted the items into 9
piles and each item was assigned a numerical value 1-9. The item values were then
averaged between the two visitors and correlated with the ideal secure sort to come
up with the dyads security score. The AQS has been shown to have construct validity
23


in assessing attachment security due to its strong association with the Strange
Situation (Vaughn & Waters, 1990).
Child Care Activity Questionnaire
The Child Care Activity Questionnaire (CCAQ) is a parent report that
assesses the percentage of the time child care activities are performed by the mother
alone, the father alone, by both parents together, and by other family members
(Montague & Walker-Andrews, 2002). The CCAQ was used to assess the percent of
time the father spent taking care of the infant alone and playing with the infant alone.
For infant care, 14 items related to infant care were averaged to come up with the
mean percentage of time the father spent taking care of the infant alone. For playing
with infant, 5 items related to playing with the infant were averaged to come up with
the mean percentage of time the father spent playing with the infant alone.
Procedure
When the mother and infant arrived, the researcher went over the consent
form with the mother and obtained her consent. After that, the mother and infant
participated in two 5-minute associative learning tasks. The mother then filled out the
BDI-II, the MacArthur Short Form Vocabulary Checklist, and demographic
questionnaires. After that was done, the infant participated in the Bayley-III
assessment and was administered the cognitive and language subscales. After the
24


infant finished, the researcher administered the SCID-I to the mother. That concluded
the first visit.
If the father is participated in the study as well, the mother, father, and infant
came to the laboratory together for a second visit. The researcher first explained the
study and obtained consent from both parents. Then each parents voice was recorded
trying to interest their child in a toy with the order counterbalanced. After that was
complete, each parent was videotaped playing with their child for a 10-15 minute
period, and the order was counterbalanced. Then the mother and infant participated in
two 5-minute associative learning tasks. Lastly the father filled out the BDI-II and
demographic questionnaires. That concluded the second visit.
For the third and fourth sessions, home visits were conducted and attachment
was assessed using the Attachment Q-Set (see above).
25


CHAPTER 3
RESULTS
Demographic Data
Demographic data for mothers with elevated (BDI > 13) and non-elevated
(BDI < 13) BDI-II scores and their infants are displayed in table 3.1. Chi-square tests
showed that mothers with elevated BDI-II scores were significantly less likely to be
married (x2 = 5.44,/? = .02) and significantly more likely to be of minority ethnic
status (x2 = 4.30,/? = .04). Maternal age, infant age, family income, and maternal
education were not significantly different as a function of maternal elevated vs. non-
elevated BDI-II categories.
Within the sub-sample that had fathers participate, there were a few
significant demographic differences between the maternal elevated vs. non-elevated
BDI-II categories. Mothers in the elevated BDI-II differed significantly from the
mothers in the non-elevated BDI-II category in amount of education (M 5.22, SD =
1.30; M= 6.38, SD = 1.36), (F(l, 29) = 4.69,/? = .04). In addition, the level of
education of the fathers differed significantly between the elevated maternal BDI-II
category and the non-elevated maternal BDI-II category (M= 4.89, SD 0.93; M-
6.10, SD = 1.45), (F(l, 29) = 5.27,p- .03). A chi-square test showed that mothers
with elevated BDI-II scores were significantly less likely to be married (x2 = 5.00,
26


p = .03). Maternal age, paternal age, infant age, and family income were not
significantly different as a function of maternal elevated vs. non-elevated BDI-II
categories.
Table 3.1 Demographic Data
Variable Elevated BDI Non-elevated BDI
N 12 32
Maternal age 30.8(6.6) 31.0(4.3)
Infant age (days) 372 (31) 365 (32)
Family income 5.6 (2.8) 6.8 (1.9)
Mothers education 5.3 (1.6) 6.2 (1.3)
Ethnicity
Caucasian 6 (50%) 26 (81.2%)
Latina 4 (33.3%) 3 (9.4%)
African-American 0 1 (3.1%)
Asian 1 (8.3%) 2 (6.2%)
Native American 1 (8.3%) 0
Percent married 66% 94%*
Note: Income (family yearly), 1.0 = $0-6,000, 2.0 = $6,000-10,000, 3.0 = $11,000 -
20,000, 4.0 = $21,000-25,000, 5.0 = $26,000-30,000, 6.0 = $31,000-40,000, 7.0 =
$41,000-50,000, 8.0 = above $50,000; Education, 3.0 = no high school degree, 4.0 =
high school degree, 5.0 = 2-year degree, 6.0 = 4-year college degree, 8.0 = advanced
degree. Parentheses indicate standard deviations. *p <.05, **p <.01
27


