EN EL AMBITO DEL MERCOSUR, LA CUALIDAD DE ESTAR DIRIGIDAS A LA GENERALIDAD, APARECE MAS CLARA EN EL CASO DE LAS

The first research aim of this thesis was to investigate the associations of measures of sustained attention, behavioral inhibition, and conflict-monitoring with the categorical diagnosis of ADHD and compare the strength of these associations. The results revealed a deficit in sustained attention in children with ADHD as compared to unaffected children. An additional but weaker deficit was found for conflict monitoring, whereas children with ADHD exhibited no deficit in behavioral inhibition relative to children without ADHD.

This strong impairment in sustained attention is consistent with previous studies and meta-analyses, in which children with ADHD had higher omission error rates, committed two or more omission errors in a row more often, and displayed a higher RTV than unaffected control children (Helps et al., 2011; Karalunas, Geurts, Konrad, Bender, & Nigg, 2014; Kofler et al., 2013; Metin et al., 2012; Tamm et al., 2012). However, these findings could be explained by at least two different, but not excluding theoretical accounts. First, the results are in line with theories explaining a larger RTV and higher omission error rates by the presence of periodic attentional lapses (Leth-Steensen, Elbaz, & Douglas, 2000; Sonuga-Barke & Castellanos, 2007; Weissman et al., 2006). According to the DMN hypothesis, periodic attentional lapses occur at low frequency bands of response times due to a failure to suppress activity of the DMN and a simultaneously reduced activity of the task positive network (Castellanos, Kelly, & Milham, 2009; Sonuga-Barke & Castellanos, 2007).

However, the DMN hypothesis has recently been challenged (Karalunas et al., 2014; Karalunas, Huang-Pollock, & Nigg, 2013) as a higher RTV was associated with both very low and high frequency bands of response times (Karalunas et al., 2013). Similarly, both hypo- and hyperactivation of the DMN have been proposed for ADHD and disagreement exists whether differences in DMN activity between children with and without ADHD are present at rest (Konrad & Eickhoff, 2010). Moreover, attentional lapses merely define an event and besides a failure to suppress DMN activity, various processes might contribute to a higher RTV, as for instance, stimulus encoding, processing speed, speed-accuracy trade-offs, post-error slowing, motor preparation, response execution, effort and arousal (Karalunas et al., 2014).

Focusing on arousal, a second explanation of higher RTV in ADHD has been derived from diffusion models. Meta-analyses of diffusion models revealed a slower drift rate in

children with ADHD as compared to unaffected children (Huang-Pollock et al., 2012; Karalunas et al., 2014). Like the signal sensitivity parameter d’ (Ratcliff & McKoon, 2008), drift rates can be interpreted as an index of arousal (Karalunas et al., 2014; Sanders, 1983; van der Meere & Sergeant, 1988) and in line with the Petersen-Posner perspective on brain networks involved in sustained attention (Petersen & Posner, 2012; Posner & Petersen, 1990), arousal is responsible for both establishing and maintaining an alert stage. From a neurobiological perspective, optimal arousal and better sustained attention are accounted for by phasic norepinephrine activation of the LC, whereas a less phasic and more tonic norepinephrine activation of the LC is associated with less optimal arousal and higher distractibility (Aston-Jones & Cohen, 2005; Langner & Eickhoff, 2013; Petersen & Posner, 2012; Posner & Petersen, 1990; Sarter, 2001). Pharmacological interventions further support the account, as methylphenidate and especially atomoxetine, two of the most common pharmacological substances for the treatment of ADHD, influence norepinephrine transmission and reuptake (Bari & Aston-Jones, 2013; Hannestad et al., 2010).

While non-optimal arousal is a promising alternative explanation to a failure to suppress DMN activity, several aspects of the account can be challenged. First, the phasic/tonic dynamics of norepinephrine and the LC activity have recently been transferred to the DMN hypothesis (Aboitiz, Ossandón, Zamorano, Palma, & Carrasco, 2014). Within this account the DMN was associated with tonic catecholamine activity in a sense that a deregulated tonic activity results in an unbalanced DMN leading to a higher distractibility (Aboitiz et al., 2014). Second, the account strongly relies on the interpretation of drift rates obtained from diffusion models. However, it is unclear whether cognitive tasks like the Flanker task, or CPT really meet the criteria for a diffusion model analysis. For instance, the diffusion model approach assumes a continuous sampling process of information (Ratcliff & McKoon, 2008; Voss, Nagler, & Lerche, 2013). While this assumption is obviously plausible for ambivalent visual stimuli like fields of pixels with different colors used in brightness or color-discrimination tasks (Ratcliff, 2002), it is less clear whether there is a continuous sampling process in typical cognitive tasks like the Flanker task, or CPT. Yet, even if a continuous sampling process could be assumed various diffusion models exist to model decision processes in the Flanker task (Hübner et al., 2010; Hübner & Töbel, 2012; White et al., 2011) so that different drift rates can be obtained from the same data leaving it unclear whether the association of slower drift rates and ADHD holds for different diffusion models. Furthermore, drift rates are generally not interpreted in terms of arousal. Instead the drift rates are though to map the speed of information uptake or the quality of information extracted

