Correlates of Mother–Premature Infant Interactions Diane Holditch-Davis, Todd Schwartz,

Research in Nursing & Health, 2007, 30, 333–346
Correlates of Mother–Premature
Infant Interactions
Diane Holditch-Davis,1{ Todd Schwartz,2z Beth Black,2§ Mark Scher3k
1
School of Nursing, Duke University, Durham, NC
School of Nursing, University of North Carolina at Chapel Hill, Chapel Hill, NC
3
Case Western Reserve University School of Medicine, Cleveland, OH
Accepted 25 November 2006
2
Abstract: This study’s purpose was to examine whether child characteristics, child illness severity, maternal characteristics, maternal psychological well-being, and paternal support influenced interactions between
108 premature infants and their mothers. Mothers with singletons or more
infant illness stress showed more positive involvement. Mothers with less
infant illness stress, less education, or less participation in caregiving by
fathers showed more negative control. First-time mothers and mothers of
singletons provided more developmental stimulation. Children of younger
and White mothers showed more social behaviors. Less maternal education
and shorter period of mechanical ventilation were associated with greater
developmental maturity. Greater maternal worry was related to more child
irritability. These findings are consistent with the developmental science view
that the mother–premature relationship is a complex, reciprocal process.
ß 2007 Wiley Periodicals, Inc. Res Nurs Health 30:333–346, 2007
Keywords: premature infants; mother–infant interactions; parenting; infant
behavior
Interactions between prematurely born children
and their mothers are of interest to nurses because
these interactions are known to affect child
developmental outcome (Forcada-Guex, Pierrehumbert, Borghini, Moessinger, & MullerNix, 2006; Smith, Landry, & Swank, 2006). More
maternal positive involvement, lower use of
negative control strategies, and more developmental stimulation (talking, teaching) are related
to better outcomes for premature infants (Berlin,
Brooks-Gunn, Spiker, & Zaslow, 1995; ForcadaGuex et al.; Olsen, Bates, & Kaskie, 1992;
Poehlmann & Fiese, 2001). However, interactions
of mothers with prematures are less mutually
satisfying and developmentally appropriate than
those with fullterms (Muller-Nix et al., 2004;
Schmucker et al., 2005). Mothers of premature
infants work harder to initiate and maintain
Contract grant sponsor: National Institute for Nursing Research, National
Institutes of Health; Contract grant number: NR01894.
Correspondence to Diane Holditch-Davis, Duke University School of Nursing, Trent
Drive, DUMC 3322 Durham, NC 27710.
{
Professor.
z
Research Assistant Professor.
§
Assistant Professor.
k
Professor and Director of Division of Pediatric Neurology.
We thank Sola Park, Lisa Moorehead, Donna Harris, Mary Barkey, Leslie Miller,
Paula Anderson, Jason Dickenson, William Wooten, Chithra Jeyaram, Tzu-Ying Lee,
Laura Pence, Samia Shreitah, David Sterka, Daisy Wilson, Tanya Kewson, and Mark
Johnson for technical assistance.
Published online in Wiley InterScience (www.interscience.wiley.com)
DOI: 10.1002/nur.20190
ß 2007 Wiley Periodicals, Inc.
334
RESEARCH IN NURSING & HEALTH
interactions but receive fewer positive responses
from their infants than mothers of fullterms
(Singer et al., 2003).
Many factors—both related and unrelated to
prematurity—influence these interaction difficulties. According to the developmental science
perspective (Cairns, Elder, & Costello, 1996;
Thoman, Acebo, & Becker, 1983), children
develop in a continuous, reciprocal interaction
with the environment and are both influenced by
and influence it. The child and environment form a
complex system, made up of elements that are also
systems, such as the mother and child (Thoman
et al.). Interactions between mothers and children
are affected by factors within the child, such as
child characteristics and illness severity; factors
within the mother, such as maternal characteristics and psychological well-being; and factors
in the larger family environment, such as paternal
support for the mother. Understanding how these
factors relate to mother–premature interactions
may help nurses plan interventions to improve
interactions. The literature is inconsistent on the
effects of these factors, and few investigators
have studied them from a systems perspective. The
purpose of this study, therefore, was to examine
the degree to which selected child characteristics
and illness severity, maternal characteristics,
and psychological well-being, and a family
characteristic—paternal support—are related to
interactions between mothers and their premature
infants.
A child characteristic that has been reported to
affect mother–premature infant interactions is
sex. From 3 to 36 months corrected age, mothers
have been reported alternatively to be less
responsive to high-risk infant girls than to boys
(Engelke & Engelke, 1992) or to respond more
verbally and express more positive affect to highrisk premature girls than boys (Cho, HolditchDavis, & Belyea, 2004; Feldman & Eidelman,
2004). Three-year-old prematurely born girls
looked at their mothers more and expressed more
positive affect than did boys (Cho et al.). Feldman
and Eidelman found that mothers provided more
proprioceptive touch and less functional touch to
boys than to girls from 3 to 24 months, but Weiss,
Wilson, Hertenstein, and Campos (2000) found no
differences in touching based on sex at 3 months.
All of these studies included ethnically and socioeconomically diverse mothers; most infants were
premature and had high-risk medical courses,
although Feldman and Eidelman and Weiss et al.
studied fullterms and healthier prematures.
Multiple birth also affects interactions between
mothers and premature infants, and a greater
Research in Nursing & Health DOI 10.1002/nur
percent of prematures than fullterms are multiple
births (Blondel & Kaminski, 2002). Ethnically
and socio-economically diverse mothers of multiples had less time for play and were less sensitive
than mothers of singletons from birth to 24 months
(Feldman & Eidelman, 2004; Feldman, Eidelman,
& Rotenberg, 2004; Ostfeld, Smith, Hiatt, &
Hegyi, 2000). In the first month after term,
multiple birth infants were left alone more
and looked at, talked to, and held less often by
middle-class mothers (Holditch-Davis, Roberts,
& Sandelowski, 1999; Ostfeld et al.). Triplets
looked at and vocalized less often to their mothers
than twins or singletons from birth to 24 months
(Feldman & Eidelman). Caring for multiple birth
infants is more difficult and stressful for mothers
than caring for a single infant (Feldman et al).
Despite this, in one study, interactions between
socio-economically diverse mothers and their
very-low-birthweight multiple birth infants did
not differ from those of similar mothers and
singletons from 1 to 9 months corrected age
(Goldberg, Perrotta, Minde, & Corter, 1986). Sick
premature infants were included in few of these
studies; thus, more research is needed on the
effects of high-risk multiple birth infants on
mother–child interactions.
Infant illness severity also affects the development of positive mother–infant interactions. Lowincome mothers of 12–24-month-old prematurely
born toddlers with larger birthweights interacted
more positively with their children than did
mothers of toddlers with smaller birthweights
(Feingold, 1994), and middle-class mothers
interacted more positively with multiple birth
and premature infants with fewer medical complications than with infants with more complications between birth and 24 months (Feldman
et al., 2004; Muller-Nix et al., 2004). At 4 and
8 months corrected age, socio-economically
diverse mothers of premature infants with bronchopulmonary dysplasia (BPD) were less sensitive
to infant cues and distress than similar mothers of
fullterms or prematures without BPD (Jarvis,
Myers, & Creasey, 1989), and infants with BPD
were less responsive to maternal interactive
behaviors at 8 and 12 months than low-risk infants
(Singer et al., 2003). At 6 and 12 months corrected
age, premature infants with severe neonatal
medical complications, including BPD and
intraventricular hemorrhage, were at greater risk
for interactive difficulties than infants with
milder illnesses (Landry, Smith, Miller-Loncar,
& Swank, 1997). However, other investigators
found that between 6 months and 4 years, mothers
of sicker children showed more social behaviors to
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
compensate for their infant’s immaturity and
lack of responsiveness (Holditch-Davis, Cox,
Miles, & Belyea, 2003; McGrath, Sullivan, &
Seifer, 1998).