Power Analysis
The power of the study was .8, the typically accepted power value. With a
power of 0.8 and alpha = .05, a sample size of 52 was needed to detect a large effect
size (f = .40) for a one-way ANOVA. For single predictor linear regressions, a sample
size of 25 was needed to detect large effect sizes (f2 = .35) and a sample size of 54
was needed to detect medium effect sizes (P = .15). For correlations, sample size of
21 was needed to detect a large effect size (r = .50) and a sample size of 64 was
needed to detect a medium effect size (r = .30). This study had a sample size of 44
and a subsample size of 30. The sample size was not large enough to detect large
effect sizes when conducting one-way ANOVAs and not large enough to detect small
and medium effect sizes when conducting bivariate correlations and linear
regressions. Small and medium effect sizes are the effect sizes that are typically
associated with studies looking at postpartum depression (e.g. Beck, 1998).
Language Development and Maternal Depression
There was complete data for the language subscale of the Bayley-III for 44 of
the 47 infants that participated. Two were unable to finish the language subscale of
the Bayley-III due to fussiness, and 1 had Spanish spoken primarily in the home so
the results were not used. All three mothers were non-depressed and did not have
elevated BDI-II scores (BDI < 13). Throughout the results section the results for the
total language score of the Bayley-III will be reported and the expressive and
28


receptive subscale scores will only be reported when there is a significant difference
from the total language score.
It was hypothesized that maternal depression would have a significant effect
on infant language development. The mean total language score of the Bayley-III for
infants with mothers who were in the elevated BDI-II category was slightly lower (M
= 19.00, SD = 3.95) than of the non-elevated category (M= 19.63, SD 2.83),
although there was no significant difference (F(l, 43) = 0.34,/? = .562). There was
also no significant correlation between maternal BDI-II score and total language
subscale score of the Bayley-III (r =-.03,/? = .86).
We also tested whether maternal depression at any point in the postpartum
period affected infants Bayley-III language scores. Infants of mothers who had a
major depressive episode since birth (n=13) did not differ significantly from those of
non-depressed mothers (n=30) with the total Bayley-III language score (F(l, 42) =
1.04,/? = .31) or receptive score (F(1,42) = 0.02,/? = .89). However, there was a
significant difference the expressive score (F(l, 42) = 6.879,/? = .01), with infants of
mothers who had been depressed since birth scoring lower (M= 9.54, SD = 1.33) than
infants of non-depressed mothers (M= 10.80, SD = 1.49).1
Infant Gender. Due to a larger proportion of boys than girls in the maternal
elevated BDI-II score group (9 boys and 3 girls), an analysis examining the
1 There were no other significant effects found for whether the mother had a major depressive episode
postnatally or if the mom was nondepressed.
29


relationship between infant gender, maternal BDI-II category, and infant Bayley-III
language score was not run because there was not enough power to detect a
difference.
Maternal Sensitivity and Structuring
Maternal Sensitivity. The subsample was used to examine maternal
sensitivity. There were 2 mothers in the subsample who were rated as oversensitive
and were excluded from the analysis. Inconsistent with the a priori hypothesis, there
was a significant negative correlation between the total language subscale score and
maternal sensitivity (r= -.39, p =.04). The correlation between maternal sensitivity
and the receptive (r= -34, p = .08) and expressive (r= -.35, p = .07) language
subscale scores approached significance.
Maternal sensitivity may look different depending on the infants age. It is
possible that infants who are older and more verbal may need mothers to give them
space to explore. Older infants may be better able to let mothers know when they
need something, so less verbal children may elicit higher levels of maternal
sensitivity. Mothers of older infants may therefore appear to be less sensitive while in
reality being more sensitive to their infants needs. A partial correlation between
maternal sensitivity and Bayley-III language score was conducted, controlling for
30


infant age. This correlation between Bayley-III language score and maternal
sensitivity was significant (r= -.40, p = .04).2
Maternal Structuring. The subsample was used to study maternal structuring.
Maternal structuring categories were used to examine the differential effects of
mothers who were withdrawn and who were intrusive on infant language
development. Structuring scores 4.5 and below were classified as under-structured,
score between 5 and 6.5 were classified as optimally structured, and scores that were
7 and above were classified as over-structured to test for differences between the
categories. It was hypothesized that infants whose mothers were optimally structuring
would have higher language scores than infants whose mothers were under- or over-
structuring. Mothers who are more withdrawn would tend to be classified as under-
structured and mothers who are more intrusive would tend to be classified and over-
structured. Maternal structuring category did not have a significant effect on language
scores (F(l, 28) = 1.54,/? = .23), and mothers who were optimally structured (M =
18.73, <£D =1.8 7) were not significantly different from mothers who were over-
structured (M = 20.30, SD = 4.15) or under-structured (M= 21.20, SD = 3.77).
2 The mediation model of maternal depression, maternal sensitivity, and infant Bayley-III language
subscale score could not be tested (Baron & Kenny, 1986) because the relationship between elevated
maternal BDI-II scores and infant language development was not significant in step one, standardized
B = -.108, <26) = -.54, p = .59.
31


Infant Associative Learning Score
It was hypothesized that there would be a significant positive relationship
between infant maternal and paternal associative learning scores and the Bayley-III
language scale score. There was complete data for the conditioning trials for 33 of the
44 infants tested with the unfamiliar fathers voice (75%) and for 36 of the 44 infants
tested with the unfamiliar mothers voice (82%). Data was not available for the 11
and 8 conditioning trials respectively due to fussiness and crying by the infants. There
was no significant relationship between the associative learning score with the
unfamiliar mothers ID speech and the total language scale score (r = -.12,p = .49) of
the Bayley-III. In addition, there was no significant relationship between the
associative learning score with the unfamiliar fathers ID speech and the total
language scale score (r =- .09, p = .62).
There was complete data for the conditioning trials for 21 of the 30 infants
tested in the subsample that had fathers participate in the study (70%). Data was not
available for 9 infants due to fussiness and crying. There was no significant
relationship between the associative learning score with the familiar mothers ID
speech and the total language subscale score (r .05, p .84) of the Bayley-III. In
addition, there was no significant relationship between the associative learning score
with the familiar fathers ID speech and the total language subscale score (r = .08,/?
= .72).
32