from the stimulus (Ratcliff, Love, Thompson, & Opfer, 2012; Voss et al., 2013). Even if changes in a person’s level of arousal influence drift rates in cognitive tasks, drift rates are affected by other processes as well and therefore cannot be regarded as a pure indicator of arousal. Finally, diffusion models cannot deal with omission errors as omission errors are excluded from the analysis (Voss et al., 2013). However, given the factor-analytical results from this study, omission errors and RTV should be considered indicators of the same construct (sustained attention). Therefore, theoretical accounts explaining RTV need to explain omission errors as well. The DMN hypothesis addresses both RTV and omission errors, and omission errors and especially two or more omission errors in a row literally correspond to attentional lapses. Hence, although effects of arousal cannot be excluded, the results of the current study might be better explained by the DMN hypothesis than by arousal.

While the present study revealed deficits in sustained attention in children with ADHD, no deficits were found in measures of behavioral inhibition. Neither did children with ADHD commit more inhibition errors in the Go/No-Go task than unaffected children, nor did they exhibit a larger Flanker interference effect for errors. Indeed, although deficits in behavioral inhibition lie at the core of theories of ADHD, as for instance the executive dysfunction theory (Barkley, 1997) or multiple pathway models (Sonuga-Barke, 2002; Sonuga-Barke et al., 2010), several studies failed to report deficits in children with ADHD with regard to No-Go inhibition (Börger & van der Meere, 2000; Collings, 2003; Fallgatter et al., 2004; Smith et al., 2006) and Flanker interference errors (Booth et al., 2007; Crone, Jennings et al., 2003; Drechsler, Brandeis, Földényi, Imhof, & Steinhausen, 2005; Johnstone & Galetta, 2013). Furthermore, it is assumed that only about 20%–50% of children with ADHD exhibit deficits in behavioral inhibition (Coghill et al., 2014; de Zeeuw et al., 2012; Sonuga-Barke et al., 2010; Tsal et al., 2005). Thus, taking into account the sample size of this study, it is not unusual that we did not find any differences between children with and without ADHD on behavioral inhibition. Furthermore, the behavioral inhibition indicators from the Go/No-Go and Flanker task did not correlate. This absence of a correlation indicates that withholding of responses and interference control might at least be separable components of behavioral inhibition, if not even unrelated constructs (see Section 5.2.2. for a detailed discussion on measurement and validity concerns). For instance, it was previously suggested that a Go/No-Go task involves response related inhibition only, while the Flanker task involves both, response related inhibition, and stimulus (distractor) related inhibition, and consequently correlations between these measures should be small (Stahl et al., 2014).

Within the Flanker task measures of behavioral inhibition can be separated from measures of conflict-monitoring (Botvinick et al., 2001, 2004). Unaffected children showed a reduction in the Flanker interference effect for response times after incongruent trials (sequential congruency effect for response times), whereas children with ADHD did not display such a modulation of the Flanker interference effect or rather in the opposite direction. While there were no differences between children with and without ADHD in the Flanker congruency effect for response times, the differences in the sequential congruency effect might reflect a deficit in conflict-monitoring. Interestingly, neuronal activity in parietal networks implicated in sustained attention has been related to activity in the ACC (Fassbender et al., 2004). Similarly, in our study correlations of medium size have been observed between various measures of conflict-monitoring and sustained attention (see Table C2 in Appendix C), indicating that the absence of conflict-monitoring in children with ADHD might be a byproduct of deficits in sustained attention. This interpretation is further supported by results from mind wandering. By definition, mind wandering state implies a deficit in sustained attention and higher mind wandering has been associated with deficits in conflict-monitoring (Kam et al., 2012).

Taken together and in line with one previous study (Tsal et al., 2005) it can therefore be concluded that deficits in sustained attention might be more pronounced and characteristic for ADHD than deficits in behavioral inhibition. In addition, reconsidering the deficit in conflict-monitoring and its relation to sustained attention and mind wandering, sustained attention might even be a prerequisite for conflict-monitoring. Hence, incorporating a sustained attention pathway in multiple pathway models of ADHD should be beneficial for understanding the disorder in more detail.