Maternal characteristics such as ethnicity,
education, age, and parity also may affect
mother–premature infant interactions. From 1 to
3 years corrected age, African-American mothers
of prematures showed less warmth, talked less,
touched less, provided fewer learning activities,
and use more negative control strategies than
White mothers even when SES was controlled
(Brooks-Gunn, Klebanov, & Duncan, 1996; Cho
et al., 2004; Tesh & Holditch-Davis, 1997). From
birth to 36 months, ethnically diverse mothers
with more education made less use of negative
control and provided more stimulation for learning and more nurturing touch, looking, talking,
and interactions than mothers with lower education levels (Feeley, Gottlieb, & Zelkowitz, 2005;
Gordon, Chase-Lansdale, & Brooks-Gunn, 2004;
Tesh & Holditch-Davis; Weiss et al., 2000). Older
and first-time mothers of prematures provided
more nurturing touch at 3 months corrected age
(Weiss et al.) and were more emotionally responsive at 6–36 months than younger or multiparous
mothers (Engelke & Engelke, 1992). However,
Gerner (1999) found that parity had no
affect on mother–premature infant interactions
at 3 and 6 months even though it did affect
interactions with fullterms. Sick premature infants
were included in these studies; thus, maternal
characteristics of ethnicity, education, age, and
parity affected interactions even with high-risk
prematures.
Maternal psychological well-being also may
affect mother–premature infant interactions. The
experience of delivering a low-birthweight infant
and the possibility of infant death place mothers at
increased risk for psychological distress during
the neonatal period (Singer, Davillier, Bruening,
Hawkins, & Yamashita, 1996). Eight months after
term, middle class, White mothers of prematures
reported more parenting stress than mothers of
fullterms, even when infant behaviors did not
differ (Moran & Pederson, 1998). Mothers with
better overall mental health had more responsive
premature infants at 3 months corrected age
(Schmucker et al., 2005; Weiss & Chen, 2002).
Depressive symptoms in mothers of high-risk
premature infants had a negative effect on
maternal feeding competency and caregiving
evaluations over the first year of life (Pridham
et al., 2005; Pridham, Lin, & Brown, 2001), and
depressed mothers were less responsive to their
prematures, showed less positive interactive
Research in Nursing & Health DOI 10.1002/nur
335
behaviors, provided less cognitive growth promoting stimulation, and reported less attachment to
them than non-depressed mothers (Feldman,
Weldman, Leckman, Kuint, & Eidelman, 1999;
Singer et al., 2003). With the exception of Moran
and Pederson, these studies included ethnically
and socio-economically diverse mothers and
premature infants with high-risk medical courses.
Finally, paternal support for the mother probably affects mother–premature infant interactions. Fathers of newborns have a uniquely
important role in supporting mothers (Mercer,
2004). Although ethnically and socio-economically diverse mothers of prematures reported that
fathers were their major source of support (Miles,
Carlson, & Funk, 1996) and, for most of them,
their closest relationship (Coffman, Levitt, &
Guacci-Franco, 1993), how this factor affects
maternal interactions with premature infants has
been studied only rarely. Mothers of disabled
children described fathers as being supportive
when they provided financial support and helped
with child care (Traustadottir, 1991). Fathers of
premature infants provided more infant caregiving
than fathers of fullterms at 1 month after discharge
(Brown, Rustia, & Schappert, 1991) and had
more positive interactions with their infants at
3 months corrected age (Harrison, 1990). AfricanAmerican, but not White, fathers who were more
involved with their prematures had children with
better developmental outcomes (Yogman, Kindlon, & Earls, 1995). How these effects relate to the
interactions of mothers with their premature
infants needs to be explored. Over the infant’s
first year, paternal support was related to better
mental health in mothers of prematures (Weiss &
Chen, 2002), and unmarried mothers of medically
fragile infants did not differ from married mothers
in reporting the helpfulness of paternal support
although they were less satisfied with it (Lee,
Miles, & Holditch-Davis, 2006).
Although factors that affect mother–premature
infant interactions have been examined in previous studies, the direction of the effects, in many
cases, have been contradictory. Researchers rarely
have examined more than one or two maternal and
infant factors. One exception is Feeley et al.
(2005), who examined variables in all of these
factors except child characteristics. However, they
only used a global measure of maternal interactive
behaviors, the total score from the Nursing-Child
Assessment Teaching Scale (Barnard, 1983).
Multiple maternal and infant factors have not
been studied along with specific maternal and
child interactive dimensions, as required by the
developmental science perspective. Therefore,
336
RESEARCH IN NURSING & HEALTH
the aim of the current study was to examine the
degree to which selected child characteristics,
child illness severity, selected maternal characteristics, maternal psychological well-being, and
paternal support were related to interactions
between mothers and their premature infants at
6 and 18 months corrected for prematurity. Child
characteristics examined were sex and singleton
versus multiple birth. Child illness severity
variables were birthweight, days of mechanical
ventilation, and number of rehospitalizations.
Maternal characteristics were ethnicity, education, age, and parity. Maternal psychological wellbeing was measured as perceived stress in the
NICU due to unit and baby factors, perceived
stress in the NICU due to parental role alteration,
worry about child health, and depressive
symptoms. Paternal support was determined by
maternal marital status and maternal report of
the amount of father participation in child
caregiving.
METHODS
Participants
Participants were 108 infants and their mothers
who were enrolled in a larger study of biological
and social risks of prematurity conducted from
1997 to 2003 (Holditch-Davis, Scher, Schwartz, &
Hudson-Barr, 2004). Infants were required to be
less than 35 weeks gestational age and either
have a birthweight less than 1,500 g or require
mechanical ventilation or continuous positive
airway pressure. Seventy-three (68%) had both
problems. Infants with congenital problems
affecting development (such as Down Syndrome)
were excluded, but infants with postnatal neurological insults were eligible. Infants were
recruited from two tertiary hospitals in the United
States: 58 from a southeastern perinatal center and
50 from a midwestern children’s hospital.
All mother–child dyads with interaction data at
6 or 18 months were included in this report. These
contacts occurred at times of different levels of
stress in the mother–child relationship. Six
months is a relatively stable time, when the
transition home is complete and initial interactive
differences between fullterms and prematures
have decreased (Crawford, 1982; Wille, 1991).
At 18 months, toddlers’ developing motor and
language skills present new challenges in mother–
child interactions. The infants had a mean
gestational age of 28.9 weeks (SD, 2.7; range
Research in Nursing & Health DOI 10.1002/nur
23–34) and a mean birthweight of 1,232 (SD,
430). All infants, except two, one of whom was
large for gestational age, were under 2,500 g at
birth. They were 52% male, 69% singletons, and
31% multiples (25% twins, 6% triplets). They had
a mean of 10.3 days of mechanical ventilation (SD,
17.6) with 87 infants requiring ventilation; 17%
had a Grade I or II intraventricular hemorrhage
(IVH), and 5% had a Grade III or IV IVH. Their
mothers averaged 28.6 years of age (SD, 6.6) and
13.7 years of education (SD, 2.3). Fifty-nine
percent were married; 43% were AfricanAmerican, 56% White, and 1% Asian. No mother
was married to or lived with a partner other than
the child’s father during the study. About 56%
were first-time mothers. Four delivered another
baby between the 6- and 18-month contacts. The
only differences between infants and mothers
from the two hospitals were that infants from
the southeastern hospital were more likely to be
multiple births (41.4% vs. 18.4%) and first births
(67.2% vs. 42.0%).