Attachment
Attachment was studied using the subsample of mothers, fathers, and infants.
The Attachment Q-Set (AQS) was used to examine the effects of infant attachment on
infant language development. The AQS is strongly related through scientific research
and theory to the Ainsworth Strange Situation classification, which has been used to
find that two-thirds of a sample from the normal U.S. population is securely attached
and one-third insecurely attached (Lamb, Thompson, Gardner, & Chamov, 1985).
Based on that and prior literature using the AQS to classify parent-child relationships,
the top two-thirds of maternal and paternal AQS scores were designated as secure and
the bottom one-third of AQS scores were designated insecure (Howes & Ritchie,
1999; Teti & Ablard, 1989). The total language score was not significantly related to
whether or not the infant was classified as securely attached to the mother (F(l, 29) =
-0.50, p = .48). Whether or not the infant was classified as securely attached to the
father was significantly related to the total language score (F(l, 26) = 4.48, p <.05).
Infants who were securely attached to their father had higher language scores (M=
20.61, SD = 3.35) than infants who were insecurely attached to their father (M=
18.00, SD = 2.18) (See Figure 3.1).
33


22.00-
<11
E 20.00
o
u
(/)
Of
03
W
3
O)
c
18.Q0-
>16.00-
m
if
m
2 14.00
V
s
12.00-
----------------------1----------------------------------(---------------------
Insecure Secure
Infant Father Attachment
Error ban: 95* Cl
Figure 3.1 Infant language score means as a function of father attachment category.
Attachment and Maternal Depression. Previous research has found that
maternal depression is significantly related to insecure attachment in infants and
posits that attachment may be one of the mechanisms responsible for the varied
outcomes of children of depressed mothers (e.g. Teti, Gelfland, Messinger, &
Isabella, 1995). A 2 (maternal secure vs. insecure) x 2 (maternal elevated BDI vs.
non-elevated BDI) univariate ANOVA was conducted to test for an interaction effect
34


with infant Bayley-III language score as the dependent variable. There was not a
significant main effect in for infant maternal attachment security (F(l, 29) = 2.74,p =
.11) or for maternal BDI-II category (F(l, 29) = \.21,p = .27). There was a
significant infant maternal attachment security x maternal BDI category interaction
(F(l, 29) = 9.39, p = .005), such that infants whose mothers had elevated BDI-II
scores and were securely attached had higher language scores than the infants whose
mothers had elevated BDI-II scores and were insecurely attached (See Figure 3.2).
Infant Maternal
Attachment
- Insecure
---Secure
Figure 3.2 Infant language score means as a function of infant mother attachment
category and maternal BDI-II category.
35


Further analysis was conducted to see if there was a significant difference
between Bayley-III language scores for infants of non-elevated BDI-II mothers
depending on whether the infant was classified as securely or insecurely attached to
the mother. There was no significant difference between the two infant attachment
groups (F(l, 20) = 1.31 ,p = .27).
We also tested whether infant paternal attachment had an effect on Bayley-III
language scores when the mother had elevated or non-elevated BDI-II scores. A 2
(paternal secure vs. insecure) x 2 (maternal elevated BDI vs. non-elevated BDI)
univariate ANOVA was conducted to test for an interaction effect with infant Bayley-
III language score as the dependent variable. As was already reported, there was a
significant main effect in for infant father attachment security (F(l, 26) = 6.18, p =
.02), but there was no main effect for maternal BDI-II category (F(l, 26) = 0.68,/? =
.42), and no paternal attachment security x maternal BDI category interaction (F(l,
26) = 2.07,/? = .16); but the means were in the predicted direction, for infants whose
mothers had elevated BDI-II scores and were securely attached to the father had
higher language scores, M= 21.20, SD = 3.27; for infants whose mothers had
elevated BDI-II scores and were insecurely attached to the father, M= 16.00, SD =
1.73.
Attachment and Maternal Sensitivity. Maternal attachment security category
was significantly related to maternal sensitivity (F(l, 26) = 4.93,/? = .036), with
mothers classified as being in the insecure attachment category having a lower
36


maternal sensitivity score (M= 6.83, SD = 0.83) than mothers classified as being in
the secure attachment category (M= 7.70, SD = 1.00). Maternal sensitivity ratings
were then categorized as high or low to test for an interaction effect with attachment
security for infant language scores. Ratings between 7 and 9 were included in the
high maternal sensitivity group {n- 19) due to including the generally sensitive and
highly sensitive rating categories as defined in EAS rating descriptors. Ratings of 6.5
or below were included in the low sensitivity group (n 8) due to including the less
sensitive and insensitive rating categories as defined in EAS rating descriptors.
Ratings above 9 (n = 2) were excluded due to mothers being rated as oversensitive.
A 2 (maternal secure vs. insecure) x 2 (maternal low sensitivity vs. high
sensitivity) univariate ANOVA was conducted to test for an interaction effect with
infant language score as the dependent variable. It was found that there was not a
significant main effect for maternal attachment security (F(3, 24) = 2.17,/? = .15) or
maternal sensitivity (F(3, 24) = 1.65,/? = .21). The interaction between maternal
attachment and maternal sensitivity approached significance (F(3, 24) = 3.72,p =
.066), such that that infants whose mothers had low sensitivity scores and were
securely attached had higher language scores than the infants whose mothers had low
sensitivity scores and were insecurely attached (see Figure 3.3).
37


