Link to the complete thesis

Examining the clinical utility of the modified Alarm Distress Baby Scale (m-ADBB) for the detection of early signs of Autism Spectrum Disorder

Preliminary Literature Review

Background / Aim

Although behavioural signs of Autism Spectrum Disorder (ASD) can be observed within the first two years of life, early detection and diagnosis of ASD remains challenging. Existing routine screening instruments have significant limitations, either being too time-consuming to use or not being reliably validated for routine screening. Therefore, no reliable and easy to use routine screening instrument for ASD in infants and toddlers is currently available to clinicians.

Some of the early ASD related deficits in social communication and interaction behaviours are also typical of sustained social withdrawal in infants and toddlers. Only one study has investigated this link to date, and its findings suggest that sustained social withdrawal in infants and toddlers may be indicative of ASD (Wendland, Gautier, Wolff, Brisson, & Adrien, 2010).

The current study aims to test whether a brief observational screening instrument for social withdrawal in infants, the modified Alarm Distress Baby Scale (m-ADBB) may be clinically useful for the detection of ASD in the first two years of life. It is hypothesised, that children with ASD will score higher on the m-ADBB than typically developing (TD) children, indicating more symptoms of social withdrawal.


Home-video recordings of children with ASD and children with typical development from approximately age 12 months and 24 months were analysed using the m-ADBB.


Eleven children with a diagnosis of ASD and eleven children with typical development were recruited to the study; videos for 20 children were available at each age. Children with a diagnosis of ASD scored statistically significantly higher on the m-ADBB than children with typical development at 12 month (Z=-2.54; p=0.023; r=-0.57) and at 24 month (Z=-2.40; p=0.023; r=-0.54). Five of ten children with ASD met the m-ADBB criterion for social withdrawal in their 12 month videos, and four out of ten in their 24 month videos. Using a lower cut-off score increased detection rates (7 at 12 month; 8 at 24 month). False positive results were low at both ages and when both cut-offs were applied (range 1 to 3 out of 10).


The findings suggest that observing only five social withdrawal behaviours as operationalized by the m-ADBB is useful in flagging possible presence of ASD in children during their first two years of life. Further research is required to establish the scale’s sensitivity and specificity for ASD detection.

Expert’s note: Extremely interesting dissertation work that confirms the results of a first study done on the use of the detection of sustained relational withdrawal as an early indication of an ASD. If the results were confirmed, the M-ADBB and the ADBB would then be tools of first choice for early identification of babies at risk of ASD without extensive training of front-line professionals who could then rapidly refer families to specialized teams, thus allowing for increasingly early management of ASD.