Calculating a power analysis for the mixed
model is not feasible. An approximation based on
a regression model showed that with a sample size
of 108, 7 independent variables in the model, and
an R2 of at least 0.125 (or 3 independent variables
and an R2 of at least .095), the power to reject
R2 ¼ 0 would exceed 0.80 at the .05 significance
level (Gatsonis & Sampson, 1989).
Variables
Mother–child interactions. Videotapes of
mother–infant interactions in the home were
made at 6 and 18 months corrected age. They
were scored several months later with the mother–
child coding system used in previous studies
(Holditch-Davis, Cox, et al., 2003; HolditchDavis, Miles, et al., 2001). The occurrence of
child and mother behaviors were scored in each
10-second interval with each behavior only
counted once in each interval. For scoring, the
codes were divided into five subsets of behaviors,
each of which was scored by two or three coders.
The first author trained the coders until they
achieved acceptable inter-rater reliability (kappas
greater than .70) on their subset of behaviors.
Ongoing inter-rater reliability was checked every
couple of months by having two coders score the
same subset of behaviors on the same tape.
Cohen’s kappas for the maternal behaviors used
in this report ranged from .76 to .91 with a mean of
.82; for child behaviors, kappas ranged from .84 to
.96 with a mean of .90.
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
337
Table 1. Interactive Behaviors Scored During the Videotapes
Mother Behaviors
Interact
Proximity
Uninvolved
Play with
Child
Positive
Negative
Child Behaviors
—
—
—
—
Positive
Negative
Touch
Talk
Teach
Talk
—
Together
—
—
—
—
—
Vocalize
Object play
Locomotion
Walk
—
—
—
Look
Gesture
Fuss
—
Definitions
Talks to, touches, gestures toward, or plays with child
Close enough to achieve eye contact with the child
Not interacting with, playing with, or looking at the child
Initiates or takes part in child’s play activity or games
Directs positive affect to mother or child (e.g., smile or kiss)
Directs negative affect towards mother or child (e.g., frown
or scold)
Touches child in a non-negative way (e.g., pat or caress)
Speaks words to mother or child
Instructs child (e.g., names an object, demonstrates an
activity)
Interacts with the child but is not caregiving.
Makes a non-fussy sound but does not say words
Plays with an object but not with a person
Moves body but without walking (e.g., belly creep or scoot)
Walks independently or by supporting self on objects or
people
Focuses eyes on face of mother
Makes a gesture, such as showing a toy or shaking head
Makes a negative sound (e.g., crying or whining)
These behaviors were combined into maternal
and child variables, as described in Table 1. To
adjust for variation in the lengths of videotapes,
most variables were measured as a percentage of
the scoreable videotape. Play with objects was
measured as a percentage of the time that the child
was not playing with the mother so that it
measured the child’s ability to play independently.
HOME. The HOME inventory (0–3 version)
was designed to identify children under 3 years
who are at risk for developmental delay due to a
home environment lacking appropriate stimulation (Caldwell & Bradley, 1984). The HOME
consists of 45 items arranged in 6 subscales. Each
item is scored as present or absent, and the score
equals the number of present items. The HOME is
administered using both maternal interview and
observation. The first author trained research
assistants at both sites until they achieved 90%
inter-rater reliability. Test-retest reliability for the
total scale over 6-month intervals was .76, and
internal consistency was .89 (Holditch-Davis,
Tesh, Goldman, Miles, & D’Auria, 2000). Two
HOME subscales were used in this report—
subscale II, Acceptance of the Child’s Behavior
and subscale V, Maternal Involvement. Two items
from subscale II that were not directly related to
maternal negative control—the presence of pets
and more than 10 books in the house—were
omitted. Similar to other studies (Bradley, Rock,
Caldwell, & Brisby, 1989; Caldwell & Bradley,
1984; Holditch-Davis, Tesh, et al.), internal
Research in Nursing & Health DOI 10.1002/nur
consistency of the modified subscale II for this
sample combining 6 and 18 months was .73;
subscale V was .60.
Maternal-child interactive dimensions. Ten
maternal behaviors from the observation and
two HOME subscales were grouped into three
interactive dimensions: positive involvement,
developmental stimulation, and negative control.
Maternal dimensions were derived theoretically to
capture the major types of maternal interactive
behaviors shown to affect child development and
to be affected by maternal depression (Berlin et al.,
1995; Field, 1995; Poehlmann & Fiese, 2001;
Singer et al., 2003; Smith et al., 2006). Ten child
behaviors were grouped into three child interactive dimensions: child social, developmental
maturity, and irritability. These child dimensions
were derived theoretically from the major types of
child behaviors shown to be related to child
developmental outcomes and affected by maternal
depression (Field, 1995; Holditch-Davis, Bartlett,
& Belyea, 2000; Holditch-Davis, Docherty, Miles,
& Burchinal, 2001). Dimension scores were
calculated by converting each variable to Z-scores
(calculated from the data from both 6 and
18 months) and then averaging the Z-scores for
each dimension’s variables. Because the mean and
SD were based on both ages, these Z-scores
preserved any existing differences between the
two ages. These dimensions were developed using
data from two different samples: 3-year-old
prematurely born children and 12–24-month-old
338
RESEARCH IN NURSING & HEALTH
toddlers of HIV-infected mothers. Internal consistency of these dimensions ranged from .57 to
.98 in both samples, and correlations over ages
were significant (Holditch-Davis, 2002). Table 2
gives each dimension’s component variables,
alphas for internal consistency, means, and SDs.
Child characteristics and illness severity. Sex,
multiple birth, birthweight, and days of mechanical ventilation were obtained from the neonatal
medical record. Because the length of mechanical
ventilation was skewed (M 10.3; SD, 17.6), infants
with a day or less of ventilation were scored as
1 day, and each subject’s score was transformed
by using its logarithm. The number of rehospitalizations was obtained from a child health history
completed by the mother at 2, 6, 9, 13, and
18 months. The number of rehospitalizations
between neonatal hospital discharge and 6 months
and between 6 and 18 months were used in
analyses.
Maternal characteristics. Maternal characteristics of ethnicity, education, and age were
recorded on the Demographic Information
Questionnaire that mothers completed at enrollment and at 2, 6, 9, 13, and 18 months. Ethnicity
was scored as White or minority. One minority
mother was Asian; the rest were African-American. Parity was obtained from the infant’s medical
record.
Maternal psychological well-being. Maternal
psychological well-being was measured with the
Parental Stressor Scale: NICU, the Worry Index,
and the Center for Epidemiologic Studies Depression Scale. The mother’s perception of the stress
Table 2. Internal Consistency, Means, and SDs of the Mother–Child Interactive Dimensions for 99 Mother–
Premature Infant Dyads at 6 Months and 92 Dyads at 18 Months
6 Months
18 Months
M (SD)
M (SD)
Interaction
Proximity
Uninvolved with childc
Play with child
Mother positive
Mother touch
HOME subscale V
Total
Mother negative
HOME subscale IIc
Total
Mother talk
71.63 (19.26)
28.38 (20.98)
20.09 (17.94)
15.70 (11.19)
10.08 (8.82)
23.56 (13.24)
5.43 (0.85)
0.06 (0.58)
0.23 (0.96)
5.77 (0.92)
0.23 (0.73)
40.94 (18.61)
57.38 (21.90)
84.47 (14.12)
34.21 (21.72)
14.08 (12.41)
8.56 (7.27)
9.28 (6.37)
5.55 (0.78)
0.07 (0.60)
0.73 (1.28)
5.20 (1.15)
0.24 (0.87)
48.29 (22.59)
Teach
Total
Child positive
Child look
Child gesture
Total
Vocalize and talk
1.81 (3.50)
0.33 (0.53)
6.25 (4.62)
22.34 (10.70)
13.57 (7.85)
0.28 (0.65)
25.21 (12.72)
12.31 (12.62)
0.35 (1.04)
7.91 (5.24)
29.72 (12.98)
21.01 (8.72)
0.30 (0.83)
36.70 (16.26)
Talk as % of vocalize þ talk
Play with objects
Locomote and walk
Walk as % of locomote þ walk
Total
Child negative
Child fuss as % of together
Total
0.00 (0.00)
27.23 (14.86)
3.99 (6.58)
8.73 (24.71)
0.63 (0.27)
6.76 (6.90)
7.75 (9.37)
0.19 (1.09)
30.07 (21.25)
48.28 (18.70)
37.66 (15.88)
84.20 (24.75)
0.68 (0.50)
4.49 (4.77)
4.38 (6.62)
0.20 (0.74)
Dimensiona
Positive involvement, alpha ¼ .70
Negative control, alpha ¼ .56
Developmental stimulation,
alpha ¼ .71
Child Social, alpha ¼ .71
Child developmental maturity,
alpha ¼ .82
Child irritability, alpha ¼ .91
Variablesb
a
Dimensions totals are means of Z-scores for the component variables.
