A validity and reliability study of assessment and screening for sustained withdrawal reaction in infancy The Alarm Distress Baby Scale (ADBB)

Abstract

Sustained withdrawal behavior in infancy is an important alarm signal to draw attention to both organic and relationship disorders. A withdrawal scale, the Alarm Distress Baby scale (ADBB), for infants between 2 and 24 months of age was built. This article describes the construction of the scale and the assessment of its psychometric properties. The ADBB has good content validity, based on the advice of seven experts. The scale has good criterion validity: first, as a measure of the infant's withdrawal reaction, with a very good correlation between nurse and pediatrician on the ADBB (rs = 0.84), and second, as a screening procedure for detecting the developmental risk of the infant. The cutoff score of 5 with a sensitivity of 0.82 and a specificity of 0.78 was determined to be optimal for screening purposes. The scale has good construct validity, with good convergent validity with both the Spitz (1951) and the Herzog & Rathbun (1982) lists of symptoms of infant depression (rs = 0.61 and 0.60, respectively). Exploratory factor analysis showed two different factors, consistent with the scale's construct. Reliability was satisfactory with good internal consistency for both subscales (the Cronbach = 0.80 for the first subscale and 0.79 for the second) and for the global scale ( = 0.83). The test-retest procedure showed good stability over time (rs = 0.90 and 0.84 for the two different raters). The scale could be used in different clinical settings, provided a sufficient level of social stimulation is given to the infant in a relatively brief period of time. The scale can be used by nurses and psychologists or by medical doctors after a short period of training. © 2001 Michigan Association for Infant Mental Health.

Full text

IMHJ (Wiley) RIGHT BATCH
short
standard
top of AH
559 base of drop
INFANT MENTAL HEALTH JOURNAL, Vol. 22(5), 559– 575 (2001)
䊚2001 Michigan Association for Infant Mental Health
ARTICLE
A VALIDITY AND RELIABILITY STUDY OF
ASSESSMENT AND SCREENING FOR SUSTAINED
WITHDRAWAL REACTION IN INFANCY: THE
ALARM DISTRESS BABY SCALE
ANTOINE GUEDENEY
Department of Child and Adolescent Psychiatry, Hoˆpital Bichat-Claude Bernard
JACQUES FERMANIAN
Department of Biostatistics, Hoˆpital Necker-Enfants Malades
ABSTRACT: Sustained withdrawal behavior in infancy is an important alarm signal to draw attention to
both organic and relationship disorders. A withdrawal scale, the Alarm Distress Baby scale (ADBB),for
infants between 2 and 24 months of age was built. This article describes the construction ofthe scale and
the assessment of its psychometric properties. The ADBB has good content validity, based on the advice
of seven experts. The scale has good criterion validity: first, as a measure of the infant’s withdrawal
reaction, with a very good correlation between nurse and pediatrician on the ADBB (r⫽0.84), and
s
second, as a screening procedure for detecting the developmental risk of the infant. The cutoff score of
5 with a sensitivity of 0.82 and a specificity of 0.78 was determined to be optimal for screeningpurposes.
The scale has good construct validity, with good convergent validity with both the Spitz (1951) and the
Herzog & Rathbun (1982) lists of symptoms of infant depression (r⫽0.61 and 0.60, respectively).
s
Exploratory factor analysis showed two different factors, consistent with the scale’sconstruct. Reliability
was satisfactory with good internal consistency for both subscales (the Cronbach
␣
⫽0.80 for the first
subscale and 0.79 for the second) and for the global scale (
␣
⫽0.83). The test-retest procedure showed
good stability over time (r⫽0.90 and 0.84 for the two different raters). The scale could be used in
s
different clinical settings, provided a sufficient level of social stimulation is given to the infant in a
This study would not have been possible without the support of Marcelle Delour, M.D., Head of medical services,
Protection Maternelle et Infantile de Paris. Janine Charon, M.D. (Head of PMI, 14th district of Paris), Franc¸ois Roge´,
M.D., pediatrician (PMI, IPP), Malika Benjellal-Zamoun, pue´ricultrice de jeunes enfants (PMI, IPP), Annie Gauvain-
Piquard, M.D. (IGR, Villejuif), Martine Vermillard, nurse (PMI, IPP), Caroline Dumont (psychologist), He´loise Court-
ier (psychologist), and Edith Thoueille, director (PMI, IPP) have directly participated in the study. Rachael Henry,
Ph.D. (Wollongong, Australia), R. Kumar, Ph.D. M.D. (Maudsley, London), and Campbell Paul, M.D. (Royal Chil-
dren’s Hospital, Melbourne) have kindly provided expert advice at different points of the study. Fre´de´ric Atger, M.D.
(Institut Mutualiste Monsouris, Paris) has been very helpful in the use of the statistical software. Annie Nataf-Cooper
Rusconi, M.D., Marianne Kumar, Gise`le Danon, M.D., and Denise Parise made translations and back-translations of
the scale from French to English. Kaija Purra, M.D. (Tampere University, Finland), Priti Patel, M.D., Dilys Daws
(Tavistock Clinic, London), and Stephen Matthey (South Western Sydney Area Health Service, Australia) havekindly
provided editing help. The study has benefited from the financial help of the Nestle´ Corporation. Direct correspondence
to: Dr. Antoine Guedeney, Hoˆpital Bichat-Claude Bernard, policlinique Ney, 124 blvd Ney, 75018 Paris, France; e-
mail: guedeney@bch.ap-hop-paris.fr.

