Many have emphasized the need to chart how cognitive abilities change over development for individuals with genetic disorders, rather than limiting studies to the comparison to typically developing controls, the traditional mental age matching approach, as the latter strategy removes, rather than investigates, age-related changes (e.g. Thomas, Annaz, Ansari, Scerif, Jarrold & Karmiloff-Smith, 2009). Recent attention and control findings highlight the value of this approach. Children with WS (on average, 12-year-olds) performed more poorly than expected, given their level of cognitive functioning, on a task that required visuo-motor control, although they were also, though more subtly, impaired on a perceptual matching task, and in both tasks their performance was comparable to that of 4-year-old children (Dilks, Hoffman & Landau, 2008). Adults (on average, 23-year-olds) with WS did not differ from children with the condition on either task (and thus from typically developing 4-year-olds), a pattern that to the authors suggested ‘developmental arrest’. However, data from other genetic disorders suggest that such a conclusion may be premature without longitudinal insights. Cornish, Cole, Karmiloff-Smith and Scerif (in preparation) studied attentional control in a sample of 4- to 10-year-olds with FXS and typically developing children by asking them to monitor a rapidly presented stream of visual stimuli for infrequent targets under a number of attentionally demanding conditions. Children with FXS performed more poorly than expected given their chronological age and developmental level, and older children with FXS were no better than younger children in the group, a pattern that may have suggested not just severe developmental delay, but ‘developmental arrest’. However, when followed longitudinally, the same children showed improvements in performance that were comparable to those of typically developing children matched to their developmental level, underscoring the danger of drawing conclusions about developmental change (or none thereof), from cross-sectional data, as these intrinsically confound age-related and individual differences.
Longitudinal studies have also become increasingly used in the study of behaviourally defined developmental disorders affecting attention, as a way of mapping the interaction between genetic and environmental factors in predicting risk for a disorder (e.g. S.E. Stevens, Kumsta, Kreppner, Brookes, Rutter & Sonuga-Barke, 2009, for ADHD). In the case of ASD, cognitive development can be studied prospectively from infancy in the siblings of older children with autism, because siblings are at higher risk of developing ASD given high heritability estimates for the condition (Elsabbagh & Johnson, 2010). At 10 months, for example, siblings take longer to disengage attention from a central stimulus to a concurrently presented peripheral stimulus and electrophysiological recordings show atypical latency for the P400 and atypical gamma band activity when processing direct eye-gaze (e.g. Elsabbagh, Volein, Csibra, Holmboe, Garwood, Tucker, Krljes, Baron-Cohen, Bolton, Charman, Baird & Johnson, 2009a; Elsabbagh, Volein, Holmboe, Tucker, Csibra, Baron-Cohen, Bolton, Charman, Baird & Johnson, 2009b). Of note, as for disorders of known genetic origin, cognitive abilities need to be followed longitudinally, as illustrated by recent data: Young, Merin, Rogers and Ozonoff (2009) assessed outcomes at multiple time-points until the age of 24 months for toddlers who, at 6 months, had been assessed with a still face procedure. At that time-point, a significantly greater proportion of siblings at risk for autism than control infants had preferred looking at their mother’s mouth, compared to eyes, both during the still face episode and in baseline periods preceding and following it. When tested longitudinally, gaze preferences in infancy did not predict later likelihood of developing autistic symptoms, underscoring the need to re-evaluate the utility of such measures and/or discover protective factors leading to later absence of symptomatology. Surprisingly, greater preference for the mouth region predicted faster growth of expressive vocabulary, suggesting novel hypotheses about the role of attention to audiovisual events at the very onset of language acquisition. More broadly, these data challenge notions derived from studying the relationship between visual attention and other cognitive domains at single time-points, or even through cross-sectional comparisons.