Stability of the ADI-R
15
The relative improvements found in the different domains of the ADI-R have been reported in several
other studies, although profiles of change seem to be more variable in retrospective studies or those
involving shorter follow-up periods (Moore & Goodson, 2003; Seltzer et al., 2003; Shattuck et al.,
2007). On the whole, improvements in the RBSP domain tend to be less marked than improvements in
the RSI and NVC/VC domains (Charman et al., 2005; Fecteau et al., 2003; Piven et al., 1996). Whilst
measurement issues may play some role here (there are fewer items in the ADI-R RBSP domain than in
the RSI, NVC and VC domains; see also Piven et al., 1996), the consistency of the reported pattern of
change across the three domains suggests that this may be more than just an artefact of measurement
and that the developmental trajectory of each of the components of the triad of impairments may well
differ.
Few studies to date have considered the stability of ADI-R symptom severity beyond the algorithm and
domain levels. In the current study, changes in the severity of specific symptoms were also considered.
Whilst findings should be interpreted with caution due to the use of multiple comparisons, the results
highlight several behaviours that showed significant improvement over time. The symptoms showing
most change in the RSI domain were: ‘interest in other children’, ‘response to other children’,
‘appropriate facial expression’ and ‘use of other’s body to communicate’. Within the NVC domain,
‘gestures’, ‘nodding’ and ‘social play’ were the specific behaviours that improved. Improvement in
‘repetitive use of objects’ was identified within the RBSP domain. No significant improvements were
identified within the VC domain although this is likely to reflect the small number of participants who
were verbal at T1. Fecteau et al. (2003) and Starr et al. (2003) also reported improvements in similar
items of the ADI-R although Cox et al. (1999) noted improvements in different items (gaze, quality of
social overtures, pointing and headshaking) between the ages of 20 and 42 months, suggesting that
patterns of change may vary according to age group.