Effect
of the
Pediatrician
on the
Mother-Infant
Relationship
Patrick H. Casey,
MD, and J. Kenneth
Whitt,
PhD
From the Robert Wood Johnson Clinical Scholars Program, Departments of Pediatrics
and Psychiatry, University of North Carolina School of Medicine, Chapel Hill
ABSTRACT. Child development literature suggests a
re-lationship between mother-child interaction and
en-hanced infant development. We conducted a randomized
clinical trial to determine if a pediatrician’s guidance improves the mother-infant relationship and the infant’s
development. Thirty-two normal mothers and their
healthy first-born infants were followed by one
pediatri-cian at 2, 4, 8, 15, and 21 weeks of age. These dyads were
randomly assigned to a control group, who received
cus-tomary care, or to an intervention group, who also
re-ceived guidance based on the infant’s developmental sta-tus at each age. Just prior to a 27-week visit, the
mother-infant relationship was assessed by a person blind to
group assignment. Infant development was assessed with
the Bayley Mental Scales of Infant Development and two
of the Uzguris-Hunt Ordinal Scales. Intervention group
mothers were rated significantly higher on sensitivity, cooperation, appropriateness of interaction, and appro-priateness of play (P < .05). Experimental infants were advanced on the Vocal Imitation ordinal scale. This study shows the effectiveness of this intervention on both the
mother-infant relationship and infant development and
supports pediatric involvement in this biosocial approach to well child care. Pediatrics 65:815-820, 1980; well child
care, child health supervision, attachment, mother-child interaction.
Child health supervision comprises a major
pro-portion of contemporary pediatric practice.’
De-spite the traditional importance of child health
supervision visits to pediatric practice, current
rec-ommendations concerning the content of these
vis-its are based on a consensus of what should be done
in practice, rather than on research designed to
determine the effectiveness of these clinical
ap-proaches. Perhaps due to this lack of empirical data,
there has been discussion in the recent literature
questioning the appropriate content and frequency
Received for publication April 6, 1979; accepted May 30, 1979.
Reprint requests to (P.H.C.) University of Arkansas for Medical
Sciences, Department of Pediatrics, Little Rock, AK 72203.
PEDIATRICS (ISSN 0031 4005). Copyright © 1980 by the
American Academy of Pediatrics.
of child health supervision visits. For example,
Hoe-kelman2 reported that an abbreviated three-visit
schedule for the first year of life does not reduce
the adequacy of care as measured by maternal
knowledge, compliance, satisfaction, and the
attain-ment of planned health supervision. Chamberlin3
questioned the conclusions of this study because of
the end points that were used, and he suggested
that these visits be focused on child development
and behavior.4 Clearly, there is a need for research
to direct the practice of well child care.
The content of child health supervision visits as
commonly practiced is primarily medical-physical
interests.5 It is suggested by research that
improve-ment in the physical outcomes of mortality and
morbidity is unlikely to result from these clinical
visits.8 The American Academy of Pediatrics has
suggested that these visits should aim not only to
prevent mortality and morbidity, but also to
pro-mote optimal childhood growth and development.9
Improvements in these behavioral developmental
outcomes will also be difficult to measure without
large-scale longitudinal studies.
In order to bypass the feasibility problems of
measuring these long-term results, short-term
prox-imate outcomes can be used to determine
appro-priateness of care if their relationship to improved
long-term outcomes has been established.’0 One
important short-term outcome for child health
su-pervision visits is the early mother-infant
relation-ship. Research by developmental psychologists over
the last decade indicates an association between
the quality of the early mother-infant relationship
and the child’s later intellectual performance and
social development.”2 For example, a
cross-sec-tional study of 36 mother-infant pairs at 9 to 18
months of age found that infant cognitive, language,
and social development was strongly and positively
related to maternal stimulation, responsiveness,
and positive emotion.’3 As suggested by this litera..
relation-ship may be a viable short-term goal for the
pedia-trician in child health supervision seeking to
pro-mote the infant’s long-term cognitive and social
competence.
This investigation examines the effectiveness of
child health supervision designed to enhance the
relationship between the mother and her child.