Infant Maternal
Attachment
Insecure
Secure
Figure 3.3 Infant language score means as a function of mother attachment category
and maternal sensitivity rating category.
To test for any effects of father attachment category, a 2 (paternal secure vs.
insecure) x 2 (maternal low sensitivity vs. high sensitivity) univariate ANOVA was
conducted to test for an interaction effect with infant language score as the dependent
variable. It was found that there was a significant main effect for father attachment
security (F(3, 24) = 6.11, p = .02), but not for maternal sensitivity (F(3, 24) = 0.15,p
38


= .70). There was a non-significant paternal attachment x maternal sensitivity
interaction (F(3, 24) = 232,p = .14).
Father Involvement
The amount of time the father spent taking care of the infant and playing with
the infant was examined in relation to Bayley-III language scale score. Neither the
reported amount of time the father spent taking care of the infant nor the reported
amount of time the father spent playing with the infant were significantly correlated
with the Bayley-III language score (r .07,/? = .68, and r = .06,/? = .68,
respectively).
MacArthur Vocabulary Checklist
The MacArthur Short Form Vocabulary Checklist: Level I was also used to
assess infant language development and to compare to the Bayley-III language scale.
There was not a significant relationship between the mothers report of receptive
vocabulary and the Bayley-III language scale score (r =- .02,/? = .89) or the mothers
report of productive vocabulary and the language scale score (r =- .13,/? = .42). In
addition, there was not a significant relationship between the mothers report of
receptive vocabulary and the Bayley-III receptive language subscale score (r = .02,/?
= .89), or between the Bayley-III expressive language subscale and the mothers
report of productive vocabulary (r = .05,/? = .76). The relationship between the
39


fathers report of receptive vocabulary and the Bayley-III language scale score was
not significant (r .1 l,p = .62), but the relationship between the fathers report of
productive vocabulary and the total language score approached significance (r = .41,
p = .057). There was not a significant relationship between the fathers report of
receptive vocabulary and the Bayley-III receptive language subscale score (r = .09,/?
= .69), but the relationship between the expressive language subscale and the fathers
report of productive vocabulary approached significance (r = .38, p = .08).
Whether the mother was in the elevated vs. not elevated BDI-II category was
not significantly associated with the mothers report of receptive vocabulary (F(l, 39)
= 1.44,/? = .24) or the fathers report of productive (F(l, 24) = 4.04,/? > .05) or
receptive (F(l, 24) = 1.291,/? = .27) vocabulary. The mothers report of productive
vocabulary was significantly associated with maternal BDI-II category, (F(l, 39) =
4.65,/? = .04) such that mothers reported higher levels of productive vocabulary if
they were in the elevated BDI-II category (.M= 8.09, SD = 12.14) than if the mother
was in the non-elevated BDI-II category (M= 2.90, SD = 3.20).
40


CHAPTER 4
DISCUSSION
This study examined the relationship between postpartum depression and
infant language development. It was hypothesized that infants whose mothers had
elevated symptoms of depression, low levels of sensitivity, or low rated security of
attachment with their infants would show slower language development. In addition,
links between infant associative learning in response to maternal and paternal ID
speech and infant language development, and the roles of father-infant attachment
and father involvement, were explored. Infants of mothers who had had an episode of
major depression at some point after the birth had lower expressive language scores
than infants of mothers who did not have a depressive episode. Although there were
no main effects of current elevated BDI-II scores or mother-infant attachment
security, there was a significant interaction between the two factors such that infants
of mothers with current elevated depression scores showed lower levels of language
development, except when they were securely attached to their mothers. In addition,
infants who were rated as securely attached to their fathers showed higher levels of
language development. Curiously, maternal sensitivity was negatively correlated with
infant language development. No other effects were significant.
41


Infants of mothers who had a major depressive episode at some point in the
postpartum period had significantly lower expressive language scores on the Bayley-
III than if the mother did not have a major depressive episode postpartum. Our
finding that infants of mothers who had a major depressive episode at some point in
the postpartum period had significantly lower expressive language scores on the
Bayley-III is consistent with Cox et al. (1987) and the NICHD Early Child Care
Research Network (1999) who found that maternal depression had a significant effect
on child expressive language development with 2 and 3-year old children. These
researchers proposed that the effects on expressive language development could be
due to depressed mothers being less sensitive caregivers, but this study did not find
that mothers were less sensitive with their infant at 12 months when they had a major
depressive episode postpartum. However, these researchers did not specifically
examine maternal depression within the first year of the childs life. Sharp et al.
(1995) found that 4-year-old children of mothers who had been depressed during the
childs first year of life had significantly lower verbal scores on the McCarthy Scales
of Childrens Abilities than the children of non-depressed mothers. Sharp et al. (1995)
discussed that it is possible that when the mother was experiencing the major
depressive episode she may have had problems interacting with the child that would
not continue when the depressive episode was over but would still impact the childs
cognitive and language development down the road. These problems with interacting
early in the childs life could affect the childs capacity to regulate emotions and
42