Videotape variables are percentages of the total videotape time, except for variables with their divisors in their titles
and Play with Objects, which was measured as a percentage of the time that the child was not playing with the mother.
c
These variables were reverse scored when combined with the other variables in the dimension.
b
Research in Nursing & Health DOI 10.1002/nur
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
due to parenting her infant during hospitalization
was measured with the Parental Stressor Scale:
NICU (PSS:NICU; Miles, Funk, & Carlson,
1993). This tool has two subscales: the unit
environment and infant illness (19 items) and
parental role alterations (7 items). Mothers rate the
items on a 5-point scale ranging from ‘‘not at all
stressful’’ to ‘‘extremely stressful.’’ At enrollment
and the 2-month contact, mothers were asked to
retrospectively rate their experiences during
hospitalization. PSS:NICU scores were correlated
with anxiety scores and have good test-retest
reliability (Miles et al., 1993; Shields-Poe &
Pinelli, 1997). Internal consistency for the current
study at enrollment and 2 months was .92 and .94
for the NICU environment and infant illness
subscale and .93 and .92 for the parental alterations subscale. Because scores at enrollment and
2 months were correlated at .69 for the unit and
baby subscale and .62 for the parental alterations
subscale, the mean of item scores for each
subscale over the two administrations was used
in analyses. The value for the single administration was used for 14 mothers with only the
enrollment data and 10 with only the 2-month
data.
The Worry Index (Miles & Holditch-Davis,
1995) was administered at all data collection
points. On the Worry Index, mothers rate how
much they worry about their child’s medical
problems, normality, length of illness and possible
death, re-hospitalization, caring for the child
at home, the infant getting enough to eat, and
sleeping enough. The mother rates the seven items
using a 5-point scale from ‘‘not at all’’ to ‘‘very
much.’’ The mean of the item scores at 6 and
18 months were used in analyses. Worry scores of
mothers of premature and medically fragile
infants were related to depressive and posttraumatic stress symptoms and to maternal
personal growth (Holditch-Davis, Bartlett, Blickman, & Miles, 2003; Miles, Holditch-Davis,
Scher, & Schwartz, in press). Internal consistency ranged from .71–.90 in previous studies
(Holditch-Davis, Bartlett, et al., 2003; Miles &
Holditch-Davis) and was .86 at 6 months and .78 at
18 months in the current study.
The Center for Epidemiologic Studies Depression Scale (CESD), used to measure depressive
symptoms, focuses on depressed mood, particularly guilt, worthlessness, helplessness, and hopelessness (Radloff, 1977). The frequency of
occurrence for each of 20 items is rated on a
4-point scale indicating how often in the last week
they were experienced, ranging from rarely or
none of the time (less than 1 day) to most or all of
Research in Nursing & Health DOI 10.1002/nur
339
the time (5–7 days). Scores on the CESD range
from 0 to 60; higher scores indicate more
depressive symptoms. The CESD is correlated
with other measures of depression (Radloff;
Weissman, Sholomskas, Pottenger, Prusoff, &
Locke, 1977) and with other measures of psychological well-being in mothers of premature
infants, including hospital environmental stress
and worry about the child’s health (Miles et al., in
press). Internal consistency for the current study
ranged from .81 at 6 months to .86 at enrollment.
CESD scores at enrollment, 6 months, and
18 months were used in analyses.
Paternal support. Paternal support was measured using maternal marital status and the
number of child caregiving activities performed
by the child’s biological father. Marital status was
obtained from the Demographic Information Questionnaire. At 6 and 18 months, the
mother completed a questionnaire asking which of
18 caregiving activities were performed by the
infant’s father, whether or not the mother was
married to him and whether or not he lived in the
child’s household. Internal consistency (KR-20)
was .96 at 6 months and .97 at 18 months.
Procedures
The committees for protection of human subjects
for the two study sites approved the study. Infants
were enrolled when they were no longer critically
ill (not receiving mechanically ventilation or
with medical conditions that were immediately
life-threatening) as long as an additional hospital
stay of at least 1 week was anticipated and the
mother provided informed consent. After enrollment, mothers filled out the demographic and
psychological well-being questionnaires.
Follow-up at home was conducted at 6 and
18 months corrected for prematurity. The 6-month
videotapings were scheduled at a time when the
infant was awake and due for a feeding. Given
the increasing independence of older infants, the
18-month observation was scheduled at a time
when the child was awake and not eating and when
the mother was at home. The mother was told that
the purpose of the videotape was to record the
child’s spontaneous behaviors and play in the
home. The mother was asked to do what she would
typically do if at home with the child and to act
as though the videotaper were not present.
The videotaper remained with the infant even if
the mother left the room. At the conclusion of a
family’s involvement with the study, the mother
was given a copy of the infant’s and her previous
340
RESEARCH IN NURSING & HEALTH
videotapes. Although we planned to visit all
families at 6 and 18 months, 9 families missed
the 6-month contact, and 16 missed the 18-month
contact.
The HOME Inventory was administered after
the videotaped interaction, and the scorer used
information about the social environment gained
during the videotaping as well as during the rest
of the visit. Scoring the HOME took about
15 minutes. Although the HOME was partially
scored using the same situation as the mother–
child videotapes, the two measures were not
redundant, as the interaction codes could be used
to score only 8 of 45 items on the HOME. The
mothers also completed child health history
questionnaires at 2, 6, 9, 13, and 18 months.
Mothers were paid $10 each time they completed
questionnaires. The child was given a small gift at
each home visit.
Data Analyses
Alpha was set at .05 for two-tailed tests. The
number of covariates in the model was based on
the general rule of thumb of having at least 5–10
observations per model term, which for 108
mother–infant dyads (and 191 observations)
allowed us to fit 10–20 explanatory variables
(Muller & Fetterman, 2002). Maternal and child
interactive dimensions at 6 and 18 months were
regressed over age using the general linear mixed
model (GLMM, hereafter referred to as mixed
model), a flexible statistical procedure for analyzing continuous longitudinal data that accommodates missing values and mistimed data (HolditchDavis, Edwards, & Helms, 1998). The mixed
model, a generalization of the standard linear
regression, allows analysis of data with several
sources of variation, rather than just one. Like the
standard model, it assesses the relationship
between continuous outcome measures and explanatory variables. Unlike a standard regression, R2
cannot be determined in an unambiguous manner
for the general linear mixed model because the
model has correlated errors and multiple variance
terms, not just a single one. Parameters of the
mixed model include population (fixed) effects
and individual (random) effects. Because the ages
of the children at the 6- and 18-month contacts
varied slightly, the actual age of the child in weeks
past term (40 weeks post-menstrual age) at the
time of 6- and 18-month contacts was used in
analyses. Age was centered so that the intercept
equaled the value at 26 weeks (6 months). Only the
intercept and age were included in the random
Research in Nursing & Health DOI 10.1002/nur
effects component, providing mother–child dyadspecific intercepts and slopes across age.
