The ADBB scale : an overview

Social withdrawal in infancy

It has been well documented that infants, from just a few hours old, demonstrate the ability to socially interact with adults. From a few months old infants’ social behaviour can be observed in a variety of ways, including vocalising, eye contact with the observer, imitating physical gestures, showing a range of facial expressions, and responding to the adult. Using facial expressions and vocalisations as indicators of feeling states, most infants show three emotions by two months of age (interest, contentment and distress), and eight emotions by seven months (joy, contentment, anger, disgust, surprise, interest, distress and sadness).

Smiling in response to a person’s voice and face occurs by eight weeks old, while anxiety towards strangers is generally considered to be evident by around nine months of age.
Such anxiety when held by a stranger will usually produce active behaviour, such as vocal protestations and squirming, in an attempt to get closer to the infant’s parent.

Withdrawn social behaviour from just two months old, indicated by a lack of either positive (e.g., smiling, eye contact) or negative (e.g., vocal protestations) behaviours is more akin to a state of learned helplessness, and should alert the clinician to the possibility that the infant is not displaying age-appropriate emotional/social behaviour.

Such low infant sociability can be due to many factors, including both organic and non-organic disorders (Guedeney & Fermanian, 2001). In 70% of infants with nonorganic failure to thrive, where there has been considerable emotional or social understimulation, Powell and Low (1983) found that such infants had poor eye contact, diminished vocalization and a lack of response to an adult. Tronick and Weinberg (1997) have reviewed how maternal depression, which can result in either an intrusive or withdrawn parenting style, can also affect the social behaviour of the infant.
Such infants have less eye contact and more self-comforting behaviours (self-stimulating gestures).
This connection between depressed mood in the caretaker (usually the mother) and either withdrawn behaviour in the infant, or impaired mother-infant interaction, has been found in a number of studies.
Of note is that such withdrawn infant behaviour has been found to be associated with decreased frontal EEG activity, which is present when the infant is interacting with his mother and also when interacting with other adults (Dawson et al., 1999).

This would suggest that there is some carry-over effect - that the infant, exposed to prolonged periods of interacting with an unresponsive caregiver, has become ‘depressed’ and continues to exhibit withdrawn behaviour even when someone else interacts in a positive way with him (for the sake of clarity the infant will be referred to as a male).
The diagnostic classificatory system for infants and toddlers, DC: 0-3 (Zero to Three, 1994), has recognized the importance of impaired affect in infants by including a category of ‘Depression in Infancy and Early Childhood’.

Field (1995) states that "Infants who interact with depressed mothers (particularly chronically depressed mothers) are at risk for later social interaction problems" (p.1).
It is important to note, however, that depression in the postpartum period does not necessarily mean the mother-infant interaction will be impaired, nor should one presume that all impaired interactions will necessarily lead to withdrawn infant behaviour.

Undoubtedly there will be buffers to such effects, such as the role of fathers and other caregivers and possibly the temperament of the infant.

Given these caveats, there is an increasing realisation that assessing an infant’s social behaviour is important.
While diminished social behaviour in the infant may not necessarily be an indicator of pathology, it should alert the clinician to undertake further assessment of both the infant and the environment.
In some cases it may reveal that the mother is experiencing psychosocial difficulties (such as postnatal depression or bonding difficulties) which she may have been reluctant to disclose.
Assessment of the infant’s social behaviour

Matthey et al (in press) state :

"While measurement of the infant’s temperament may have some overlap with his social behaviour, it is important to realise that these two constructs are separate.
Temperament refers to the infant’s degree and style of responsiveness to varying internal and external stimuli (e.g., noise, heat, as well as social stimuli), whereas social behaviour in infancy refers to degree and style of responsiveness just to social stimuli.
Thus while an infant may, within temperament measures, be considered ‘shy’ or ‘slow to warm up to others’, he will still be responsive to the adult. A socially withdrawn infant however will lack many of the features of responsiveness to others." (in press).

Most studies that have investigated infant social behaviour have been done within an experimental setting directing the mother to interact with her child.
These experimental settings have sometimes required special apparatus (e.g., Murray’s use of an infant chair and mirror placed on a table) while the mother and child are videotaped.
Trained assessors then rate these videotaped interactions, with part of the rating giving an indication of the social behaviour of the infant towards the mother.
A different approach has been taken by Rosenn, Loeb and Jura (1980), who describe an assessment measure that can be used within a clinical setting, investigating the social behaviour of the infant to a relative stranger - the examiner (e.g., child health nurse or paediatrician).
They used a semi-structured social interaction, consisting of the examiner approaching the infant, showing him a toy, picking the infant up, holding him, and putting him down again.
The infant’s overall social behaviour during this interaction is then rated on a 7-point scale, from 1 (extremely negative - screaming, flailing, self-stimulation) to 7 (extremely positive - laughing, cooing, showing interest).
In their paper they discuss the inter-rater reliability of this measure across samples of failure-to-thrive and control infants.