impact the childs brain development (Tronick & Weinberg, 1997) and may have
lasting impacts on the childs language development. It could be that mothers who
had an episode of postpartum depression but are now in partial or full remission are
able to be sensitive caregivers currently but may not have been earlier in the infants
life and we are seeing those effects on the infants language development. Further
research is needed that assesses maternal sensitivity during the postpartum depressive
episode and then later assesses infant language development before any conclusions
can be drawn.
This study did not find that infants of mothers with elevated BDI-II scores had
any significant differences in language development relative to infants of mothers
with non-elevated BDI-II scores, or that infant maternal attachment security had any
significant main effects on language development. However, infant attachment to the
mother was found to have a moderating effect on infant language development when
the mother had elevated levels of depression symptoms. An infant with secure
attachment to a mother who has elevated levels of depression symptoms may act as a
protective factor against negative language development outcomes. Infants who are
securely attached may have mothers who are more responsive and talk more to their
child. This finding is consistent with prior research that has found that infant maternal
attachment can be a protective factor against negative outcomes; previous research
has found that maltreated toddlers who were securely attached to the mother appeared
to be protected from lower levels of language development (Gersten, et al., 1986).
43


We found that there was a significant effect of infant paternal attachment
security on infant language development. Secure infant attachment to the father may
affect infant language development through fathers being likely to use more
complicated language with the infant than mothers (McLaughlin, White, McDevitt, &
Raskin, 1983) and also being better able to informally teach the infant by being more
responsive and not having the relational stress of insecure attachment (Gersten et al.,
1986). There have been studies that have shown that father sensitivity and
involvement are associated with better language development, but there have not been
studies that have looked at father attachment and infant language development. This
points to the importance of the father in child language development, and is a find
that needs further research to gain an understanding of how exactly paternal
attachment affects language development. There were few fathers in our sample with
elevated BDI-II scores, so it was not possible to test for an interaction effect with
infant attachment to the father and father BDI-II score on language development. This
would be important to examine in future research, and see if the finding this study had
with mothers also applies to fathers.
Surprisingly, this study found that maternal sensitivity was negatively related
to infant language development. This finding was not consistent with prior literature
that has found that higher levels of maternal sensitivity are related to higher levels of
language development (e.g., NICHD Early Child Care Research Network, 1999). One
possible explanation is that maternal sensitivity may look different depending on the
44


infants verbal ability. Infants that have more advanced language development may
be better able to communicate with the mother, so infants with less language
development may elicit higher levels of maternal sensitivity. Maternal sensitivity
ratings using EAS are also influenced by the infants behavior in the interaction, so
infants that have more advanced language development may not need to involve the
mother as much as infants with less advanced language development. Mothers of
more verbal infants may therefore appear to be less sensitive while in reality being
sensitive to their infants needs. Although an analysis controlling for infant age did
not reveal any significant differences on the effects of maternal sensitivity on infant
language development, this is still an avenue to explore and research further.
There was a near significant interaction between infant attachment to the
mother and level of maternal sensitivity for infant language development. Infants of
mothers who were low in sensitivity and were securely attached had much higher
Bayley-III language scores than the other infants. This may support the idea advanced
above about the effects of children who have more advanced language on perceived
maternal sensitivity. Sensitive caregiving is associated with secure infant attachment,
and attachment security speaks to the history of the relationship between the infant
and the caregiver. It is also possible that there may have been something else going
on in this subgroup that would make their level of language development more
advanced than the other groups. For instance, these infants may have been in daycare
so the effects of low maternal sensitivity would not be as salient for them. Due to the
45


small sample size, these effects would need to be replicated and investigated with
further research.
There was a near significant relationship between fathers report of productive
vocabulary and infant language development with the Bayley-III, but there was not a
significant relationship between the fathers report of infant receptive vocabulary and
the mothers report of infant productive and receptive vocabulary and language
development as measured by the Bayley-III. It is possible if the father is not around as
much he might be a more objective reporter than the mother. For the mothers report
of vocabulary, it was found that she reported significantly more expressive
vocabulary for the infant if she was in an elevated BDI-II category than if she was
not. This is interesting given that the mean Bayley-III language score was slightly
lower for infants of mothers who had elevated BDI-II scores than for infants of
mothers who did not have elevated BDI-II scores. It is possible that mothers with
elevated BDI-II scores sense that they are not talking to and interacting with their
infant as much and are overcompensating on the parental report of vocabulary.
Another possibility is that mothers who have elevated BDI-II scores may be more
sensitive to infants expressive language because the child is making noises and the
mothers are easily irritated due to the depression symptoms so they over perceive it.
There are many reasons why we may not have found a significant relationship
between elevated maternal BDI-II scores and infant language development. The
sample had few currently clinically depressed mothers, so it is possible that direct
46