Specifically, this study sought to answer two
ques-tions: (1) Can a pediatrician in well child care
promote the mother-child interaction in the infant’s
first six months of life? (2) Will this intervention
affect the cognitive development of the infant at six
months of age?
METHODS
Subjects and Allocation
This study was conducted in the primary care
clinic of the North Carolina Memorial Hospital
between July 1977 and July 1978. A group of normal
mothers and their healthy first-born infants were
recruited for this study within three days of
deliv-ery. Nursery infants were eligible for inclusion in
the study whenever chart review revealed: (1)
five-minute Apgar score greater than 6, (2) gestation
age greater than 36 weeks, (3) birth weight greater
than 2,500 gm, and (4) no significant congenital
anomalies or systemic disease. Mothers included
those who (1) were primiparous, (2) exceeded 16
years of age, (3) had family income of less than
$15,000 per year, (4) resided in the surrounding five
county area, (5) planned to provide the primary
care of the infant, and (6) had no identified source
of medical care. During the six months of nursery
chart review, 59 of 543 mother-infant pairs met the
selection criteria, and 47 (80%) agreed to be followed
by the author (P.C.) for well child care. The
major-ity of those who declined to participate did so
because of problems with transportation; the
re-mainder desired to obtain medical care in their local
communities. Following discharge from the
hospi-tal, the mother-infant pairs were stratified by
ma-ternal education and randomly assigned to
experi-mental and control groups. Three of the 47
mother-infant pairs who agreed to participate never came
for a visit. Of the 44 mother-infant pairs who were
randomly assigned, ten dropped out after the first
visit, six from the experimental group and four from
the control group. Half of these dropped out
be-cause of transportation problems. The other half
were lost to follow-up. Two others moved from the
surrounding area after the fourth visit. A total of 15
mother-infant pairs in the experimental group and
17 in the control group completed the study.
Intervention
All of the mother-infant pairs were followed
din-ically by a single peditrician at 2, 4, 8, 15, 21, and 27
weeks of age. No charges were made for these
scheduled child health supervision visits during the
study.
Infants in both groups received physical care
consistent with the American Academy of
Pediat-rids’ 1972 standards of preventive health care in
terms of number and content of visits.’4 The timing
of the visits was different from the Academy’s
red-ommendations because we believed that early visits
were necessary in order to affect the mother-child
relationship. The control group received thorough
physical examinations and discussions of physical
and preventive care such as accident prevention
and nutrition. The mothers in the experimental
group received discussions at all visits designed to
enhance the affective interaction between mother
and infant and to stimulate the infants’ cognitive
development. These discussions occurred
concomi-tant with and in place of certain aspects of the
physical examination. The duration of clinic visits
for the two groups was not different and lasted 25
to 30 minutes.
The objectives of these age-appropriate
interven-tions were as follows: (1) to improve the mothers’
understanding of normal infant development, (2) to
increase the mothers’ understanding of the
individ-uality of their infants, (3) to promote the mothers’
awareness of the social nature of infant behaviors,
(4) to encourage maternal responsiveness to these
social behaviors, and (5) to promote the mothers’
feelings of confidence and competence to affect
their infant’s development. To accomplish these
goals, the pediatrician discussed with the mothers
the normal developmental sequences of infant
be-haviors in a social context, as illustrated in the
Figure. For example, infant crying at the 2- and
4-week visit and infant vocalization at the 15-week
visit were presented as infant efforts at social
com-munication. Normal variation in these individual
behaviors was discussed, and maternal responses to
these behaviors was recommended, as gleaned from
the developmental psychology literature. In
gen-eral, mothers were advised to respond to the infant
behaviors in a contingent way using vocal and
sup-portive physical behaviors. Physician modeling (eg
maintaining eye contact, soothing crying, or
vocal-izing in response to infant vocalization) was utilized when appropriate.
In order to control for the potential beneficial
effects of the frequent visits with the same
pedia-trician, a third group of 12 mother-infant pairs was
also evaluated at 27 weeks of age.This no-contact
comparison group was selected from the same birth
cohort and met the same selection criteria as the
other groups. They had not been invited to
psi-tic-ipate in the study either because they did not live
Figure. Temporal sequence of infant developmental patterns as focus of pediatric intervention.