560 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of text
relatively brief period of time. The scale can be used by nurses and psychologists or by medical doctors
after a short period of training.
RESUMEN: Cuando la conducta de ensimismarse o apartarse es sostenida durante la infancia nos encon-
tramos frente a una importante sen˜al de alarma que nos hace dirigir la atencio´n hacia los trastornos
orga´nicos y los que se presentan dentro de una relacio´n. Se ha establecido una escala para determinar el
ensimismamiento en infantes entre los 2 y 24 meses de edad, la cual se denomina “Alarm Distress Baby
scale” (ADBB). Este ensayo describe la estructuracio´n de dicha escala y la evaluacio´n de sus propiedades
sicome´tricas. La ADBB tiene una muy buena validez de contenido, basa´ndose en el consejo de siete
expertos. La escala tambie´n posee una validez en cuanto al buen criterio: primero como una forma de
medir la reaccio´n de ensimismamiento del infante, con una buena correlacio´n entre enfermera y pediatra
en la ADBB (rs ⫽0.84); y segundo, como un procedimiento investigativo para detectar el riesgo de
desarrollo del infante. Se determino´ que un puntaje ma´ximo de 5, con una sensibilidad de 0.82, y una
especifidad de 0.78, era el puntaje o´ptimo para los propo´sitos investigativos. La escala tiene ası´mismo
una buena validez de estructuracio´n, con buena validez convergente con referencia a las listas desı´ntomas
de la depresio´n infantil de Spitz (1951) y de Herzog & Rathbun (1982) (rs: 0.61 y 0.60). El ana´lisis del
factor de exploracio´n mostro´ dos diferentes factores, consistentes con la estructuracio´n de la escala. La
fiabilidad es satisfactoria con buena consistencia interna tanto para las sub-escalas (alpha de Cronbach
⫽0.80 para la primera sub-escala y 0.79 para la segunda) como para la escala global (alpha ⫽0.83). El
procedimiento de examinacio´n y re-examinacio´n mostro´ buena estabilidad con el pasar del tiempo (rs ⫽
0.90 y 0.84 para los dos sistemas de evaluacio´n). Esta escala pudiera ser usada en diferentes escenarios
clı´nicos, siempre y cuando que se provea un nivel suficiente de estı´mulo hacia el infante en un perı´odo
de tiempo relativamente breve. La escala puede ser usada por las enfermeras y los sico´logos ası´ como
tambie´n por los me´dicos, despue´s de un perı´odo corto de entrenamiento.
RE
´SUME
´:Un comportement prolonge´ de repli et retrait est un signal d’alarme important attirant l’attention
sur a` la fois des troubles organiques et des troubles relationnels. Une e´chelle de retrait et repli (e´chelle
d’alarme pour le de´sarroi chez le be´be´, abre´ge´ ADBB en anglais) pour les nourissons entre 2 et 24 mois
ae´te´ construite. Cet article de´crit la construction de l’e´chelle et l’e´valuation de ses proprie´te´s psychom-
e´triques. L’ ADBB a une bonne validite´ de teneur, base´e sur les conseils de sept experts. L’e´chelle a une
bonne validite´ de crite`re: d’abord en tant que mesure de la re´action de repli et retrait du nourrisson, avec
une tre`s bonne corre´lation entre le pe´diatre et l’infirmie`re sur la ADBB (rs⫽0,84); et deuxie`mement en
tant que proce´dure de de´pistage du risque dans le de´veloppement du nourrisson. La limite place´ea` 5 avec
une sensibilite´ de 0,82 et une spe´cificite´ de 0,78 fut de´termine´e afin d’eˆtre optimale pour le de´pistage.
L’e´chelle a un bonne validite´ conceptuelle, avec une bonne validite´ de convergence avec les listes de
symptoˆmes de de´pression du nourrisson de Spitz (1951) et Herzog et Rathbun (1982) (rs: 0,61 et 0,60).
Les analyses exploratoires de facteur ont montre´ deux facteurs diffe´rents, correspondant a` la conceptu-
alisation de l’e´chelle. La fiabilite´ est satisfaisante avec une bonne cohe´rence interne pour a` la fois les
sous-e´chelles (l’alpha de Cronbach ⫽0,80 pour la premie`re sous-e´chelle et 0,79 pour la seconde) et pour
l’e´chelle globale (alpha⫽0,83). La proce´dure de test-retest a montre´ une bonne stabilite´ au fil du temps
(rs⫽0,90 et 0,84 pour deux e´valuateurs diffe´rents). L’e´chelle pourrait eˆtre utilise´e dans des contextes
cliniques diffe´rents, a` condition qu’un niveau suffisant de stimulation sociale soit donne´e au nourrisson
dans un pe´riode relativement bre`ve. L’e´chelle peut eˆtre utilise´e par les infirmie`res et les psychologues ou
par les docteurs, apre`s un courte pe´riode de formation.
ZUSAMMENFASSUNG: Ein dauernder Ru¨ckzug von Kleinkindern ist ein wichtiges Alarmsignal sowohl bei
organischen Krankheiten, als auch bei Beziehungssto¨rungen. Eine Bewertungsskala fu¨r Ru¨ckzug von
Kleinkindern (Alarm: Unwohlsein bei Babys–Bewertungsskala: ADBB) zwischen 2 und 24 Monaten
wurde erstellt. Diese Arbeit beschreibt den Aufbau und die Messung der psychometrischenMo¨glichkeiten
der Skala. Nach der Meinung von sieben Experten hat der ADBB eine gute Inhaltsvalidita¨t. Die Skala hat
eine gute Konstruktvalidita¨t: Zuerst als Meßinstrument der Ru¨ckzugsreaktion des Kleinkinds, mit einer

Alarm Distress Baby Scale ●561
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
sehr guten Korrelation zwischen der Krankenschwester und dem Kinderarzt im ADBB (rs ⫽0.84);
zweitens eine Reihenuntersuchung, um Entwicklungsrisiken bei Kleinkindern zu entdecken. DerGrenz-
wert von 5, mit einer Sensitivita¨t von 0.82 und einer Spezifita¨t von 0.78 wurde fu¨r die optimale Reihen-
untersuchung festgestellt. Der Fragebogen hat eine gute Konstruktvalidita¨t, mit einer guten konvergenten
Validita¨t, sowohl mit der Liste der Symptome der kleinkindlichen Depression nach Spitz (1951) und
Herzog & Rathburn (1982) (rs: 0.61 und 0.60). Bei der Exploration zeigten sich zwei Faktoren, die
passend zur Konstruktion des Fragebogens positiv geladen waren. Die Reliabilita¨t ist zufriedenstellend
mit einer guten inneren Konsistenz fu¨r beide Subskalen (Cronbachs alpha ⫽0. 80 fu¨r die erste Subskala
und 0.79 fu¨r die zweite) und fu¨r die globalen Skalen (alpha ⫽0.83). Das Test–Restest Verfahren zeigte
eine gute Stabilita¨t u¨ber die Zeitachse (rs ⫽0.90 und 0.84 fu¨r zwei verschiedene Untersucher). Der
Fragebogen kann in unterschiedlichen klinischen Anwendungen verwendet werden, vorausgesetzt, das
eine entsprechende Stimulation des Kleinkinds in einer relativ kurzen Zeit sichergestellt werden kann.
Der Fragebogen kann von Krankenschwestern, Psychologen und A
¨rzten nach einer kurzen Trainingszeit
verwendet werden.
***
The concept of withdrawal in infants remains somewhat underdefined, despiteitsfrequent
uses in clinical practice and in the process of clinical evaluation (Guedeney & Lebovici, 1997).
Brief withdrawal is a normal mechanism in the mother-infant interaction when observed at a
microanalytic level with nonclinical populations (Brazelton, Klosowski & Main, 1974; Wein-
berg & Tronick, 1994; Beebe, Lachmann & Jaffe´, 1997). On macroscopic and clinical levels,
sustained infant withdrawal is observed in several clinical situations, apart fromthe well-known
organic causes, such as fever, dehydration, postcritical state after seizure,intoxication, diseases
of the central nervous system, or visual and auditory sensory impairment (Behrman, Vaughan
& Nelson, 1983).