effects of depression on infant language development would be greater with mothers
whose depressive symptoms are significantly affecting their functioning. BDI-II
assesses depressive symptoms but not the amount of clinically significant distress and
impairment. Mothers who have elevated BDI-II scores but are not clinically
depressed may be able to still be fairly responsive and verbal with their infant, and
therefore have fewer effects on their language development.
The finding with the interaction between infant maternal attachment security
and maternal level of depression may also help to explain why there was not a
significant association between levels of maternal depression and infant language
development. There may have been enough mothers with elevated BDI-II scores that
had infants that were securely attached to not show significant effects. This
interaction will be important to examine in further research.
Due to the small number of female infants of mothers with elevated BDI-II
scores, the effects of infant gender and maternal depression on infant language
development could not be examined. Previous research has found an effect of
maternal depression on infant gender in relation to language development in 4-year-
olds (Sharp et al., 1995), but other research has not found an effect on language
development in younger children. Further study needs to be done to find if effects on
language development would be more pronounced for male infants of depressed
mothers than for female infants.
47


Level of maternal structuring was used as a proxy to investigate if there was
an effect the intrusive versus the withdrawn style of depressed parenting on infant
language development. Previous research has found that the infants of withdrawn
depressed mothers experience more negative developmental effects than infants of
intrusive depressed mothers (Jones et al., 1997). There was no effect for maternal
structuring when looking at the total sample of mothers. Unfortunately, there were not
enough mothers with elevated BDI-II scores in the different structuring categories to
examine the effects that depression has on structuring and language development,
which will be important to look at in future studies.
Exploratory analyses were conducted to find what affect father involvement
has on infant language development. The amount of time the father spent taking care
of the infant and playing with the infant alone was not found to have a significant
effect on infant language development. This differs from previous literature that has
found an effect of father involvement on child cognitive development (Easterbrooks
& Goldberg, 1984; Shannon, Tamis-LeMonda, London, & Cabrera, 2002). This study
used the mothers report of father involvement, so it is possible that the mother was
not an accurate reporter. In future studies it may be helpful to look at the fathers
report of involvement and see if there is an effect there. This finding was interesting,
especially given the effect that was found for infant father attachment security on
infant language development. It is possible that it is not the amount of time that the
48


father spends with the infant, but the quality of time spent with the infant that is
associated with more advanced language development.
There was no relationship found between infant associative learning from
maternal or paternal ID speech and infant language development. This was not
consistent with the literature that has found that ID speech affects language
development (Thiessen, Hill, & Saffran, 2005). It may be that ID speech does not
have as great of a direct effect on infant language development as previously thought.
Further research needs to be done before drawing any conclusions.
This study was limited by the small sample size. The sample was also
primarily Caucasian and middle class, so its generalizability may be limited. In
addition, there may have been confounding variables that were not taken into account.
For instance, there may be some variability in language scores due to infant fussiness.
An episode of depression in the postpartum period was found to have a
significant effect on infant expressive language development. Further research is
needed to investigate exactly how maternal depression in the first year affects child
language development later on. Maternal sensitivity was found to be negatively
associated with infant language development. Examining the effects the infant has on
maternal sensitivity may help to understand this finding. This study also supports the
importance of attachment in infant language development. It appears that infant
attachment to the mother can act as a protective factor for infant language
development when the mother is depressed and possibly when the mother is low in
49


sensitivity. Father attachment was also found to play a significant role in infant
language development, which is something that has not been found before. Focusing
on the attachment relationship with both parents may be a good avenue to protect the
infant from developmental risks associated with maternal depression.
50


REFERENCES
Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978). Patterns of attachment: A
psychological study of the strange situation. Hillsdale, NJ: Erlbaum.
American Psychiatric Association. (2000). Diagnostic and statistical manual of
mental disorders, 4th edition, text revision. Washington, DC: American
Psychiatric Press, Inc.
Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51,1173-1182.
Bayley, N. (2005). Bayley Scales of Infant and Toddler Development- Third Edition.
San Antonio, TX: Harcourt Assessment
Beck, A. (1996). BDI-II. San Antonio, TX: Psychological Corporation.
Beck, C.T. (1998). The effects of postpartum depression on child development: A
meta-analysis. Archives of Psychiatric Nursing, 12,12-20.
Bertoncini, J., Morais, J., & Bijeljac-Babic, R. (1989). Dichotic perception and
laterality in neonates. Brain and Language, 37, 591-605.
Biringen, Z., Robinson, J.L., and Emde, R.N. (1993). Manual for scoring the
emotional availability scales infancy to early childhood version. Boulder, CO.
51


Biringen, Z. (2000). Emotional availability: Conceptualization and research findings.
American Journal of Orthopsychiatry, 70, 104-114.
Bowen, C. (1998). Ages and Stages: Developmental milestones for receptive and
expressive language development. Retrieved from http ://www. speech-language-
therapy. com/devel2.htm on (1/25/08).
Boysson-Bardies, B. (1999). How Language Comes to Children. Cambridge: MIT
Brennan, P.A., Hammen, C., Andersen, M.J., Bor, W., Najman, J.M., and Williams,
G.M. (2000). Chronicity, severity, and timing of maternal depressive symptoms:
Relationships with child outcomes at age 5. Developmental Psychology, 36, 759-
766.
Bruner, J. (1983). Childs talk. New York: Norton.
Burgess, A.P. (2006) Postpartum depression, maternal sensitivity, and infant
associative learning. Unpublished masters thesis, University of Colorado
Denver. Denver.
Carter, A.S., Garrity-Rokous, F.E., Chazan-Cohen, R., Little, C., & Briggs-Gowan,
M.J. (2001). Maternal depression and comorbidity: Predicting early parenting,
attachment security, and toddler social-emotional problems and competencies.
Journal of the American Academy of Child & Adolescent Psychiatry, 40, 18-26.
Cohn, J.F., Matias, R., Tronick, E.Z., & Connell, D. (1986). Face-to-face interactions
of depressed mothers and their infants. New Directions for Child Development.
34, 31-45.
52