Covariates were child characteristics, child
illness severity, maternal characteristics, maternal
psychological well-being, paternal support, and
their interactions with child age. Four of these
variables were measured at both 6 and 18 months:
number of rehospitalizations, worry about child
health, depressive symptoms, and number of
caregiving activities performed by the father. For
the remaining covariates (sex, multiple birth,
birthweight, days of mechanical ventilation,
ethnicity, education, age, parity, NICU stress due
to unit and baby factors, NICU stress due to
parental role alteration), a single value was used
for each dyad. Diagnostics performed on the
model residuals to determine whether model
assumptions were satisfied indicated a departure
from the assumptions for all interactive dimensions except positive involvement. Therefore, all
dimensions except positive involvement were
transformed using the natural logarithm.
An elimination process was used to simplify the
model. First, each group of variables was examined separately. Variables with p < .15 were
included in an overall mixed model. Interactions
of each main effect with age were added. A groupwise test of the interactions was performed at
p < .15, and individual interactions tested only if
the group met this criterion. Remaining variables
were reduced until all variables in a final mixed
model were p < .05. This procedure simplified the
model and led to inferences that some effects were
either zero or not large enough to be detected.
RESULTS
Examining dyads with data at both 6 and
18 months, all maternal dimensions, but not child
dimensions, were significantly correlated over
age, ranging from r(83) ¼ .28 for negative control
to r(83) ¼ .41 for developmental stimulation. As
Table 3 shows, some dimensions were intercorrelated within age. Maternal dimensions of
positive involvement and developmental stimulation were correlated at both ages. Negative control
was unrelated to other maternal dimensions.
The child dimensions were unrelated except for
a positive relation between child social behavior
and irritability at 6 months and between child
social behavior and developmental maturity at
18 months. Child social behavior was positively
related to all maternal dimensions at both 6 and
18 months. Developmental maturity was related
to maternal negative control at 6 months. Child
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
341
Table 3. Correlations Among the Mother–Child Interactive Dimensions at 6 Months (n ¼ 99, Above the
Diagonal) and at 18 Months (n ¼ 92, Bolded and Below the Diagonal)
Dimension
1. Positive involvement
2. Negative control
3. Develop. stimulation
4. Child social
5. Develop. maturity
6. Irritability
1. Pos.
2. Neg.
.02
.05
.80***
.40***
.02
.12
.12
.22*
.12
.24*
3. Stim.
.53***
.16
.22*
.08
.14
4. Social
5. Maturity 6. Irritab.
.36***
.23*
.41***
.01
.35***
.19
.11
.21*
.19
.04
.05
.03
.27**
.10
.03
Note: Pos., positive; Develop., developmental; Neg., negative; Stim., stimulation; Irritab., irritability.
*p < .05.
**p < .01.
***p < .001.
irritability was related only to maternal negative
control at 18 months.
Correlates of Interactive Dimensions
Table 4 presents the final mixed models for each
interactive dimension. As shown by the direction
of the corrected age effect, maternal positive
involvement decreased with age; maternal developmental stimulation, negative control, child
social behavior, and developmental maturity
increased. Maternal positive involvement was
higher for mothers of singletons and mothers with
more stress due to the NICU environment and the
infant’s illness. Less maternal stress due to the
NICU environment and infant illness, less maternal education, and less paternal caregiving
participation were related to more maternal
negative control. Mothers of singletons and firsttime mothers provided more developmental
stimulation than other mothers. Children of
younger and White mothers showed more social
behaviors. Shorter mechanical ventilation and less
maternal education and were related to greater
child developmental maturity. More maternal
worry about child health was related to more child
irritability. There were no significant interactions
with age.
Table 4. Relationship of Maternal and Child Interactive Dimensions at 6 and 18 Months to Child
Characteristics, Illness Severity, Maternal Characteristics, Maternal Psychological Well-Being, and
Paternal Support in the 108 Mother–Premature Infant Dyads
Covariates
Intercepta
Estimate (SE)
Corrected Age
Estimate (SE)
.14 (.16)
.003* (.001)
.41 (.32)
.01*** (.001)
Developmental stimul.
.42*** (.05)
.01*** (.001)
Child social behaviors
.82*** (.15)
.01*** (.001)
Developmental maturity
.58*** (.09)
.01*** (.001)
.76*** (.20)
.004 (.003)
Dimension
Positive involvement
Negative control
Child irritability
Variable
Multiple birth
NICU stress
NICU stress
Mat. education
Father careg.
Multiple birth
Parity
Mother age
Ethnicity
Mech. ventilat.
Mat. education
Worry
Estimate (SE)
.28** (.10)
.11* (.05)
.10* (.04)
.04* (.02)
.01* (.01)
.14* (.07)
.15* (.04)
.01* (.005)
.18* (.06)
.06* (.02)
.02** (.01)
.24** (.08)
Note: Stimul., stimulation; Mat. education, maternal education; Father Careg., number of paternal caregiving
activities; NICU Stress, stress due to baby’s illness and the NICU environment; Mech. Ventilat., log of days of
mechanical ventilation
a
The intercept is tested for significant difference from zero.
*p < .05.
**p < .01.
***p < .001.
Research in Nursing & Health DOI 10.1002/nur
342
RESEARCH IN NURSING & HEALTH
Correlates of Negative
Control Variables
Because the internal consistency of the maternal
negative control dimension was relatively low,
separate predictive models were calculated for its
two component variables. Predictors of the
modified HOME subscale II were ethnicity
(F[1,103] ¼ 6.87, p < .05), maternal education
(F[1,103] ¼ 6.17, p < .05), and maternal age
(F[1,103] ¼ 6.01, p < .05). Greater mother negative (from the videotape) was related to more
maternal stress due to the NICU environment
and infant illness, F(1,104) ¼ 8.55, p < .01. Thus,
predictors of the maternal negative control
dimension combined predictors of its two component variables but with father caregiving participation instead of maternal ethnicity and age.
DISCUSSION
The interactions of mothers and their prematurely
born children were related to child characteristics,
child illness severity, maternal characteristics,
maternal psychological well-being, and paternal
support. Different variables, and only a few of the
possible covariates, were related to each interactive dimension. Although other investigators
have examined the effects of most of these factors
on mother–premature interactions, this is the first
study in which the effects of representative
measures of all of these factors on specific
maternal and child interactive dimensions were
examined. Most of the dimensions were affected
by child age. Maternal positive involvement
decreased with age; maternal developmental
stimulation, and negative control and child social
behaviors increased. As expected, child developmental maturity also increased. No interactions
between age and child or maternal factors were
significant, indicating that similar factors are
related to mother–premature infant interactions
at both 6 and 18 months corrected age. However,
had we also studied these mothers and infants at an
earlier age when the NICU experiences were more
recent, we might have found more interactions
between age and the other covariates.
Both maternal and child variables were related
to three interactive dimensions—maternal positive involvement, maternal developmental stimulation, and child developmental maturity. The
lower amounts of positive involvement and
developmental stimulation of mothers with multiples probably were due to the time demands of
caring for two or three infants (Holditch-Davis
Research in Nursing & Health DOI 10.1002/nur
et al., 1999; Ostfeld et al., 2000). The finding that
first-time mothers provided more developmental
stimulation is similar to the findings of Weiss et al.