562 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of textThe most salient causes of withdrawal in infancy are autism and Pervasive Developmental
Disorders, where withdrawal is constant and is a key element of the diagnosis, no matter which
classification system is used. Withdrawal is a key symptom of infant depression (Spitz, 1946,
1951; Herzog & Rathbun, 1982; Guedeney, 1997, 2000). Withdrawal behavior has a strong
correlation with insecure attachment (Rubin & Lollis, 1988; Ward, Kessler & Altmann,
1993). Withdrawal is a feature of most attachment disorders (Zeanah, Boris, Baskshi & Lie-
berman, 2000). Fraiberg (1982) described the “freezing reaction,” observed among infants
when they are faced with very adverse conditions. Recently, infant withdrawal has been rec-
ognized as an element of the clinical picture of attachment disorganization(Solomon & George,
1999).
In severe forms of protein-deficiency malnutrition, such as Kwashiorkor, withdrawal is
intense and prolonged, and its abatement is a reliable sign of recovery (Guedeney, 1987, 1995,
1997). Withdrawal is also a key feature of nonorganic failure to thrive (Powell & Low, 1983;
Powell & Bettes, 1992). Menahem (1984) described two infants with possible conservation-
withdrawal reaction arising from grossly inadequate feeding and suggested that many infants
with nonorganic failure to thrive (NOFTT) can be understood better within the context of
conservation-withdrawal. Engel and Schmale (1972) developed the concept of Conservation-
Withdrawal as “a biological threshold mechanism where survival of the organism is supported
by process of disengagement and inactivity, vis-a-vis the external environment” (p.75). Engel
& Reichsman (1956, 1979) described the case of Monica, who presented with a severe state
of withdrawal during a long hospitalization, and they followed her for 25 years.
Infant withdrawal is a key element of the infant’s response in the face of of a noncontingent
kind of relationship. This can be observed in the “still-face” experiment (Cohn & Tronick,1983;
Field, 1984; Seifer & Dickstein, 1993; Tronick & Weinberg, 1997), in studies of the effect of
maternal depression (Murray & Cooper, 1997), or in parents who have other kinds of mental
impairment (Reder & Lucey, 1995). Severe withdrawal can be observed also at differentlevels
in infants with Anxiety Disorders, Post Traumatic Stress Disorders (Zeanah, 1993), and in
children who have suffered from chronic and severe pain in infancy (Gauvain-Piquard, Rodary,
Rezvani & Lemerle, 1987; Gauvain-Piquard, Rezvani, Rodary & Serbouti, 1999).
Although withdrawal is a very important alarm signal (Ironside, 1975), no assessment tool
is available to date, to our knowledge. This article describes the constructionand the validation
of a scale, the Alarm Distress Baby Scale (ADBB), designed to assess sustained withdrawal
reactions in infants aged between 2– 24 months.
METHOD
Subjects
Inclusion criteria. All infants male or female, aged 2–24 months, of any national or ethnic
background, seen for an initial follow-up visit at the Well-Baby Clinic (PMI) by F. Roge´,
pediatrician, and by M. Vermillard, infant nurse, at the Institut de Pue´riculture de Paris between
July 1996 and July 1997 were eligible for the study. Only new patients were eligible, so that
both the pediatrician and the nurse would not be influenced by any previous knowledge of the
familial and medical backgrounds of the child. The upper age limit of 24 months was picked
to demonstrate that sustained withdrawal reaction could be assessed during a period of rapid
and dramatic developmental growth and before the use of language. The lower age limit of 2
months was chosen after reviewing the literature (Cioni et al.,1977; Prechtl et al., 1997), taking
into account the duration of the fetal period and its influence after birth.

Alarm Distress Baby Scale ●563
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of textExclusion criteria. We excluded subjects from the study as follows:
1. Infants with any acute disease that could be in itself responsible for a state of withdrawal
or prostration, such as fever (above 37.5 ⬚C or 99.5 ⬚F), dehydration, recent seizures,
intoxication, acute otitis media, or any cause of severe acute or chronic pain;
2. Infants with acute and chronic neurological problems, and infants with known visual
or auditory impairment;
3. Premature and small for date babies ( ⬎37 weeks gestational age, and infants whose
birth weight was below the third standard deviation [SD] on growth charts); and
4. Infants already followed at the PMI.
Procedure. The consultation was videotaped with a VHS camera, which focused only on the
infant. Written informed consent was required from the parents just before the consultation.
All infants with clear withdrawal behavior were followed by the PMI team, and were referred,
if necessary, to our nearby infant guidance clinic for evaluation and treatment. The research
protocol was approved by the Ethics Committee (Faculte´deMe´decine Necker-Enfants Mala-
des, Paris).
Instrument Construction
The main idea in designing the scale was to build an instrument simple enough to be used
effectively by nurses in clinical practice. The medical consultation is used in the same way as
the Winnicott “set situation” (Winnicott, 1941), which provides a given stimulation and ob-
serves the way the infant makes uses of it. .
The scale was built in three steps. The 14-items scale by Powell & Low (1983) was taken
as a starting point. Their study defining behavior that differentiated between organic and non-
organic failure to thrive (FTT) showed that abnormal “interpersonal” behaviors were more
common than “non-interpersonal” behaviors in hospitalized infants with nonorganic failure to
thrive. The Powell & Low eight noninterpersonal items were as follows: general inactivity,
excessive crying and irritability, flexed hips and knees, expressionless face, infantileposturing,
rumination, excessive thumb sucking, and disproportionate hand and finger inactivity. Their
six interpersonal items were as follows: crying when approached, lack of decreased vocaliza-
tion, lack of cuddliness, poor eye contact, lack of response, and indifference to separation.
However, the Powell & Low study was made with only 21 hospitalized infants, with novalidity
or reliability studies, and some of the items in their scale (for example infantile posturing, or
flexed hips and knee) were observed with only severely distressed infants. Therefore, a first
draft of the ADBB was made with only 10 items that corresponded more appropriately with
the interpersonal and non interpersonal dimensions of the withdrawal behavior: facial expres-
sion, eye contact, general activity, self-stimulating gestures, vocalizations, response to stimu-
lation, relationship to observer, ability to attract attention, reaction to cuddling, andreactionto
separation. After the items were examined for content validity by four experts (A. Gauvain-
Piquard, S. Lebovici, M. Soule´, D. Widlo¨cher), two items, ‘reaction to cuddling’; and ‘reaction
to separation,’ were removed. The 8-item version of the scale then was tested on 29 infants
who had very different kinds of clinical conditions. Taking into account the difficulties en-
countered during these initial assessments, we modified the scale following the advice of five
experts (R. Brunod, R. N. Emde, H. Fitzgerald, M. Soul, Ch. H. Zeanah) to clarify the item
descriptions.
This provisional scale was tested on 40 infants, each of whom was rated simultaneously