Cox, A.D., Puckering, C., Pound, A. & Mills, M. (1987). The impact of maternal
depression in young children. Journal of Child Psychology and Psychiatry, 28,
917-928.
Cummings, E.M., & Davies, P.T. (1999). Depressed parents and family functioning:
Interpersonal effects and childrens functioning and development. In Joiner, T.,
and Coyne, J. C. (Eds.), The interactional nature of depression: Advances in
interpersonal approaches (pp. 299-327). Washington, DC: American
Psychological Association.
Diego, M.A., Field, T., Hart, S., Hemandez-Reif, M., Jones, N., & Cullen, C. (2002).
Facial expressions and EEG in infants of intrusive and withdrawn mothers with
depressive symptoms. Depression and Anxiety, 15,10-17.
Dobbing, J., & Sands, J. (1973). Quantitative growth and development of human
brain. Archives of Diseases of Childhood, 48, 757-767.
Easterbrooks, M.A., & Goldberg, W.A. (1984). Toddler development in the family:
Impact of father involvement and parenting characteristics. Child Development,
55, 740-752.
Fenson, L., Pethick, S., Renda, C., Cox, J. L., Dale, P.S., & Reznick, J.S. (2000).
Short-form versions of the MacArthur Communicative Development Inventories.
Applied Psycholinguistics, 21, 95-116.
Femald, A., & Simon, T. (1984). Expanded intonation contours in mothers speech to
newborns. Developmental Psychology, 20, 104-113.
53


Field, T., Fox, N.A., Pickens, J., & Nawrocki, T. (1995). Relative right frontal EEG
activation in 3- to 6- month-old infants of depressed mothers. Developmental
Psychology, 31, 358-363.
Field, T., Healy, B., Goldstein, S., & Guthertz, M. (1990). Behavior-state matching
and synchrony in mother-infant interactions of nondepressed versus depressed
dyads. Developmental Psychology, 26, 7-14.
First, M.B., Gibbon, M., Spitzer, R.L., and Williams, J.B.W. (2001). Users guide for
the Structured Clinical Interview for DSM-IV-TR Axis I disorders Research
version. New York: Biometrics Research.
Gersten, M., Coster, W., Schneider-Rosen, K., Carlson, V., & Cicchetti, D. (1986).
The socio-emotional bases of communicative functioning: Quality of attachment,
language development and early attachment. In M. Lamb, A.Brown, & B. Rogoff
(Eds.), Advances in developmental psychology. Hillsdale, NJ: Erlbaum.
Glanville, B.B., Best, C.T., & Levenson, R. (1977). A cardiac measure of cerebral
asymmetries in infant auditory perception. Developmental Psychology, 13, 54-59.
Grace, S.L., Evindar, A., and Sterwart, D.E. (2003). The effect of postpartum
depression on child cognitive development and behavior: A review and critical
analysis of the literature. Archives of Womens Mental Health, 6, 263-274.
Groth-Mamat, G. (2003). Handbook of psychological assessment, 4th edition. New
York: John Wiley & Sons, Inc.
54


Hossain, Z., Field, T.M., Gonzalez, J., Malphurs, J., & Del Valle, C. (1994). Infants
of depressed mothers interact better with their nondepressed fathers. Infant
Mental Health Journal, 15, 348-357.
Howes, C, & Ritchie, S. (1999). Attachment organizations in children with difficult
life circumstances. Development and Psychopathology, 11, 251-268.
Jacobson, J.L., Boersma, D.C., Fields, R.B., & Olson, K.L. (1983). Paralinguistic
features of adult speech to infants and small children. Child Development, 54,
436442.
Jones, N.A., Field, T., Fox, N.A., Lundy, B., & Davalos M. (1997). EEG activation in
1-month-old infants of depressed mothers. Development and Psychopathology, 9,
491-505.
Jones, N.A., Field, T., Fox, N.A., Davalos M., Malphurs, J., & Carraway, K. (1997).
Infants of intrusive and withdrawn mothers. Infant Behavior and Development,
20, 175-186.
Jones, N.A., Field, T., Davalos M., & Pickens, J. (1997). EEG stability in
infants/children of depressed mothers. Child Psychiatry & Human Development,
28, 59-70
Kaplan, P.S., Fox, K.B., & Huckeby, E.R. (1992). Faces as reinforcers: Effects of
pairing condition and facial expression. Developmental Psychobiology, 25, 299-
312.
55