Infant Behavior Birth I Month 2 Month 3 Month 4 Month 5 Month 6 Month
Physical Concerns Feeding Bowel Patterns Growth
Crying
Gazing
Smiling
Vocalizing
Laughing
Gross and Fine Motor Exploration
charts were not reviewed due to holidays or
week-ends. This no-contact group received neither the
experimental intervention nor continuity of care by
the pediatric investigator. They received at least
one well child care visit from other pediatricians in
the primary care setting. A total of 34 mothers
(virtually all those of the birth cohort who were not
invited to participate because of the above reasons)
were written to and offered a free physical
check-up, including immunizations and developmental
testing. A total of 18 accepted, and 12 (35% of the
mailing) infants were evaluated at 27 weeks of age.
Data concerning demographic and attitudinal
variables were gathered before the first clinic visit
since these variables could influence the endpoints
of interest. Table 1 shows some of these
indepen-dent variables. In order to determine whether there
were differences between the experimental group
when compared to the control and the no-contact
groups,
x2
analyses were performed for categoricalvariables and t tests for continuous variables. All
comparisons were nonsignificant except as shown.
Assessment
An assistant blind to experimental group status
assessed the mother-infant relationship based on
two hours of contact with the pair, including
labo-ratory and testing time, just prior to the 27-week
physician visit. The mother and infant were
ob-served through a two-way mirror for 21 minutes in
a naturalistic “living room” setting which allowed
for a variety of solitary or interactive activities.
This room was approximately 8 x 10 ft and consisted
of an upholstered sofa and chair, an end table with
lamp and current magazines, a rug, an infant crib,
and a box of toys. The mothers were instructed that
we were interested in observing the infant at play
and that they could do as they pleased.
The rating instruments consisted of eight
individ-ual scales selected as empirically established
mea-sures of relevant aspects of the mother-infant
rela-tionship developed by Ainsworth’s group in their
studies of infant-mother attachment. The scales
used were entitled Sensitivity-Insensitivity,
Coop-eration-Interference, Accessibility-Ignoring,
Inter-action Offered, Interaction Appropriateness, Visual
Contact, Vocal Contact, and Appropriateness of
Play.’7 Each scale had nine point dimensions, with
9 as the best score. The anchor points of 9, 7, 5, 3,
and 1 were behaviorally defined in detail. The scales
cannot be presented here because there are several
pages of instruction for each scale. A brief summary
of the behaviors at the extreme poles of the
Sensi-tivity-Insensitivity scale is provided as an example.
A sensitive mother is finely attuned to her baby’s
signals; she is aware of them, interprets them accurately,
and responds to them promptly and appropriately. An
insensitive mother is geared almost exclusively to her
own wishes, moods and activities, so that her
interven-tions are rarely contingent upon her baby’s signals. Reliability data for these scales have been published and exeed 0.85.1819
Infant development was assessed immediately
after the observation session by two methods: the
Bayley Mental Scales of Infant Development
(M.D.I.), a test of development standardized by age
and the Object Permanence and Vocal Imitation
scales of the Ordinal Scales of Psychological
Devel-opment, which are individual scales each consisting
of many increasingly sophisticated levels based on
TABLE 1. Average Values of Demographic and Attitudinal Variables Study Groups
for the Three
Variables Experimental
(n = 15)
Control
(n = 17)
No-Contact Comparison
(n = 12)
Infant
Apgar score 8.6 9Q* 9.Ot
Birth weight (mg) 3398 3335 3247
Sex (% male) 46.7 47.1 58.3
Maternal
Age (years) 20.8 21.47t 19.9
Education (years completed) 11.9 12.4 11.6
Socioeconomic status (Green)’5 53.9 57.9 53.4
Race (% caucasian) 46.7 23.5 33.3
Marital status (% married) 60 88.2 38.5t
Breast feed (% yes) 26.7 17.6 15.4
Health locus of control’6 at 27 wk 36.8 36.4 36.5
Time mother primary caregiver (mo) 4.1 3.9 5.1