564 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of textby a pediatrician, a nurse, and Antoine Guedeney, the first author of this article. The distribution
of the total scores were examined, various factor analyses were performed, and interrater re-
liability was assessed by using statistical methods described below. Taking into account the
results of these analyses, we made improvements in the formulation of some of the items,
which lead to the present and final version of the scale (see appendix). The final version of the
scale was translated into English (Annie Nataf-Cooper, M.D.), then blindly retranslated into
French (Marianne Kumar).
Assessment of Psychometric Properties
The psychometric properties of the French version of the scale were assessed in the following
manner (American Psychological Association, 1985):
Validity
Face validity. The meaning and acceptability of the items were investigated by theraters
during the administration of the questionnaire.
Content validity. Seven experts on infant psychopathology, infant depression, and affect
disorders in infancy who had experience in designing and validating assessment scales (R.
Brunod, R. N. Emde, A. Gauvain-Piquard, S. Lebovici, M. Soule´, D. Widlo¨cher, Ch. H.
Zeanah) checked independently if the items of the scale did correspond with the two hypoth-
esized interpersonal and noninterpersonal dimensions of the withdrawal reaction.
Criterion-related validity. Criterion-related validity was investigated in two ways.
First, the ADBB was used as a measure of the intensity of the withdrawal reaction. Each
of the infants was rated simultaneously by the first author (A.G.), considered an expert, using
a visual analog scale (VAS), by the pediatric nurse of the PMI, and by one of the consulting
pediatrician of the PMI, who both used the ADBB. The expert, the pediatrician, and thenurse
filled out the scales independently and simultaneously immediately after each medical exam-
ination. Considered the gold standard, the VAS ranged from 0 (absence of withdrawal) to 100
(severe withdrawal). The correlation coefficients between the expert’s VAS scores and, re-
spectively, the nurse’s scores and the pediatrician’s scores assessed the concurrent validity of
the ADBB.
Second, the ADBB was used as a screening tool, through the evaluation of its sensitivity
and specificity, for detecting the developmental risk for the infant in the following year (pre-
dictive validity). The developmental tests available and validated on the French population
have no predictive value for development (Josse, 1979). The developmental risk was, therefore,
rated high or low by the first author on the basis of the 17 items (see Appendix A), which were
systematically collected for each infant by the PMI team (nurse, pediatrician, head nurse) since
its participation in an epidemiological study in 1982 that aimed to determine the developmental
risk factors in infancy (Choquet, Facy, Laurent & Davidson, 1982). In their study, a sample of
415 infants representative of the 14 districts of Paris was followed for 3 years. Four question-
naires describing children’s health and behavior were completed at 3, 9, 18, and 36 months.
First, a cluster analysis was carried out on the collected variables, and three different groups
of children were identified (high, middle, and low risk). Ten percent (10‰) of the children
belonged in the high risk group and were characterized by 6 of the 17 items listedin Appendix
A. Second, a follow-up study showed that these criteria were reliablein discriminating between
risk infants and others (Choquet & Ledoux, 1985).
In our study, we considered only two categories of infants, those classified as high risk if
they had, as in the preceding study, 6 or more of the 17 items listed Appendix A and those
classified as low-risk group if they had 5 or less of these 17 items.

Alarm Distress Baby Scale ●565
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
TABLE 1. Spitz and Herzog & Rathbun’s Lists of Symptoms of Infant Depression
Spitz’s Criteria for Anaclitic Depression Herzog & Rathbun Developmentally Specific
Criteria for Infant Depression (0–36 months)
Increased demandingness Failure to thrive or rumination
Weepiness Lack of social play
Loss of weight Persistent lethargy or irritability
Arrest of developmental progress Separation or attachment problems
Insomnia Agitation, aggression
Rigid facial expression Developmental delays
Restriction of mobility Sleep disorders
Withdrawal Feeding difficulties
In this study, sensitivity is, thus, defined as the proportion of infants classified aswithdrawn
by the ADBB (score equal or above a cutoff point) from the total number of infants classified
as having a high developmental risk. Specificity is defined as the proportion of infants classified
as not withdrawn by the ADBB (score below the fixed cutoff point) from the total number of
infants classified as having a low developmental risk. The nurse’s total scores with the ADBB
on live examinations were used to evaluate sensitivity and specificity.
Construct validity. There were several components to construct validity:
1. Convergent validity was assessed by the correlation coefficients between the nurse’s
total scores with the ADBB and each of the two variables relating to withdrawal and
representing the severity of the state of depression in the infant. The first variable was
the number of pathologic events during the year after the assessment with the ADBB
and belonging to the list of symptoms given by Spitz (1951) in his description of
anaclitic depression. The second variable was the number of pathologic events, calcu-
lated as above, listed in the Herzog & Rathbun (1982) description of infant depression.
Table 1 shows the list of symptoms or behaviors present in each description. After
carefully checking with the pediatrician, we included in the calculationof thevariables
each pathologic event reported in the PMI medical record of the infant.
2. Discriminant validity was assessed through the correlation coefficient between the
nurse’s ADBB total scores and several variables thought to have no relationship with
withdrawal: age of the mother, parity, age of the father, age of the infant, birth rank,
and duration of the well-baby clinic consultation.
3. Exploratory factor analysis was made in order to investigate the internal structure of
the scale, as described in the statistical methods section.
Reliability. This was tested in three ways.
1. The interrater reliability between the nurse and the pediatrician was calculated on the
basis of their live assessments, made immediately after each consultation and without
discussing it (see criterion-related validity).
2. Two psychology students (H. Courtier and C. Dumond) were trained in the use of the
ADBB, through live examinations at PMI and through videotaped sessions, using the
documents recorded during the pilot study described above. Then, they scored blindly

