Acute neuropsychological effects of methylphenidate in stimulant drug-naïve boys with ADHD II – broader executive and non-executive domains

Attention deficit hyperactivity disorder (ADHD) combined type (DSM-IV –American Psychiatric Association, 1994) is characterised by pervasive behavioural symptoms of inattention, hyperactivity and impulsivity. Hyperkinetic disorder (ICD-10 –World Health Organisation, 1992) is a more restrictive definition requiring higher degrees of pervasiveness and impairment and, as such, describes those with severe ADHD. Current models of ADHD and hyperkinetic disorder (collectively referred to here as ADHD) emphasise a causal pathway from genetic variations generating functional abnormalities in dopaminergic and noradrenergic neurotransmission within fronto-striatal circuitry leading to deficits in executive neuropsychological functioning and, ultimately, to the behavioural manifestations of ADHD (Castellanos & Tannock, 2002).

A range of executive functioning deficits, including inhibition (Barkley, 1997), working memory (Rhodes, Coghill, & Matthews, 2004; Kempton et al., 1999) attentional set shifting and planning (Kempton et al., 1999; Rhodes, Coghill, & Matthews, 2005) have been described in ADHD. Studies utilising broad neuropsychological testing, such as the Cambridge Automated Neuropsychological Testing Automated Battery (CANTAB) (Robbins et al., 1994), suggest that ADHD is associated with deficits in non-executive as well as executive functioning. These include tasks such as spatial recognition and spatial span in addition to tasks upon which performance depends on intact temporal or parietal lobe function – Delayed Matching to Sample (DMtS) and Pattern Recognition (Kempton et al., 1999; Rhodes et al., 2004). Further, deficits in cerebellar timing tasks (Toplak, Rucklidge, Hetherington, John, & Tannock, 2003) and in delay aversion tasks for which the neuroanatomical substrates are unclear (Sonuga-Barke, 2003) have been described.

Despite this evidence for broad neuropsychological dysfunction in ADHD, prevailing models continue to emphasise an assumed primacy of inhibitory and working memory deficits (Barkley, 1997; but see Coghill, Nigg, Rothenberger, Sonuga-Barke, & Tannock, 2005; Castellanos & Tannock, 2002).

Stimulant medication, of which methylphenidate (MPH) is the most widely studied (Greenhill, Halperin, & Abikoff, 1999), remains the cornerstone of the pharmacological management of children with ADHD (Safer & Malever, 2000). However, an understanding of the relationship between the pharmacological actions of MPH and its therapeutic effects remains elusive. A generalised, simplistic explanation suggests that, since oral MPH at therapeutic doses increase the availability of extracellular dopamine (Volkow et al., 2001) and executive functioning can be modified by even small manipulations in catecholamine release (Mehta, Sahakian, & Robbins, 2001), the therapeutic effects of MPH in ADHD may arise from catecholamine-mediated improvements in executive functioning (Volkow et al., 2001). Indeed, it has been proposed that MPH may act relatively selectively to enhance executive functioning (Mehta et al., 2000; Elliott et al., 1997). If deficits in inhibitory control and working memory reflect the core psychopathology of ADHD, and MPH acts as above, performance on inhibition and working memory tasks should improve following administration of MPH. Numerous studies support this hypothesis, whereby MPH appears to enhance inhibitory control, reducing impulsive and variable responding and errors on tasks with prominent inhibitory components (e.g. Berman, Douglas, & Barr, 1999; Shue & Douglas, 1992; Trommer, Hoeppner, & Zecker, 1991; Scheres et al., 2003). However, other studies have failed to demonstrate enhanced inhibitory performance in medicated children with ADHD (Ross, Hommer, Breiger, Varley, & Radant, 1994; Van der Meere, Gunning, & Stemerdink, 1999).

The CANTAB test battery is increasing utilised in child and adolescent neuropsychopharmacological studies because of its extensive validation (Luciana, 2003; Curtis, Lindeke, Georgieff, & Nelson, 2002), its availability in parallel forms for repeated testing, and its sensitivity to pharmacological manipulation. A previous controlled study in healthy adults (Elliott et al., 1997) and previous controlled (Bedard, Martinussen, Ickowicz, & Tannock, 2004; Mehta, Goodyer, & Sahakian, 2004) and uncontrolled (Kempton et al., 1999; Barnett et al., 2001) studies in children with ADHD have reported that MPH enhanced performance on the CANTAB Spatial Working Memory task but have reported inconsistent effects on other tasks. Each of the ADHD studies has significant methodological limitations which limit their interpretation. The two earliest studies were both uncontrolled and non-randomised (Barnett et al., 2001; Kempton et al., 1999). The studies of Kempton and colleagues (1999) and Mehta et al. (2004) reported on small samples with non-standard clinical assessment procedures. Also, both of the previously reported controlled studies included non-medication-naïve subjects (Mehta et al., 2004; Bedard et al., 2004) and utilised a within-subjects crossover design which permits practice effects in addition to potential carryover and/or withdrawal effects of MPH.