Kaplan, P.S., Jung, P.C., Ryther, J.S., & Zarleng-Strouse, P. (1996). Infant-directed
versus adult-directed speech as signals for faces. Developmental Psychology, 32,
880-891.
Kaplan, P.S., Bachorowski, J.A., Smoski, M J., & Zinser, M. (2001). Role of clinical
diagnosis and medication use in effects of maternal depression on infant-directed
speech. Infancy, 2, 537-548.
Kaplan, P.S., Bachorowski, J., Smoski, M.J., & Hudenko, W.J. (2002). Infants of
depressed mothers, although competent learning, fail to learn in response to their
own mothers infant-directed speech. Psychological Science, 13(3), 268-271.
Kaplan, P.S., Dungan, J.K., & Zinser, M. (2004). Infants of chronically depressed
mothers learn in response to male, but not female, infant-directed speech.
Developmental Psychology, 40(2), 140-148.
McLaughlin, B., White, D., McDevitt, T., & Raskin, R. (1983). Mothers and fathers
speech to their young children: Similar or different? Journal of Child Language,
10, 245-252.
Mesulam, M.M. (1998). From sensation to cognition. Brain, 121,1013-1052.
Montague, D.P.F., & Walker-Andrews, A.S. (2002). Mothers, fathers, and infants:
The role of person familiarity and parental involvement in infants perception of
emotion expression. Child Development, 73, 1339-1352.
56


Murray, A.D., & Yingling, J.L. (2000). Competence in language at 24 months:
Relations with attachment security and home stimulation. Journal of Genetic
Psychology, 161, 133-140.
Murray, L. (1992). The impact of postnatal depression on infant development.
Journal of Child Psychology and Psychiatry and Allied Disciplines, 33, 543-561.
Murray, L., Fiori-Cowley, A., Hooper, R., and Cooper, P. (1996). The impact of
postnatal depression and associated adversity on early mother-infant interactions
and later infant outcome. Child Development, 67, 2512-2526.
NICHD Early Child Care Research Network. (1999). Chronicity of maternal
depressive symptoms, maternal sensitivity, and child functioning at 36 months.
Developmental Psychology, 35, 1297-1310.
Schaffer, C.E., Davidson, R.J., & Saron, C. (1983). Frontal and parietal
electroencephalogram asymmetry in depressed and non-depressed subjects.
Biological Psychiatry 18, 753-762.
Schore, A.N. (2001a). Effects of a secure attachment relationship on right brain
development, affect regulation, and infant mental health. Infant Mental Health
Journal, 22, 7-66.
Schore, A.N. (2001b). The effects of early relational trauma on right brain
development, affect regulation, and infant mental health. Infant Mental Health
Journal, 22, 201-269.
57


Sharp, D., Hay, D.F., Pawlby, S., Schmucker, G., Allen, H., and Kumar, R. (1995).
The impact of postnatal depression on boys intellectual development. Journal of
Child Psychology and Psychiatry and Allied Disciplines, 36, 1315-1336.
Sohr-Preson, S.L., and Scaramella, L.V. (2006). Implications of Timing of Maternal
Depressive Symptoms for Early Cognitive and Language Development. Clinical
Child and Family Psychology Review, 9, 65-83.
Spieker, S.J., Nelson, D.C., Petras, A., Jolley S.N., & Barnard, K.E. (2003). Joint
influence of child care and infant attachment security for cognitive and language
outcomes of low-income toddlers. Infant Behavior & Development, 26, 326-344.
Shannon, J., Tamis-LeMonda, C.S., London, K., & Cabrera, N. (2002). Beyond
rough and tumble: Low-income fathers interactions and childrens cognitive
development at 24 months. Parenting: Science and Practice, 2, 77-104.
Teti, D. M., & Ablard, K. E. (1989). Security of attachment and infant-sibling
relationships: A laboratory study. Child Development, 60, 1519-1528.
Teti, D.M., Gelfland, D.M., Messinger, D.S., & Isabella, R. (1995). Maternal
depression and the quality of early attachment: An examination of infants,
preschoolers, and their mothers. Developmental Psychology, 31, 364-376.
Thiessen, E.D., Hill, E.A., & Saffran, J.R. (2005). Infant-directed speech facilitates
word segmentation. Infancy, 7, 53-71.
Trevarthen, C. (1986). Form, significance and psychological potential of hand
gestures of infants. In Nespoulos, J. L., Perron, P., Lecours, A. R. (Eds.), The
58


biological foundation of gestures: Motor and semiotic aspects. Hillsdale:
Erlbaum.
Tronick, E.Z., & Weinberg, M.K. (1997). Depressed mothers and infants: Failure to
form dyadic states of consciousness. In L. Murray & P. Cooper (Eds.),
Postpartum depression and child development. New York: Guilford Press.
Van Ijzendoom, M.H., Dijkstra, J., & Bus, A.G. (1995). Attachment, intelligence, and
language: A meta-analysis. Social Development, 4, 115-128.
Vaughn, B.E., & Waters, E. (1990) Attachment behavior at home and in the
laboratory: Q-Sort observations and strange situation classifications of one-year-
olds. Child Development, 61, 1965-1973.
Waters, E. (1987). Attachment Q-set (Version 3). Retrieved (1/25/08) from
http://www.iohnbowlbv.com.
Whiffen, V.E., & Gotlib, I.H. (1989). Infants of postpartum depressed mothers:
Temperament and cognitive status. Journal of Abnormal Psychology, 98, 274-
279.
59