566 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of textand independently the whole set of 60 videotapes of the present study, in random order,
in order to measure their interrater reliability on the videotaped documents.
Six months later, the two psychology students did the same assessment again,
independently, and again in random order, to provide an evaluation of the test-retest
reliability with videotaped documents. The interrater and test-retest reliabilities were
assessed using the intraclass correlation coefficient as described below.
3. The internal consistency was assessed first by using the Cronbach
␣
coefficient (Cron-
bach, 1951), considering the ADBB as a global scale. Taking the results of the factor
analysis into account, the assessment of the internal consistency then was made sepa-
rately for each of the two subscales (see below).
Statistical methods. The quantitative variables were described using means, standard devia-
tions [SD], and ranges. Qualitative variables were described using frequenciesand percentages.
The relationship between two quantitative variables was assessed using the Spearman corre-
lation coefficient [r] because the binormality of the sampled distribution could not be assumed.
s
It was tested if each correlation coefficient differed significantly from zero. As expected, cor-
relation coefficients assessing criterion and convergent validity were highly significant,whereas
all those measuring discriminant validity were nonsignificant. To simplify the presentation of
the results, we did not report the pvalues of the previous tests in the text. The interrater and
the test-retest reliability were assessed using the intraclass correlation coefficient [ICC] under
the random-effect model. After we estimated the components of total variance by analysis of
variance, we calculated the ICC in the usual manner (Shrout & Fleiss, 1979). Factor analysis
of the scale was performed using the principal component analysis as the methodof extraction.
The Cattell screen test was used for determining the number of factors extracted.Independent
factors were obtained using the Varimax rotation method. All the statistical tests were two-
tailed, and their level of significance was
␣
⫽0.05. Statview II娀(SAS Institute) was the
statistical software used.
RESULTS
Description of Subjects
One family out of 61 refused to take part in the study because the person accompanying the
child was not a parent. Sixty infants were finally included. The mean age of infants was 30.5
weeks, [SD: 21.51 wks; range: 8– 96 wks]. The mean age of mothers was 30.1 years [SD: 5.68
yrs; range: 20–44 yrs]. Mean age of fathers was 35.2 years [SD: 9.96 yrs; range: 23–67 yrs].
Following criteria listed in Appendix A (see above), the proportion of high developmental risk
infants was 11 out of 60 (18.3‰). The sociodemographic and obstetric characteristics are
summarized Table 2. Socioeconomic status (SES) was assessed following French Parisian
Inserm criteria (Inserm, 1975). Nurse’s total ADBB score had a mean value of 3.4 [SD: 3.78;
range: 0–17]. Figure 1 shows the distribution of the nurse’s total scores.
Psychometric Properties of the ADBB
Validity
Face validity. All items of the scale were judged acceptable, easy to understand, and
easy to fill out by the raters, the nurses, the pediatrician, and the psychology students.

Alarm Distress Baby Scale ●567
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
TABLE 2. Sociodemographic and Obstetric Characteristics of the Studied Subjects (N
⫽
60)
Variables N %
Girls 28 46.66
Sex Boys 32 53.33
Very high 3 5.00
High 8 13.33
Socioeconomic status (SES) Middle 24 40.00
Low 19 31.66
Very low 6 10.00
French metropolitan 35 58.33
French overseas departments 2 3.33
Geographic origin of the mother North Africa 7 11.66
Africa 6 10.00
Asia 5 8.33
Other 5 8.33
French metropolitan 32 53.33
French overseas departments 1 1.66
North Africa 9 15.00
Geographic origin of the father Africa 8 13.34
Asia 7 11.67
Other 3 5.00
Parity 1 child 32 53.33
2 children 17 28.34
3 children 5 8.33
⬎3 children 6 10
First born 49 81.67
Rank of birth Second 7 11.66
Third or later 4 6.67
F
IGURE
1.1. Distribution of the nurse’s ADBB total scores.

568 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of text
TABLE 3. Specificity and Sensitivity of the ADBB
for Different Cutoff Scores
Cutoff
scores 5678
Sensitivity 0.82 0.64 0.55 0.45
Specificity 0.78 0.80 0.86 0.96
Content validity. The seven experts independently judged that each item corresponded
to a postulated dimension and that each dimension was represented by several items. Consid-
ering that the main characteristics of the universe of items were reproduced in the proper
proportions and balance, the experts concluded independently that the content validity was
good.
Criterion validity. Criterion validity was investigated in two situations. First (concurrent
validity) as a measure of the severity of the infant’s withdrawal reaction, the expert’s VAS
scores were well correlated with the nurse’s scores on the ADBB [r⫽0.67] and with the
s
pediatrician’s score [r⫽0.63], respectively.
s
Second (predictive validity), the ADBB was used as a screening procedure for detecting
the developmental risk of the infant in the year after the assessment. The sensitivityand spec-
ificity of the scale for different cutoff scores are displayed Table 3.
The cutoff score of 5 with a sensitivity of 0.82 and a specificity of 0.78 was determined to be
optimal for screening purposes.
Construct validity. We tested three measures of construct validity.
1. Convergent validity. The correlation coefficients between the nurse’s score with the
ADBB and the number of pathologic events, different or not, during the year after the
assessment were calculated separately with the two available descriptions of infant
depression and are, respectively, r⫽0.61 for Spitz anaclitic depression (1951) and
s
r⫽0.60 for Herzog & Rathbun (1982) criteria.
s
2. Discriminant validity. As expected, no significant correlation was found between the
total score on the ADBB by the nurse and the age of the mother [r⫽0.048], parity
s
[r⫽0.175], age of the father [r⫽0.020], age of the infant [r⫽0.225], rank of birth
sss
[r⫽0.175], or duration of the consultation [r⫽0.212].
ss
3. Factor analysis extracted two orthogonal factors which accounted for 63.3‰ of the
variance. The factor loading for each item is shown in Table 4. The first factor (FI) has
5 items: 2 (eye contact), 3 (general level of activity), 4 (self-stimulating gestures), 7
(relationship), and 8 (attractivity). The second factor (FII) has 3 items: 1 (facial ex-
pression), 5 (vocalization), and 6 (response to stimulation). The loading of item 6
(response to stimulation) was close for FI and FII; however, this item was classifiedas
FII, after we took into account its clinical meaning (Harman, 1967).
Reliability
1. The interrater reliability between the nurse and the pediatrician during the live assess-
ments of the 60 infants with the ADBB was very good (ICC ⫽0.84).