C
p< ,#{216}5,compared to experimental group.t
.05 < P < .2.RESULTS
Mother-Infant Relationship
Experimental, control, and no-contact
compari-son group mean scores for the maternal rating
measures are presented in Table 2. The effects of
the intervention discussion on the mother-infant
relationship were analyzed by comparing the
exper-imental and control group means using one-tailed
t-statistics. These blind ratings uniformly favored
the experimental group. Mothers who received the
intervention were rated significantly higher than
control group mothers on measures of Interaction
Appropriateness (P < .01),
Cooperation-Interfer-ence (P < .01), Sensitivity-Insensitivity (P < .05),
and Appropriateness of Play (P < .05). The other
measures (Accessibility-Ignoring, Interaction
Of-fered, Vocal Contact, and Visual Contact) favored
the experimental group (P < .10). A similar analysis
was performed comparing the intervention group
means with the blind ratings of the no-contact
comparison group of mother-infant pairs who
re-ceived child health supervision from other pediatric
clinicians. The results obtained were similar to the
previous analysis. As reported in Table 2, all differ-ences in mean rating scores favored the intervention mothers. Statistical results indicated significantly
higher ratings of the mother-infant relationship for
the experimental group Interaction
Appropriate-ness (P < .05) and Sensitivity-Insensitivity (P <
.05) measures, as well as a trend for the Accessibil-ity-Ignoring measure (P < .10).
Infant Cognitive Development
The mean scores of infant developmental
evalu-ations administered blind to the three subject
groups at 27 weeks of age are shown in Table 3.
Statistical comparison of the experimental and
con-trol infants reveals no significant differences on the
Bayley Mental Scales of Infant Development or on
the Object Permanence Scale. However, the
inter-vention group infants tended to perform at a
some-what more advanced level than control infants on
the Vocal Imitation Scale (P
=
.08).Similarly, experimental group infant
perform-ance on the Bayley Mental Scales was not
signifi-cantly different when compared with the no-contact
comparison group scores. On the Ordinal Scales for
these groups, however, the intervention group
chil-dren demonstrated significantly higher
perform-ance on the Vocal Imitation Scale (P < .01) and a
positive trend on the Object Permanence Scale (P
= .07).
DISCUSSION
The data presented in this paper suggest that the
mother-infant pairs in the experimental group
en-joyed a more harmonious relationship as measured
by the rating scales. Also, although all infant groups
were normal as measured by the standardized test
of development, the experimental group infants
were more advanced in vocal imitation. Since a
major focus of our intervention consisted of
encour-aging maternal sensitivity and verbal
responsive-ness to infant vocalizations, a relationship is
sug-gested among the pediatric intervention, maternal
behavior, and advanced infant vocal behavior.
The comparison between the two randomly
allo-cated groups should be emphasized. Because the
group of mother-infant pairs that received child
health supervision by pediatricians other than the
investigator was not randomly allocated, it is thus
liable to a variety of selection biases, despite being
similar in sociodemographic variables. Also, there
was no control over the instruction regarding child
Experimental Control Group
Group (n = 15) (n 17)
No-Contact Group (n = 12)
Interaction Appropriateness 7.06 5.05w 5.25t
Cooperation-Interference 6.93 5.29* 6.5
Appropriateness of Play 7.35 6.Ot 6.23
Sensitivity-Insensitivity 7.46 6.23t 6.17t
Accessibility-Ignoring 8.26 7.29 7.25
Vocal Contact 7.0 5.88 6.17
Interaition Offered 7.33 6.17 6.5
Visual Contact 8.33 7.76 7.9
C p <
.oi,
one-tailed t test compared to experimental group. t.01 <P< .05.TABLE 3. Infant Development Tests at 27 Weeks (Mean Values)
Experimental Group Control Group No-Contact Group
(n=15) (n=17) (n=12)
Bayley M.D.I. 111.8 109.2 109.3
Object Permanence 3.53 3.47 3.0*
Vocal Imitation 2.93 2.41 2.07t
*
.05 < P < .1, t test, one-tailed, compared to experimental group.tP< .01.