Matthey et al (in press) state :

"One advantage of assessing an infant’s social behaviour with a comparative stranger, rather than with his parent, is that it does not put the parent under any perceived pressure.
Within a clinical setting, such as a hospital or early childhood clinic, to ask a parent to ‘play with her infant’ and for the clinician to assess the resultant infant social behaviour would undoubtedly make parents feel anxious, and possibly unwilling to visit such centres if they felt their competence was being assessed.
A scale that therefore neither requires special apparatus, nor a special sequence of prescribed interactions, and which does not require the parent’s active interaction with the infant, could be considered by clinicians interested in assessing the infant’s social behaviour as being more desirable." (in press).

Such a scale is the Alarme Distress Bébé Scale (ADBB), by Guedeney and Fermanian (2001), constructed to assess infant social behaviour during routine physical examinations given by a range of health professionals.

The ADBB Scale

This French scale consists of eight items related to the infant’s social behaviour, and is used during the clinician’s routine physical examination of the infant.
It requires the clinician to engage the infant in social behaviour - by talking to him, touching him, and smiling to him, which are practices normally undertaken during such examinations.
The eight items, each rated from zero to four (with low scores being optimal social behaviour), are: facial expression; eye contact; general level of activity; self-stimulation gestures; vocalizations; briskness of response to stimulation; relationship to the observer, and attractiveness to the observer.
The clinician keeps in mind the eight items while he/she is conducting the routine physical assessment, and then spends approximately two-three minutes completing the scale.

Matthey, Guedeney, Starakis and Barnett (in press) state :

"Recently high scores on the ADBB (indicative of withdrawn behaviour) have been show to be associated with less optimal interactive behaviours by both the mother and her infant in a Finnish study on 127 two-month-old infants (Puura, 2004) and in an Israeli study on 97 seven to eighteen-month-old infants (Dollberg, 2004).
The original paper (Guedeney & Fermanian, 2001) shows the ADBB scale to have good psychometric properties on a sample of sixty infants in France, aged between 2-24 months.
Good inter-rater reliability was found between raters using it during live (as opposed to viewing the videotape) assessments and an expert’s rating.
Inter-rater reliability, as well as acceptable test-re-test reliability (rs = 0.91, one month interval), have also been demonstrated in a Brazilian study (Lopes, 2004). A total ADBB cut-off score of 4/5 was found to be optimal in detecting those infants considered to have unusually low social behaviour.
This optimum cut-off score has recently been replicated in both a Finnish study (Puura, 2004) and a Brazilian study (Lopes, 2004).
Thus a score of 1 on only five of the eight items is sufficient to indicate possible sub-optimal social behaviour."(in press).

Matthey et al (in press) state that :

"The original paper (Guedeney & Fermanian, 2001) shows the ADBB scale to have good psychometric properties on a sample of sixty infants in France, aged between 2-24 months.... Factor analysis, using the criterion of factor loadings of 0.5 or more, identified two factors accounting for 63.6% of the variance - an interpersonal factor (five items: eye contact, level of activity, self-stimulating gestures, relationship, attractivity) and a non-interpersonal factor (3 items: facial expression, vocalization, response to stimulation), with one complex item (response to stimulation) loading greater than 0.5 on both factors (but with the authors deciding to put it under the non-interpersonal factor).
The authors of the scale recommend that further studies investigate this factor structure. This has been done in a Brazilian study on 90 infants aged between 0 and 2 years (Assumpcao, Kuczynski, Da Silva Rego, & Castanho de Almeida Rocca, 2002). This study found four factors accounting for 63.5% of the variance. While the authors specified the factor loading criterion of 0.5, examination of the data reveals that all non-included items had loadings of less than 0.3, which is the usual criterion used in factor analysis (Child, 1990). Factor 1 consisted of facial expression and level of activity; factor 2 of eye contact and response to stimulation; factor 3 of self-stimulating gestures and the relationship to the observer; and factor 4 of vocalisations. There were no items loading on more than one factor.” (in press).

Dr. Stephen Matthey: South West Sydney Area Health Service: Infant, Child & Adolescent Mental Health Service, George St. Health Centre, 157-161 George Street (2nd Floor), Liverpool, NSW 2170 AUSTRALIA.
ph: (02) 8778 0732/0700; Fax: (02) 8778 0768; email:

Available translations of adbb scale :

Catalan and Castillan

Translations credits :

Italian : Dra Vania Valoriani,Firenza
Catalan and Castillan : Dr Jorge Tizon, Barcelona
Dutch : Pr Peter De Chateau, Nijmegen
German : M. Moralès-Huet & Claudia Klier, Wien
Hebrew : Ruth Feldman & Miri Keren, Bar-Ilan university
Portugal : Barbara Figereido, Porto
Brazil : Franciso Baptista Assumçào, Sao Paulo. Dra Simone Facuri- Lopez, Belo -Horizonte.