(2000) and Engelke and Engelke (1992) but
different from those of Gerner (1999). However,
these investigators did not examine maternal
behaviors that promote child development, but
rather examined global maternal responsivity or
touch. The finding that greater maternal stress due
to the NICU environment and infant illness was
related to more positive involvement contradicts
previous findings that mothers had more positive
interactions with healthier prematures (Feingold,
1994; Landry et al., 1997; Muller-Nix et al., 2004;
Singer et al., 2003). Our findings support the
suggestions of other investigators that mothers
provide more interactions to compensate for the
behaviors of sicker infants (Holditch-Davis, Cox,
et al., 2003; McGrath et al., 1998) but also suggest
that the mother’s perception of how sick her infant
was may be more important for mothering than the
actual severity of the infant’s illness.
Maternal education and length of mechanical
ventilation were related to child developmental
maturity. These findings are consistent with other
studies showing longer mechanical ventilation is
related to poorer child development (Vohr et al.,
2003). This effect may be due to medical
complications that are correlated with mechanical
ventilation, such as intraventricular hemorrhage,
rather than the direct effects of ventilation. On the
other hand, our findings that greater developmental maturity was related to lower maternal
education contradict previous findings that poorer
child development is related to lower maternal
educational attainment (Gordon et al., 2004; Zahr,
1999). Yet this finding is not universal: Zahr found
that higher maternal education was related to
better motor skills at 8 months but not better
cognitive skills and then only for AfricanAmerican prematures. These investigators used
standardized assessments of multiple behaviors
and assessed infant capabilities. Our measure
of developmental maturity only examined the
amounts of three behaviors—talking, walking,
and playing independently with toys—that children actually exhibited during interactions,
regardless of their capabilities. Thus, the differences in our findings from the literature probably
are due to methodological differences.
The other maternal and child dimensions were
related only to maternal variables. The findings for
maternal negative control were generally similar
to those of previous studies. Findings that mothers
with less education used more negative control are
consistent with previous findings that lower
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
socioeconomic status is related to greater maternal
use of negative control strategies (Gordon et al.,
2004; Tesh & Holditch-Davis, 1997). Likewise,
the finding that more maternal stress due to the
NICU environment and infant illness was related
to less negative control is consistent with our
findings on positive involvement (a dimension that
is roughly opposite to negative control) that
mothers who were more stressed provided
more positive interactions. As suggested by others
(Holditch-Davis, Cox, et al., 2003; McGrath et al.,
1998), this may be due to mothers providing more
positive interactions to compensate for the immature behaviors of sicker infants. Few researchers
have examined the effect of paternal caregiving
activities on mother–infant interactions. Our
finding that when fathers engaged in more
caregiving (whether or not they lived in the
household), mothers used less negative control
shows that paternal support can affect mothering.
Thus, there is a need for additional research on the
activities of fathers of premature infants and the
ways they have an impact on the maternal–child
relationship and child outcomes.
The child social dimension also was related
only to maternal variables and was correlated with
the maternal dimensions at both 6 and 18 months.
Children of younger and White mothers showed
more social behaviors. Although other investigators have found that White mothers of prematures
showed more warmth, used fewer negative
control strategies, and interacted more than
African-American mothers (Brooks-Gunn et al.,
1996; Cho et al., 2004; Tesh & Holditch-Davis,
1997), the current results are the first to show that
this parenting style may affect the social behaviors
of prematurely born children. These findings also
might be due to either our sample having very few
adolescent mothers, or the younger mothers
requiring more independence of 6–18-montholds. Previous researchers have found that
younger mothers had higher expectations for the
behaviors of their infants than older mothers
(Vukelich & Kliman, 1985).
Other investigators have not examined the
relationships between premature infant irritability
and either multiple birth or maternal worry about
child health. Mothers of multiple birth prematures
have been shown to be less likely to respond to
crying than mothers of singleton prematures
(Ostfeld et al., 2000). Thus, multiple birth infants
may be forced to cry longer to attract maternal
attention. Generalized maternal anxiety has been
related to decreased responsiveness of premature
infants but not to irritability (Feeley et al., 2005;
Schmucker et al., 2005). It is possible that either
Research in Nursing & Health DOI 10.1002/nur
343
greater maternal worry may lead premature
infants to become more irritable, or greater infant
irritability may lead mothers to worry more.
Papousek and von Hofacker (1998) found that
maternal anxiety was high when 1–6-month-old
fullterm infants were highly irritable, but in
another study of fullterm infants, maternal anxiety
and depression at 4 and 8 months did not predict
infant irritability at 12 months (Pauli-Pott, Mertesacker, & Beckman, 2004).
A number of variables, including child birthweight, number of rehospitalizations, and sex, and
maternal marital status, did not reach criteria for
inclusion in any final analysis, indicating that they
were not directly related to mother–premature
interactions. Stress due to parental role alteration
and depressive symptoms also had minimal effect
on interactions. This is surprising, as a number of
investigators have shown that maternal depression
was related to the quality of mother–premature
infant interactions (e.g., Feldman et al., 1999;
Singer et al., 2003). However, the effects of stress
and worry about child health have not been
examined along with depression in previous
studies. Investigators have found that maternal
depressive symptoms are increased by worry
about child health and the stress experienced in
the NICU (Miles et al., in press). Thus, these
variables, which we found affected mother–
premature infant interactions, might explain the
effects of maternal depression on mother–
premature infant interaction found in other
studies. Additional research is needed to clarify
this issue.
Several factors may limit the generalizability of
this study. We only examined mother–premature
infant interactions at two time points, 6 and
18 months after term. Studying interactive behaviors at earlier ages and more frequently might
result in different findings. Longitudinal studies
that examine the relationship between the maternal and child factors and mother–premature
interactions over longer time periods are needed.
Another limitation was the relatively low-internal
consistency of the negative control dimension,
probably due to the low rate at which mothers
displayed this behavior. Items on the HOME, for
example, include physical punishment occurring
more often than once a week, and rates of maternal
expression of negative affect typically make up
less than 2% of mother–child interactions
(Holditch-Davis, Bartlett, et al., 2000) as we
found in the current study. Despite its infrequency,
negative control has a major impact on children
and their developmental outcomes (HolditchDavis, Miles, et al., 2001; Olsen et al., 1992).
344
RESEARCH IN NURSING & HEALTH
Our negative control dimension also showed
changes over age and was related to maternal
and child variables. Finally, we did not examine
other environmental factors that influence
mother–child interactions, such as social support
from family members other than the father. Future
research is needed to examine how relationships
with other caregivers affect mother–premature
infant interactions.
Overall, our findings are consistent with complex processes affecting maternal–premature
infant interactions, as expected by the developmental science perspective. Three of six dimensions were related to both mother and infant
variables, providing evidence of processes involving both mother and infant affecting interactions.
Maternal dimensions also were correlated with
child social behaviors, and to a lesser extent with
other child dimensions. These findings are different than those of investigators who found that
differences in the interactions of prematures and
fullterms were due primarily to the mothers
(Jarvis et al., 1989; Muller-Nix et al., 2004). The
complexity of maternal–premature interactions
was shown by the fact that different variables were
related to each interactive dimension. Previous
investigators have examined how many of these
variables affected mother–premature interactions, but the direction of the effects were contradictory in many cases. For example, mothers have
been described as both being less and more
attentive to sicker prematures (Feingold, 1994;
Holditch-Davis, Cox, et al., 2003; Landry et al.,
1997; Muller-Nix et al.), possibly because these
studies included only one or two of these variables.
However, we did not examine the reciprocal
effects of mother and child variables; this needs
to be done in future studies.
To develop the knowledge needed for nurses
working with these vulnerable infants and
mothers, studies of mother–premature infant
interactions must be designed to capture the
complex factors affecting their interactions and
to examine how maternal and infant characteristics work together to influence interactions.