TABLE 2. Ratings of Mother-Infant Relationship at 27 Weeks (Mean Values)
visits for the no-contact group. The findings of this
group are reported because the results are similar
to the control group, and they support the
differ-ences of the experimental group.
The results of this study are pertinent to two
issues in child care: the effectiveness of the
coun-seling provided by pediatric clinicians in
nonphysi-cal areas during child health supervision visits, and
the possibility of improving the early mother-infant
relationship to improve childrens’ long-term
well-being.
A recent Task Force on Pediatric Education
em-phasized the importance of “biosocial” pediatrics
and pediatric health maintenance.2’ Unfortunately,
there is little research to guide educators and
prac-titioners in these areas. While some have
ques-tioned the necessity of well child visits,22 others feel
that these visits should be continued with a focus
on child development and behavior.4 Improvements
in children’s cognitive development and long-term
behavioral outcomes have been demonstrated after
an intensive health supervision schedule that
in-cluded home visits by nurses.23’24 This randomized
controlled trial demonstrates the beneficial effects
of the relatively brief time expenditure of physician
counseling in child health supervision on the early
mother-child interaction, an outcome which is
thought to be of central importance to the physical
and psychosocial well-being of children. The use of
this short-term outcome allowed us to measure
important differences between groups which would
have otherwise required years of longitudinal
fol-low-up, with multiple intervening variables
con-founding the results.
The second issue relevent to this report is the
possible enhancement of the early mother-infant
interaction in order to improve children’s
develop-ment. Bronfenbrermer5 developed some “principles
of early intervention” after reviewing the
experi-ence of the early education intervention programs
of the previous decade. In the child’s early years
the primary objective of such an intervention is the
establishment of a stable emotional relationship
between parent and child which involves frequent
reciprocal interactions. We felt that the
pediatri-cian’s general access to the young infant and family
and the physician’s documented positive influence
on parent behavior6 made it reasonable to use the
child health supervision visits to improve infant
development by facilitating the early mother-child
relationship. This study, which resulted in
ad-vanced vocal imitation, verifies the importance of
this approach. The Ordinal Scales of Psychological
Development are thought to be particularly
sensi-tive to differences in various aspects of infant
de-velopment which result from different child-rearing
environments.20 Infant vocalization rate has been
shown to be increased by responsive social
environ-ment.27 The advanced vocal imitation of the
exper-imental group infants suggests that the mothers in
that group followed the pediatrician’s advice and
were more verbally responsive to infant
vocaliza-tions. The association ofthis skifi to future cognitive
development is unknown. However, several studies
have found significant relationships between verbal
skills in girls before one year of age and future
IQ.28’
In summary, this clinical trial with healthy
pri-miparous mother-infant pairs demonstrated that a
brief pediatric intervention during child health
su-pervision was effective in changing the relationship
months of life, as well as one aspect of infant
de-velopment. Long-term follow-up will be necessary
to measure the impact of this approch on the
psy-chosocial and cognitive functioning of these
chil-dren. Although the generalizabiity of these results
is limited because of the stringent recruitment
cri-teria and because the results were achieved by a
single pediatrician, these findings offer empirical
support for a more biosocial approach to routine
child health supervision.
IMPLICATIONS
This study suggests that there is a more active
role for the child health clinician in facilitating the
mother-child relationship beyond the immediate
perinatal period. The framework of the traditional
child health supervision visits provides the clinical
encounters to achieve these effects.
ACKNOWLEDGMENTS
The collaboration of Craig Ramey, Ron Haskins, Earl
Schaefer, and the staff in making the Frank Porter
Gra-ham Child Development Center available for this
re-search is gratefully acknowledged. Special appreciation is
expressed to Frank Loda, David McKay, and Robert
Fletcher for their constructive suggestions on earlier drafts of this paper and to Kathy Pasco and Karen Strain for assistance in the collection and analysis of test data.
REFERENCES
1. American Academy of Pediatrics, Committee on Manpower:
Projecting pediatric practice patterns. Pediatrics 62(suppl):
627, 1978
2. Hoekelman RA: What constitutes adequate well-baby care?
Pediatrics 55:313, 1975
3, Chamberlin RW: What is “adequate well baby care”?