Alarm Distress Baby Scale ●569
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
TABLE 4. Results of Factor Analysis; Loading 5
⬍
0.50 Are
Not Reported
Factor 1 Factor 2
Eigenvalues 3.882 1.201
Percentage of total variance explained 48.5 15.1
Factor loading
item 1 .841
item 2 .750
item 3 .566
item 4 .538
item 5 .902
item 6 .532 .657
item 7 .774
item 8 .761
2. The interrater reliability between the two psychology students on the 60 videotaped
examinations was also good for both the first assessment (ICC ⫽0.88) and for the
second assessment, performed 6 months later (ICC ⫽0.87). The ICC coefficients that
assessed the test-retest reliability on videotaped documents were, respectively, 0.90 for
the first psychology student and 0.84 for the second.
3. When the ADBB is considered as a global scale, the internal consistency assessed by
the Cronbach
␣
is 0.83. When we took into account the two-factor structure of the
ADBB, the Cronbach
␣
was 0.80 for the first subscale and 0.79 for the second subscale.
DISCUSSION
Psychometric Properties
Validity
Face and content validity
1. During this validation study, all the users of the ADBB, especially nurses and psy-
chology students, found that the scale had good face validity and was easy to fill out
in a short period of time (5 minutes maximum). This was confirmed during all the
training sessions performed later, after the completion of the present study.
2. The seven experts who assessed the content validity were all experienced clinicians
specializing in various fields of infant psychopathology, such as depression, pain, or
attachment regulatory disorders or post traumatic stress disorders. Their independent
judgements were all in close agreement and resulted in a good content validity.
Criterion validity. Because of his clinical experience with infants who exhibited withdrawal
behavior (Guedeney, 1987, 1995, 1997, 2000), A. Guedeney was considered an expert. His
assessments of the severity of the withdrawal through the use of a VAS and his judgement on
the further developmental risk for the infant were considered as references for calculating

570 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of textconcurrent and predictive validity. We found these two types of criterion validity satisfactory.
This suggests that the ADBB can easily be used in clinical practice, especially by nonmedical
users, both for assessing the intensity of withdrawal and for detecting with good sensibility
and specificity the developmental risk of the withdrawn infant.
Construct validity
1. Convergent validity
Several authors have pointed out that withdrawal behavior is related to infant depression
(Spitz, 1946, 1951; Herzog & Rathbun, 1982; Powell & Bettes, 1992). Although it is
commonly used in clinical practice, the concept of infant depression rarely is defined
accurately. Only Spitz (1946, 1951) and Herzog & Rathbun (1982) have, to our knowl-
edge, published a list of symptoms of depression in infants. Using each of the two lists
of symptoms of infant depression, we calculated the number of pathologic events in
the year after the assessment with the ADBB. Each event, notable enough to be reported
in the PMI medical file, was considered ‘pathologic’ after we carefully checked with
the pediatrician in charge of the infant. We, therefore, obtained an objective and valid
measure of the severity of the state of depression of the child and responded with the
usual clinical approach. We found that withdrawal behavior was related to depression
as defined by each list (r⫽0.61 and r⫽0.60, for the Spitz and Herzog & Rathbun
ss
lists of symptoms, respectively). However, this also indicates that the ADBB scale is
not identical to a depression scale.
2. Factor analysis
The observed internal structure of the scale with five items loading on FI (interpersonal
dimension) and three items loading on FII (noninterpersonal dimension) is consistent
with the two-factor structure of the construct underlying the construction of the scale
(see Methods). This structure should, however, be confirmed by further studies with a
greater number of infants.
Taking into account both the results of the factor analysis and the good convergent and
discriminant validity, we consider that the construct validity is satisfactory (Coste, Fermanian
& Venot, 1995).
Reliability. The good interrater reliability has been achieved with in-person and videotaped
training. Moreover, the good ICC coefficient. (0.84) between the nurse andpediatricianshows
that the scale is effective in enabling the pediatrician and the nurse to have a closerappreciation
of the withdrawal behavior of the infant. The test-retest coefficients show good stability of
rating over time. When the ADBB is considered as a global scale, the internal consistency
assessed by the Cronbach
␣
(0.831) demonstrates that the items are interrelated well. When
the two-factor model is retained, the values of the Cronbach
␣
for each subscale (0.80 for FI,
0.79 for FII) show that each has good internal consistency without too much redundancy of
the items.
In conclusion, this eight-items scale has good content and criterion validity. Its construct
validity is satisfactory. The interrater and test-retest reliability is very good, both for live and
on video assessments. The ADBB can be used for measuring the severity of withdrawal
behavior or as a screening tool for detecting further developmental risk with the with-
drawn infant. After a short period of training, a medical doctor or nonmedical professionals,
especially nurses or psychologists, can perform the assessment.The scale should facilitate the
recognition and the evaluation of withdrawal behavior in infants who are seen in common
clinical practice.

Alarm Distress Baby Scale ●571
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
APPENDIX A
Criteria used to determine infants with high developmental risk (Choquet, Facy, Laurent, &
Davidson, 1982). At least 6 out of 17 are needed for a child to belong in the high risk devel-
opmental group.
㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮㛮
-Wakes up frequently at night
-Disturbed sleep pattern
-Often angry
-Often dissatisfied
-Receives sedative drugs
-Hospitalized for more than 3 days between the age of 18 months and 3 years
-Poor health–according to the mother
-Has asthma
-Several accidents between 18 months and 3 years of age
-Has nightmares
-Erratic appetite
-Frequent rhinitis
-Going to bed and getting asleep is difficult
-Frequent bronchitis
-Frequent otitis media
-No thumb sucking, or substitute
-Does not play alone
APPENDIX B
Alarm Distress BaBy Scale (ADBB)
Assessment of withdrawal behavior in infancy
Each item is rated on a scale from 0 to 4:
0: No unusual behavior
1: Doubt as to unusual behavior
2: Mild unusual behavior
3: Clear unusual behavior
4: Severe unusual behavior
This scale is best rated by the observer on the basis of her/his observations, immediately
following the clinical interview. Initially, spontaneous behavior is assessed, then following
stimulation (smile, voice, gesture, touch, etc.), and the evolution along time. The rating is what
seems more significant during the whole examination procedure. In case ofdoubt betweentwo
ratings return to the definition above.
1. FACIAL EXPRESSION: Observer assesses any reduction of facial expressiveness:
0: Face is spontaneously mobile, expressive, animated.
1: Face is mobile, expressive, but limited in range.
2: Little spontaneous facial mobility is exhibited.
3: Face is fixed, sad.
4: Face is fixed, frozen, absent, appearing prematurely old.