As shown in our findings and those of other
researchers (e.g., Landry et al., 1997; Ostfeld
et al., 2000; Tesh & Holditch-Davis, 1997),
multiple birth infants, infants with longer periods
of mechanical ventilation, minority mothers,
mothers with high levels of worry about the
child’s health or low stress due to infant illness in
the NICU, and mothers lacking support from the
infant’s father may be at particular risk for having
less-positive interactions than other premature
infants and, thus, need to be a focus of nursing
Research in Nursing & Health DOI 10.1002/nur
interventions. In accordance with the developmental science framework, however, having a
single risk may not be as important to infant
outcomes as mothers and infants who exhibit
multiple risk factors. Given that the interactive
risks involved both mothers and infants and the
interactive dimensions were inter-correlated, nursing interventions that focus on the entire mother–
infant system are most likely to be effective in
improving interactions and ameliorating the
developmental handicaps associated with prematurity.
REFERENCES
Barnard, K.E. (1983). Nursing Child Assessment
Teaching Scale. Seattle, WA: University of Washington.
Berlin, L.J., Brooks-Gunn, J., Spiker, D., & Zaslow, M.J.
(1995). Examining observational measures of emotional support and cognitive stimulation in black and
white mothers of preschoolers. Journal of Family
Issues, 16, 664–686.
Blondel, B., & Kaminski, M. (2002). Trends in the
occurrence, determinants, and consequences of
multiple births. Seminars in Perinatology, 26, 239–
249.
Bradley, R.H., Rock, S.L., Caldwell, B.M., & Brisby,
J.A. (1989). Use of the HOME Inventory for families
with handicapped children. American Journal on
Mental Retardation, 94, 313–330.
Brooks-Gunn, J., Klebanov, P.K., & Duncan, G.J.
(1996). Ethnic differences in children’s intelligence
test scores: Role of economic deprivation, home
environment, and maternal characteristics. Child
Development, 67, 396–408.
Brown, P., Rustia, J., & Schappert, P. (1991). A
comparison of fathers of high-risk newborns and
fathers of healthy newborns. Journal of Pediatric
Nursing, 6, 269–273.
Cairns, R.B., Elder, G.H., & Costello, E.J. (1996
Developmental science. Cambridge, Britain: Cambridge University Press.
Caldwell, B., & Bradley, R. (1984). Administration
manual for the home observation for measurement of
the environment (revised ed.). Little Rock, AR:
University of Arkansas at Little Rock.
Cho, J., Holditch-Davis, D., & Belyea, M. (2004).
Gender and ethnicity and the interactions of prematurely born children and their mothers. Journal of
Pediatric Nursing, 19, 163–175.
Coffman, S., Levitt, M.J., & Guacci-Franco, N. (1993).
Mothers’ stress and close relationships: Correlates
with infant health status. Pediatric Nursing, 19, 135–
140.
Crawford, J.W. (1982). Mother-infant interaction in
premature and full-term infants. Child Development,
53, 957–962.
CORRELATES OF MOTHER–PREMATURE INTERACTIONS / HOLDITCH-DAVIS ET AL.
Engelke, M.K., & Engelke, S.C. (1992). Predictors of
the home environment of high-risk infants. Journal of
Community Health Nursing, 9, 171–181.
Feeley, N., Gottlieb, L., & Zelkowitz, P. (2005). Infant,
mother, and contextual predictors of mother-very low
birth weight infant interaction at 9 months of age.
Journal of Developmental and Behavioral Pediatrics,
26, 24–33.
Feingold, C. (1994). Correlates of cognitive development in low-birth-weight infants from low-income
families. Journal of Pediatric Nursing, 9, 91–97.
Feldman, R., & Eidelman, A.I. (2004). Parent-infant
synchrony and the social-emotional development of
triplets. Developmental Psychology, 40, 1133–1147.
Feldman, R., Eidelman, A.I., & Rotenberg, N. (2004).
Parenting stress, infant emotion regulation, maternal
sensitivity, and the cognitive development of triplets:
A model for parent and child influences in a unique
ecology. Child Development, 75, 1774–1791.
Feldman, R., Weldman, J., Leckman, J.F., Kuint, J., &
Eidelman, A.I. (1999). The nature of the mother’s tie
to her infant: Maternal bonding under conditions of
proximity, separation, and potential loss. Journal of
Child Psychology and Psychiatry, 40, 929–939.
Field, T. (1995). Infants of depressed mothers. Infant
Behavior and Development, 18, 1–13.
Forcada-Guex, M., Pierrehumbert, B., Borghini, A.,
Moessinger, A., & Muller-Nix, C. (2006). Early
dyadic patterns of mother-infant interactions and
outcomes of prematurity at 18 months. Pediatrics,
118, e107–e114. Retrieved 8/2/2006 from http://
www.pediatrics.org/cgi/content/full/118/1/e107.
Gatsonis, C., & Sampson, A.R. (1989). Multiple
correlation: Exact power and sample size calculations. Psychological Bulletin, 106, 516–524.
Gerner, E.M. (1999). Emotional interaction in a group of
preterm infants at 3 and 6 months of corrected age.
Infant and Child Development, 8, 117–128.
Goldberg, S., Perrotta, M., Minde, K., & Corter, C.
(1986). Maternal behavior and attachment in lowbirth-weight twins and singletons. Child Development, 57, 34–46.
Gordon, R.A., Chase-Lansdale, P.L., & Brooks-Gunn, J.
(2004). Extended households and the life course of
young mothers: Understanding the associations using
a sample of mothers with premature, low birth weight
babies. Child Development, 75, 1013–1038.
Harrison, M.J. (1990). A comparison of parental
interactions with term and preterm infants. Research
in Nursing and Health, 13, 173–179.
Holditch-Davis, D. (2002, February). Methodological
issues in systematic observation. In D. HolditchDavis (Chair), Systematic observation in maternalchild research: Methods and issues. Symposium
presented at the 16th annual meeting of the Southern
Nursing Research Society, San Antonio, TX.
Holditch-Davis, D., Bartlett, T.R., & Belyea, M. (2000).
Developmental problems and the interactions
between mothers and their three-year-old prematurely born children. Journal of Pediatric Nursing, 15,
157–167.
Research in Nursing & Health DOI 10.1002/nur
345
Holditch-Davis, D., Bartlett, T.R., Blickman, A., &
Miles, M.S. (2003). Post-traumatic stress symptoms
in mothers of premature infants. Journal of Obstetric,
Gynecologic and Neonatal Nursing, 32, 161–171.
Holditch-Davis, D., Cox, M.F., Miles, M.S., & Belyea,
M. (2003). Mother-infant interactions of medically
fragile infants and non-chronically ill premature
infants. Research in Nursing and Health, 26, 300–
311.
Holditch-Davis, D., Docherty, S., Miles, M.S., &
Burchinal, M. (2001). Developmental outcomes of
infants with bronchopulmonary dysplasia: Comparison with other medically fragile infants. Research in
Nursing and Health, 24, 181–193.
Holditch-Davis, D., Edwards, L., & Helms, R. (1998).
Modeling development of sleep-wake behaviors: I.
Using the mixed general linear model. Physiology
and Behavior, 63, 311–318.
Holditch-Davis, D., Miles, M.S., Burchinal, M.,
O’Donnell, K., McKinney, R., & Lim, W. (2001).
Parental caregiving and developmental outcomes in
infants of mothers with HIV. Nursing Research, 50,
5–14.
Holditch-Davis, D., Roberts, D., & Sandelowski, M.
(1999). Early parental interactions with and perceptions of multiple birth infants. Journal of Advanced
Nursing, 30, 200–210.
Holditch-Davis, D., Scher, M., Schwartz, T., & HudsonBarr, D. (2004). Sleeping and waking state development in preterm infants. Early Human Development,
80, 43–64.
