Discussion
This study set out to characterise inhibitory information processing in 22q11DS, using multiple measures (active/passive SEM (PPI and PPF); antisaccade inhibition) and examine potential associations with executive functioning. This is the first study to systematically examine automatic and controlled components of information processing and their relationship to executive functioning in 22q11DS. We report evidence of impaired controlled attentional processes, i.e. antisaccade inhibition in 22q11DS individuals, as well as some evidence (though statistically non-significant) of sensorimotor gating deficits—in a prepulse inhibition paradigm with reduced attentional modulation of the startle eyeblink response.
Previous studies of PPI in 22q11DS have found impaired sensorimotor gating compared to typically developing controls when employing a passive PPI task [14]; however, group differences on the passive PPI component of the task were not reported in the present study. Differing methodologies including lead intervals, levels of white background noise, relative stimulus intensities and number of trials presented may account for this disparity [18, 32]. Additionally, the effects of age on PPI may explain the difference in results, with a younger sample recruited by Sobin and colleagues [14]. Others have recently reported that levels of PPI increase with age [33], indicating that the age of our cohort may have contributed to the differences between this study and previous reports of PPI in 22q11DS [18, 17]. We are currently limited to speculation regarding the maturational trajectory of PPI in clinical groups; however, based on the current findings, differences in a PPI paradigm may best explain the low levels of PPI reported for both 22q11DS and HC groups in this study compared to previous reports in both 22q11DS [14] and other populations which typically report PPI in healthy control participants around 50%–60% [18]. These floor effects in our passive PPI condition may explain our failure to replicate associations between measures of executive functioning and PPI that have been reported previously in 22q11DS [13] and clinical conditions prevalent in 22q11DS [autism spectrum disorder; ASD [34]].
The results of the present study indicate attention-modulated disruption of sensorimotor gating in 22q11DS. One of the purposes of this study was to investigate higher- and lower-order levels of information processing in 22q11DS. Therefore, we incorporated a PPI paradigm that systematically examined relatively automatic (passive task) and controlled (active task) attention. During active attendance, the effects of an acoustic pre-pulse are typically magnified, with an increase in startle suppression (i.e. increase in PPI) at short lead intervals and increased startle (i.e. decrease PPI or increase PPF) at longer lead intervals [17]. We showed that individuals with 22q11DS appear to exhibit appropriate inhibition and facilitation of the startle response during the passive task, indicating normal early or automatic information processing. However, by failing to display the normal pattern of greater PPI and PPF during attended, rather than ignored, pre-pulses, the 22q11DS group exhibited a dysfunction in controlled attentional processing, rather than a pre-attentive dysfunction.

Antisaccade inhibition
This is the first time antisaccade inhibition has been examined in 22q11DS. Our results were consistent with reports from other clinical populations [schizophrenia [25], first episode psychosis [35]] indicating that 22q11DS participants were less able to inhibit an automatic pre-potent response (i.e. increased antisaccade error). Increased antisaccade error in 22q11DS may reflect the inhibitory demands of the task and support our PPI findings of impaired controlled attention processes in 22q11DS. Our findings are also consistent with studies that show poor antisaccade performance in ultra-high risk for schizophrenia [36] and first episode schizophrenia groups [37].
Our passive PPI task was uncorrelated with the antisaccade task. This is consistent with previous studies in schizophrenia that show no association between passive PPI and antisaccade accuracy [38]. We partly concur with Swerdlow et al. [39] that the absence of association between passive PPI and antisaccade indicates that the measures are partly "dissociable and non-redundant" (p336) measures of information processing, resembling tasks tapping early versus late portions of the information processing chain. The present study shows poorer antisaccade accuracy, as well as attention-modulation deficits in PPI in 22q11DS. Taken together, these findings suggest a more cortical, in particular prefrontal, than subcortical (i.e. superior collicular) dysfunction in our 22q11DS sample.

Limitations
Several limitations are associated with this study. The lack of an intellectually matched control group and convenience recruitment approach to participant selection is inherent to research of this nature and a limitation in 22q11DS research noted by our group previously [40]. In addition, 22q11DS is associated with a range of mental health conditions (see Participants); and these diagnoses and the medications used to treat them may have influenced our findings. However, these clinical diagnoses also reflect a defining characteristic of 22q11DS and were too few in number and varied in type to control statistically. Regarding medication, it is unclear how best to control for the influence of medication status given the medication participants reported using at the time of testing may have improved or impaired performance on inhibition tasks. In the future, our findings may be strengthened with additional controls for medication and clinical status in a larger sample.

Clinical implications
The path of cognitive maturation in 22q11DS remains unclear; and future studies may benefit from the inclusion of participants aged across the lifespan to help determine whether inhibition processes are underdeveloped, delayed or constant across the lifespan in this group. Future studies may also benefit from exploring the influences of GABA, dopamine and glutamate transmission [13], as they are likely to be involved in the disturbed attention-modulated PPI and, we would suggest, the poor antisaccade performance reported in the current sample.