The aim of the present study was to address these methodological limitations. We have used a randomised, placebo-controlled, double-blind, parallel design in a large sample of rigorously diagnosed medication-naïve boys with ADHD. We have examined their neuropsychological responses to the first administration of MPH using a broad range of range neuropsychological tasks from the CANTAB battery. We predicted that MPH would enhance performance on a range of tasks with (Go/NoGo, SWM; SOC, ID/ED), and without (e.g. Spatial Span, DMtS), a significant executive component. MPH enhancement of performance was further predicted to be dose-dependent.

We have previously reported the effects of acute MPH on the SWM, DMtS and Pattern Recognition tasks from the CANTAB battery (Rhodes et al., 2004). In this paper we reported that acute MPH did not affect performance on the SWM Task; MPH had no significant effect on Between or Within Search Errors or Strategy Score. Acute MPH also did not affect performance on the Pattern Recognition Task or performance accuracy under simultaneous test conditions of the DMtS task. In contrast, acute MPH at a .6 mg/kg dose restored performance accuracy across each of the delay conditions of the DMtS task to the levels observed in healthy controls (see Rhodes et al., 2004) and this improvement was not associated with significant changes in response latencies. Impaired functioning, however, continued to be seen across each delay condition with placebo and MPH .3 mg/kg. Here we report the effects of acute MPH on performance of the previously unpublished tasks from the CANTAB battery and the Go/NoGo task.

Method

This study was approved by the Tayside Committee on Medical Research Ethics. All participants and their parents/guardians provided written informed consent.

Results

Mean performance and statistical comparisons for each of the treatment groups on each task are summarised in Table 4. F and p values reported in Table 4 represent SESSION × TREATMENT interactions. Main effects of TREATMENT GROUP, TASK DIFFICULTY and SESSION are reported only when significant.

Discussion

The acute administration of oral MPH (.3 and .6 mg/kg) under randomised, placebo-controlled, double-blind conditions to a cohort of 73 medication-naïve boys with ADHD improved neuropsychological performance on three tasks without a prominent executive component. Shortened response latencies were reported on a complex reaction time task and lengthened response latencies (which were not associated with increased task accuracy) were found on a Spatial Recognition memory task. As we previously reported (Rhodes et al., 2004), performance on a Delayed Matching to Sample task was restored to that described for healthy developing children. Contrary to predictions, acute MPH failed to improve performance on neuropsychological tasks with a prominent executive component. This absence of effect was observed on sensitive tests of inhibition, working memory, strategy formation, planning, and attentional set-shifting. These findings are striking in that they contradict the existing literature. Specifically, the hypothesis that MPH selectively enhances performance on neuropsychological tasks with a prominent executive component was not supported.

Whilst it was not appropriate to examine clinical response to a single dose of medication, the results presented do not support the hypothesis that the therapeutic effects of MPH in children with ADHD are mediated by improved executive neuropsychological performance. Further, the present study fails to support the findings of two earlier studies which utilised the Go/NoGo task and demonstrated an enhancement of inhibitory processes with acute MPH (Trommer et al., 1991; Broyd et al., 2005). However, Van der Meere and colleagues also conducted a double-blind, placebo-controlled study of strictly defined drug-naïve children with ADHD (Van der Meere et al., 1999) and found that stimulant medication (chronically administered) failed to enhance performance on a Go/NoGo task. It is also possible that the differences between these studies could be a consequence of the Go/NoGo tasks used. The current study and that of Van der Meere used tasks with visual targets whereas the Trommer and Broyd studies used an auditory Go/NoGo task. Further, the current study used a relatively fast presentation rate with an inter-stimulus interval of .9 whilst both the Trommer and the Broyd studies used a medium presentation rate with an inter-stimulus interval of between 3 and 4 seconds. This explanation seems less likely as Van der Meere and colleagues investigated performance across several inter-stimulus intervals (1, 4 and 8 secs) and found no relationship between presentation rate and medication effects.