Holditch-Davis, D., Tesh, E.M., Goldman, B.D., Miles,
M.S., & D’Auria, J. (2000). Use of the HOME
Inventory with medically fragile infants. Children’s
Health Care, 29, 257–277.
Jarvis, P., Myers, B., & Creasey, G. (1989). The effects
of infants’ illness on mothers’ interactions with
prematures at 4 and 8 months. Infant Behavior and
Development, 12, 25–35.
Landry, S.H., Smith, K.E., Miller-Loncar, C.L., &
Swank, P.R. (1997). Responsiveness and initiative:
Two aspects of social competence. Infant Behavior
and Development, 20, 259–262.
Lee, T., Miles, M.S., & Holditch-Davis, D. (2006).
Father’s support to mothers of medically fragile
infants. Journal of Obstetric, Gynecologic and
Neonatal Nursing, 35, 46–55.
McGrath, M.M., Sullivan, M.C., & Seifer, R. (1998).
Maternal interaction patterns and preschool competence in high-risk children. Nursing Research, 47,
309–317.
Mercer, R.T. (2004). Becoming a mother versus
maternal role attainment. Journal of Nursing Scholarship, 36, 226–232.
Miles, M.S., Carlson, J., & Funk, S. (1996.) Sources of
support reported by mothers and fathers of infants
hospitalized in a neonatal intensive care unit.
Neonatal Network, 15, 45–54.
Miles, M.S., Funk, S.G., & Carlson, J. (1993). Parental
Stressor Scale: Neonatal intensive care unit. Nursing
Research, 42, 148–152.
346
RESEARCH IN NURSING & HEALTH
Miles, M.S., & Holditch-Davis, D. (1995). Compensatory parenting: How mothers describe parenting their
3-year-old prematurely born children. Journal of
Pediatric Nursing, 10, 243–253.
Miles, M.S., Holditch-Davis, D., Scher, M., &
Schwartz, T. (In press). Depressive symptoms in
mothers of prematurely-born infants. Journal of
Developmental and Behavioral Pediatrics.
Moran, G., & Pederson, D.R. (1998). Proneness to
distress and ambivalent relationships. Infant Behavior and Development, 21, 493–503.
Muller, K.E., & Fetterman, B.A. (2002). Regression and
ANOVA: A unified approach using SAS software
(p. 225). Cary, NC: SAS Institute, Inc.
Muller-Nix, C., Forcada-Guex, M., Pierrehumbert, B.,
Jaunin, L., Borghini, A., & Ansermet, F. (2004).
Prematurity, maternal stress and mother-child interactions. Early Human Development, 79, 145–158.
Olsen, S.L., Bates, J.E., & Kaskie, B. (1992). Caregiverinfant interaction antecedents of children’s schoolage cognitive ability. Merrill-Palmer Quarterly, 38,
309–330.
Ostfeld, B.M., Smith, R.H., Hiatt, M., & Hegyi, T.
(2000). Maternal behavior toward premature twins:
Implications for development. Twin Research, 3,
234–241.
Papousek, M., & von Hofacker, N. (1998). Persistent
crying in early infancy: A non-trivial condition of risk
for the developing mother-infant relationship. Child:
Care, Health, and Development, 24, 395–424.
Pauli-Pott, U., Mertesacker, B., & Beckman, D. (2004).
Predicting the development of infant emotionality
from maternal characteristics. Development and
Psychopathology, 16, 19–42.
Poehlmann, J., & Fiese, B.H. (2001). Parent-infant
interaction as a mediator of the relation between
neonatal risk status and 12-month cognitive development. Infant Behavior and Development, 24, 171–
188.
Pridham, K., Brown, R., Clark, R., Limbo, R.K.,
Schroeder, M., Henriques, J., et al. (2005). Effect of
guided participation on feeding competencies of
mothers and their premature infants. Research in
Nursing and Health, 28, 252–267.
Pridham, K., Lin, C.-Y., & Brown, R. (2001). Mothers’
evaluation of their caregiving for premature and fullterm infants through the first year: Contributing
factors. Research in Nursing and Health, 24, 157–
169.
Radloff, L. (1977). The CES-D scale: A self-report
depression scale for research in the general population. Applied Psychological Measurement, 1, 385–
401.
Schmucker, G., Brisch, K.-H., Kohntop, B., Betzler, S.,
Osterle, M., Pohlandt, F., et al. (2005). The influence
of prematurity, maternal anxiety, and infants’ neurobiological risk on mother-infant interactions. Infant
Mental Health Journal, 26, 423–441.
Shields-Poe, D., & Pinelli, J. (1997). Variables associated with parental stress in neonatal intensive care
units. Neonatal Network, 16, 29–37.
Research in Nursing & Health DOI 10.1002/nur
Singer, L., Davillier, M., Bruening, P., Hawkins, S., &
Yamashita, T. (1996). Social support, psychological
distress, and parenting strains in mothers of very low
birthweight infants. Family Relations, 45, 343–350.
Singer, L.T., Fulton, S., Davillier, M., Koshy, D.,
Salvator, A., & Baley, J.E. (2003). Effects of infant
risk status and maternal psychological distress on
maternal-infant interactions during the first year of
life. Journal of Developmental and Behavioral
Pediatrics, 24, 233–241.
Smith, K.E., Landry, S.H., & Swank, P.R. (2006). The
role of early maternal responsiveness in supporting
school-aged cognitive development for children
who vary in birth status. Pediatrics, 117, 1608–1617.
Tesh, E.M., & Holditch-Davis, D. (1997). HOME
Inventory and NCATS: Relation to mother and child
behaviors during naturalistic observations. Research
in Nursing and Health, 20, 295–307.
Thoman, E.B., Acebo, C., & Becker, P.T. (1983). Infant
crying and stability in the mother-infant relationship:
A systems analysis. Child Development, 54, 653–
659.
Traustadottir, R. (1991). Mothers who care: Gender,
disability, and family life. Journal of Family Issues,
12, 211–228.
Vohr, B.R., Allan, W.C., Westerveld, M., Schneider,
K.C., Katz, K.H., Makuch, R.W., et al. (2003).
School-age outcomes of very low birth weight infants
in the indomethacin intraventricular hemorrhage
prevention trial. Pediatrics, 111, e340–346.
Retrieved 8/29/2004 from http://www.pediatrics.
org/cgi/content/full/111/4/e340.
Vukelich, C., & Kliman, D.S. (1985). Mature and
teenage mothers’ infant growth expectations and use
of child development information sources. Family
Relations, 34, 189–196.
Weiss, S.J., & Chen, J.-L. (2002). Factors influencing
maternal mental health and family functioning during
the low birthweight infant’s first year of life. Journal
of Pediatric Nursing, 17, 114–125.
Weiss, S., Wilson, P., Hertenstein, M., & Campos, R.
(2000). The tactile context of a mother’s caregiving:
Implications for attachment of low birth weight
infants. Infant Behavior and Development, 23, 91–
111.
Weissman, M.M., Sholomskas, D., Pottenger, M.,
Prusoff, B.Q.A., & Locke, B.Z. (1977). Assessing
depressive symptoms in five psychiatric populations:
A validation study. American Journal of Epidemiology, 106, 203–214.
Wille, D.E. (1991). Relation of preterm birth with
quality of infant-mother attachment at one year.
Infant Behavior and Development, 14, 227–240.
Yogman, M., Kindlon, D., & Earls, F. (1995). Father
involvement and cognitive/behavioral outcomes in
preterm infants. Journal of the American Academy of
Child and Adolescent Psychiatry, 34, 58–66.
Zahr, L.K. (1999). Predictors of development in
premature infants from low-income families: African
Americans and Hispanics. Journal of Perinatology,
19, 284–289.
`