Pedi-atrics 58:772, 1976
4, Chamberlin RW: Management of pre-school behavior
prob-lems. Pediatr Cliii North Am 21:33, 1974
5, Korsch BM, Negrete VF, Mercer AS: How comprehensive
are well child visits? Am J Dis Child 122:483, 1971
6. Rogers KD, Ernst R, Schulman I, et at: Effectiveness of
aggressive follow-up on Navajo infant health and medical
care use. Pediatrics 53:721, 1974
7. Gordis L, Markowitz M: Evaluation of the effectiveness of
comprehensive and continous pediatric care. Pediatrics 48:
766, 1971
8. Lewis C: What is the evidence? Am J Dis Child 122:469, 1971
9, Statement of child health supervision, in Bulletin of
Pedi-atric Practice. Evanston, IL, American Academy of Pediat-rics, 1972, vol 6, p 3
10. Brook RH, Davies-Avery A, Greenfield 5, et al: Assessing
the quality of medical care using outcome measures: An
overview of the method. Med Care 15(suppl to No. 9), 1977
11. Clarke-Stewart A: Child Care in the Family: A Review of
Research and Some Propositions for Policy. New York,
Academic Press, 1977
12. Ainsworth MDS, Bell SM: Mother-infant interaction and
the development of competence, in Connelly K, Bruner J
(eds): The Growth of Competence. New York, Academic Press, 1974
13. Clarke-Stewart A: Interactions between Mothers and Their
Young Children: Characteristics and Consequences.
Mon-ographs of the Society for Research in Child Development,
1973, vol 38, nos 6 and 7
14. Standards of Child Health Care, ed 2. Evanston, IL,
Amer-ican Academy of Pediatrics, 1972; Revised ed 3, 1977
15. Green LW: Manual for scoring socioeconomic status for
research on health behavior. Public Health Rep 85:815, 1970
16. Wallston BS, Wallston KA, et al: Development and
valida-tion of the health locus of control scale. J Consult Clin
Psychol 44:580, 1976
17. Boyer EG, Simon A, Karafin GR: Measures of Maturation:
An Anthology ofEarly Childhood Observation Instruments.
Philadelphia, Research for Better Schools, mc, 1973
18. Ainsworth MDS, Bell SM, Stayton DJ: Individual
differ-ences in strange-situation behavior in one-year olds, in
Schaffer HR (ed): The Origins ofHuman Social Relations.
London, Academic Press, 1971
19. Ainsworth MDS, Bell SM: Some Contemporary Patterns of
Mother-Infant Interaction in the Feeding Situation, in
Am-brose JA (ed): Stimulation in Early Infancy. London,
Aca-demic Press, 1969
20. Uzguris IC, Hunt JMV: Assessment in Infancy. Urbana, University of Illinois Press, 1975
21. The Future of Pediatric Education: A Report by the Task Force on Pediatric Education. Evanston, IL, American
Academcy of Pediatrics, 1978
22. Stickler GB: How necessary is the “routine check-up”? Clin
Pediatr 6:454, 1967
23. Gutelius MF, Kirsch AD, MacDonald 5, et al: Promising
results from a cognitive stimulation program in infancy. Clin Pediatr 11:585, 1972
24. Gutelius MF, Kirsch AD, MacDonald 5, et al: Controlled
study of child health supervision: Behavioral results. Pedi-atrics 60:294, 1977
25. Bronfenbrenner U: Is early intervention effective? Facts and
principles of early intervention: A summary, in Clarke AM,
Clarke ADB (eds): Early Experience: Myth and Evidence. London, Open Books, 1976
26. Stolz LM: Influences on Parent Behavior. Stanford, CA,
Stanford University Press, 1967
27. Millar WS: Operant acquisition of social behaviors in
in-fancy: Basic problems and constraints, in Reese HW (ed):
Advances in Child Development and Behavior. New York,
Academic Press, 1976, pp 107-140
28. Cameron J, Livson N, Bayley N: Infant vocalizations and
their relationship to mature intelligence. Science 157:331,
1967
29. Moore T: Language and intelligence: A longitudinal study of