572 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of text2. EYE CONTACT: Observer assesses the reduction of eye contact:
0: Eye contact is spontaneous, easy and sustained.
1: Eye contact is brief and spontaneous.
2: Eye contact is possible only when initiated by observer.
3: Eye contact is fleeting, vague, elusive.
4: Eye contact is totally avoided.
3. GENERAL LEVEL OF ACTIVITY: Observer assesses any failure of motion of the
head, torso, and limb without taking into account hands and fingers activity:
0: Frequent and well coordinated, spontaneous head, torso and limb motions.
1: Reduced general level of activity, few head and limb movements.
2: No spontaneous activity, but reasonable level in response to stimulation.
3: Low level of activity in response to stimulation.
4: Immobile, rigid, stiff, whatever the stimulation.
4. SELF-STIMULATING GESTURES: Observer assesses the frequency with which the
child plays with his/her own body: fingers, hand, hair, thumb sucking, repetitive rubbing
etc., in relation to the general level of activity. One significant sign of self-stimulating
gesture is sufficient to rate a 1 or higher.
0: Absence of self-stimulation; autoexploration is appropriate to the level of general
activity.
1: Self-stimulation occurs fleetingly.
2: Self-stimulation is rare but obvious.
3: Self-stimulation is frequent.
4: Self-stimulation is constant.
5. VOCALIZATIONS: Observer assesses the lack of vocalization expressing pleasure
(cooing, laughing, babbling, babbling with consonant sounds, squealing with pleasure),
but also lack of vocalization expressing displeasure or pain (screaming or crying):
0: Frequent, cheerful, modulated spontaneous vocalizations; brief crying or screaming
in response to an unpleasant stimulation or sensation.
1: Brief spontaneous vocalizations, frequent screaming or crying (even if only in re-
sponse to stimulation).
2: Constant crying.
3: Whimpering only in response to stimulation.
4: Absence of vocalization.
6. BRISKNESS OF RESPONSE TO STIMULATION: Observer assesses the sluggish-
ness of response to pleasant or unpleasant stimulation during the examination (smile,
voice, touch). The amount of response is not being assessed here, but the delay in
response.
0: Appropriate, brisk and swift response to stimulation.
1: Slightly delayed and sluggish response to stimulation.
2: Sluggish, delayed response to stimulation.
3: Markedly sluggish response to even unpleasant stimulation.
4: Very delayed response to stimulation, or absence of any response to stimulation.
7. RELATIONSHIP: Observer assesses the infant’s ability to engage in a relationship
with him/her, or with anyone present in the room, other than his/her caretaker. Rela-
tionship is assessed through attitude, visual contact, reaction to stimulation,and possible
reaction to the end of the examination session.
0: Relationship clearly and quickly established, rather positive (after a possible initial
phase of anxiety).

Alarm Distress Baby Scale ●573
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text1: Relationship identifiable, positive or negative, but less than described in 0.
2: Relationship mildly evident, delayed, positive or negative.
3: Doubt as to the existence of a relationship.
4: Absence of identifiable relationship to others.
8. ATTRACTION: The effort needed by the observer to keep in touch with the child is
assessed here, along with the pleasure initiated by the contact with the child:
0: The child attracts attention through his/her initiative and contact, generatinga feeling
of interest and enjoyment.
1: There is interest towards the child, but without less pleasure than as described in 0.
2: Neutral feelings towards the child, possibly with a tendency to forget to focus on
the child.
3: Uneasy feeling towards the child, feeling of being maintained at a distance.
4: Disturbing feeling with the child, impression of a child beyond reach.
TOTAL:
LAST NAME: FIRST NAME:
DATE:/// AGE: / / MONTHS / / DAYS
EXAMINER
REFERENCES
American Psychological Association. (1985). Standards for educational and psychological testing. Wash-
ington, DC: Author.
Beebe, B., Lachmann, F., & Jaffe, J. (1997). Mother-infant structures and pre-symbolic self and object
representations. Psychoanalytic Dialogues, 2, 133–182.
Behrman, R.E., Vaughan, V.C., & Nelson, W.E. (Eds.). (1983). Nelson textbook of pediatrics (13thed.).
Philadelphia: W.B. Saunders.
Brazelton, T.B., Koslowski, B., & Main, M. (1974). Origins of reciprocity. In M. Lewis & L. Rosenblum
(Eds.), Mother infant interaction (pp. 57–70). New York: John Wiley & Sons.
Cioni, G., Ferrari, F., Einspieler, CH., Paocelli, P., Barbani, M.T., & Prechtl, H.F.R. (1997). Comparison
between observation of spontaneous movements and neurologic examination in pre-term infants.
Journal of Pediatrics, 130, 704–711.
Choquet, M., Facy, F., Laurent, F.,& Davidson, F. (1982). Discovery of ‘risk’ groups in a population of
children of pre-school age by a method of typological analysis. Early Child Development and Care,
9, 255–274.
Choquet, M., & Ledoux, S. (1985). La valeur pronostique des indicateurs de risque pre´coces. Etude
longitudinale des enfants a` risque a` 3 ans. Archives Franc¸aises de Pe´diatrie, 42, 541–546.
Cohn, J.F., & Tronick, E.Z. (1983). Three months old infant’s reaction to simulated maternal depression.
Child Development, 54, 334–235.
Coste, J., Fermanian, J., & Venot, A. (1995). Methodological and statistical problems inthe construction
of composite measurement scales: a survey of six medical and epidemiological journals. Statistics
in Medicine, 14, 331–345.
Cronbach, L.S. (1951). Coefficients alpha and internal structure of tests. Psychometrica, 16, 297–334.
Engel, G.L., & Reichsman, F. (1956). Spontaneous and experimentally induced depression in an infant
with gastric fistula: a contribution to the problem of depression. Journal of the American Psycho-
analytical Association, 4, 428– 452.

574 ●A. Guedeney and J. Fermanian
IMHJ (Wiley) LEFT BATCH
short
standard
top of rh
base of rh
cap height
base of text
Engel, G.L., & Reichsman, F. (1979). Monica: a 25 years follow-up longitudinal study of the conse-
quences of trauma in infancy. Journal of the American Psychoanalytical Association, 27, 107–126.
Engel, G.L., & Schmale, A.H. (1972). Conservation withdrawal: a primary regulatory process for orga-
nismic homeostatas. In Physiology, emotion and psychosomatic illness, CIBA Foundation Sympo-
sium 8 (new series) (pp. 57–85). Amsterdam: Excerpa Medica.
Field, T. (1984). Early interactions between infants and their postpartum depressed mothers. Infant Be-
havior and Development, 7, 537–532.
Fraiberg, S. (1982). Pathological defenses in infancy. The Psychoanalytical Quarterly, 4, 612–635.
Gauvain-Piquard, A., Rodary, Ch., Rezvani, A., & Lemerle, J. (1987). Pain in children aged 2–6 years:
a new observational rating scale elaborated in a pediatric oncology unit. Pain, 31, 177–188.
Gauvain-Piquard, A., Rodary, C., Rezvani, A., & Serbouti, S. (1999). The development of the DEGR威:
A scale to assess pain in young children with cancer. European Journal of Pain, 3, 165–176.
Guedeney, A. (1987). Les aspects psychosomatiques des malnutritions Prote´ino-caloriques de la premie`re
enfance. La Psychiatrie de l’Enfant, 1, 155–190.
Guedeney, A. (1995). Kwashiorkor, depression and attachment disorders. The Lancet, 8985 (34), 1293.
Guedeney, A. (1997). From early withdrawal reaction to infant depression. A baby alone does exist.
Infant Mental Health Journal, 18(4), 339–349.
Guedeney, A. (2000). Infant depression and withdrawal: clinical assessment. In J. D. Osofsky & H. E.
Fitzgerald (Eds.), WHAIM handbook of infant mental health, Vol. 4 (pp. 455–484). New York:
John Wiley & Sons.
Harman, H. H. (1967). Modern factor analysis. Chicago: University of Chicago Press.
Herzog, D.B., & Rathbun, J.M. (1982). Childhood depression: developmental considerations. American
Journal of Diseases of Childhood, 13, 115–120.
Inserm (1975). Classification socio-professionnelles, Documentation Inserm. Paris: Author.
Ironside, W. (1975). The infant development distress (I.D.D.) syndrome: a predictor of impaired devel-
opment? Australian and New Zealand Journal of Psychiatry, 9, 153–158.
Josse, D. (1979). Brunet-Le´zine re´vise´. Issy-les Moulineaux: Editions et Applications Psychologiques.
Kendall, M.G. (1975). Rank correlation methods. London: Griffin & Co.
Lourie, R.S., & Nover, R.A. (1980). Applied psychoanalysis and assessment of psychopathology in the
first year of life (pp. 365–379). In S.I. Greenspan & G.H. Pollock (Eds.). The course of life: psy-
choanalytic contributions toward understanding personality development: Vol. 1. Infancy and early
childhood. Bethesda: NIMH.
Menahem, S. (1984). Possible conservation withdrawal reaction in two infants. Developmental and Be-
havioral Pediatrics, 5, 361–363.
Murray, L., & Cooper, P.J. (1997). Postpartum depression and child development. New York: Guilford.
Powell, G.F., & Low, J. (1983). Behavior in nonorganic failure to thrive. Developmental andBehavioral
Pediatrics, 4, 1, 26–33.
Powell, G.F., & Bettes, B.A. ( 1992). Infantile depression, nonorganic failure to thrive and DSM-II-R:
A different perspective. Child Psychiatry and Human Development, 22(3), 185–198.
Prechtl, H.F.R., Einspieler, Ch., Cioni, G., Bos, A.F., Ferrari, F., & Sontheimer, D. (1997). An early
marker for neurological deficits after perinatal brain lesions. The Lancet, 349(10), 1361–1363.
Reder, P., & Lucey, C. (1995). Assessment of parenting. London: Routledge.
Rosenn, D.W., Stein Loeb, L., & Bates M. (1980). Differentiation of organic from non organicfailure to
thrive syndromes in infancy. Pediatrics, 66, 698–704.
Rubin, K.H., & Lollis, S.P. (1988). Origins and consequences of social withdrawal. In J. Belsky & T.
Nezworski (Eds.), Clinical implications of attachment (pp. 219–252). Hillsdale, NJ: Erlbaum.
SAS Institute. Statview II [Computer software]. Carry, NC: Author.

Alarm Distress Baby Scale ●575
IMHJ (Wiley) RIGHT BATCH
short
standard
top of rh
base of rh
cap height
base of text
Seifer, R., & Dickstein, S. (1993). Parental mental illness and infant development. In C. H.Zeanah (Ed.),
Handbook of infant mental health (pp.120–142). New York: Guilford Press.
Shrout, D.E., & Fleiss, J.L. (1979). Intra-class correlation: uses in assessing raterreliability.Psychology
Bulletin, 86, 420–428.
Spitz, R.A. (1946). Anaclitic depression. Psychoanalytical Study of the Child, 2, 313–341.
Spitz, R.A. (1951). The psychogenetic diseases in infancy: an attempt at their etiologic classification. The
Psychoanalytic Study of the Child, 6, 255–275.
Tronick, E.Z., & Weinberg, M.K. (1997). Depressed mothers and infants: failure to form dyadic states
of consciousness. In L. Murray & P. Cooper (Eds.), Postpartum depression and child development
(pp. 54– 84). New York: The Guilford Press.
Ward, M.J., Kessler, D.B. & Altmann, S.C. (1993). Infant-mother attachment in children with failure to
thrive. Infant Mental Health Journal, 14(3), 208–220.
Weinberg, M.K., & Tronick, E.Z. (1994). Beyond the face: An empirical study of infant affective con-
figurations of facial, vocal, gestural and regulatory behaviors. Child Development, 65(5), 1503–
1515.
Winnicott, D.W. (1941). Observation of infants in a set situation. International Journal of Psychoanalysis,
22, 229–249.