The present findings also fail to fully replicate those from previous comparable studies using the CANTAB tasks (Bedard et al., 2004; Kempton et al., 1999; Mehta et al., 2004; Barnett et al., 2001). Each of these studies reported that MPH reduced BSE on the SWM task, and both studies that included the ID/ED task reported improvement with MPH (Kempton et al., 1999; Mehta et al., 2004). However, studies using the SOC planning task reported mixed findings (Kempton et al., 1999; Bedard et al., 2004). We found no effects on these three tasks. This may relate to differences in study design. The two earliest studies were both uncontrolled and non-randomised (Barnett et al., 2001; Kempton et al., 1999). The studies of Kempton and colleagues (1999) and Mehta et al. (2004) reported on small samples with non-standard clinical assessment procedures. Also, both of the previously reported controlled studies included non-medication-naïve subjects (Mehta et al., 2004; Bedard et al., 2004). Both also utilised a within-subjects crossover design which permits practice effects in addition to potential carryover and/or withdrawal effects of MPH. We observed improvements on several tasks between baseline and acute challenge in the present study that cannot be attributed to MPH. Finally, unlike these previous studies, the statistical analyses for the present study included unmedicated baseline performance when estimating the significance of MPH effects.

Stimulant drug effects on tasks without a prominent executive component are relatively understudied. Using an uncontrolled study design, Kempton et al. (1999) suggested that MPH enhanced performance on several tasks without a prominent executive component: Spatial Span, Pattern Recognition and Delayed Matching to Sample, but not on others: Spatial Recognition and Paired Associates Learning (PAL). Controlled studies have included only small subsets of these tasks. MPH improved performance on the Spatial Span task as reported by Bedard, Ickowicz, and Tannock (2002) but did not improve performance on the Spatial Recognition task (Bedard et al., 2004; Mehta et al., 2004) or on the Pattern Recognition task (Mehta et al., 2004). The present study found that acute MPH (.6 mg/kg) exerted significant effects on response latencies on two such tasks. MPH .6 mg/kg lengthened latencies for both correct and incorrect responses on the Spatial Recognition task and shortened reaction time latencies on the 5-choice condition of the Reaction Time task. These findings support the contention that MPH exerts therapeutic effects by improving ‘regulatory ability’– slowing performance during the difficult conditions of tasks and shortening response times during easier stages (Douglas, 1999; Berman et al., 1999). These beneficial effects may be dependent upon memory load. Berman et al. (1999) demonstrated that at a low memory load, improvement in performance accuracy occurred with no cost to reaction time, irrespective of dose of MPH, whereas at higher loads there was a dose-dependent effect of MPH to slow reaction times. In support of this model, we found that acute MPH did not uniformly slow or speed responding. MPH lengthened responding on the Spatial Recognition task, but shortened response times on the Reaction Time task. While MPH has contrasting effects on reaction times on these two tasks, both effects can be regarded as enhancements of cognitive functioning. The Spatial Recognition task is a relatively complex task in comparison to the lighter cognitive processing burden of the Reaction Time task; hence lengthened responding in the former and shortened responding in the latter can be regarded as an enhancement of performance in both tasks. Whilst these results support those found in other studies, it should be noted that the reported p values for each of these effects were in the .5 to .1 range. Considering the number of tasks investigated, it remains possible that these results represent Type 1 errors. Further significant effects were found on another task without a prominent executive component, the DMtS task, which assesses the ability to hold information in memory over a short delay (Rhodes et al., 2004).

The present study also supports the notion that MPH selectively enhances discrete aspects of neuropsychological functioning (Gao & Goldman-Rakic, 2003) and that these effects extend beyond baseline impairments (Rapport & Kelly, 1991). Within the present sample, baseline assessments revealed a wide range of deficits in neuropsychological functioning (see Table 2) (Rhodes et al., 2005). Acute MPH, however, failed to enhance performance on many of these tasks. Indeed, apart from the amelioration of deficits on the DMtS task and the slowing of incorrect responses (but without improved accuracy) on Spatial Recognition, MPH-related enhancements were restricted to aspects of neuropsychological functioning in which boys showed no impairment at baseline. Importantly, despite there being no evidence for MPH-related improvement on other aspects of neuropsychological performance, there was no evidence of MPH-related performance deficits. Specifically, there was no evidence for MPH reducing cognitive flexibility (see ID/ED results) as has previously been suggested (Robbins & Sahakian, 1979; Dyme, Sahakian, Golinko, & Rabe, 1982).

Several potential limitations of this study should be addressed. The recruitment of participants meeting criteria for both DSM-IV ADHD Combined Type and ICD-10 Hyperkinetic Disorder may have generated a more severely affected clinical sample than previous studies. However, a recent re-analysis of the Multimodal Treatment Study for ADHD (MTA) data (E. Taylor, pers. comm.) suggests that such a population is more, rather than less, likely to respond to MPH. Presumably, this ought to increase rather than decrease the chances of detecting clinically meaningful changes. The inclusion of only male subjects reduces the generalisability of the findings to females with ADHD.

No attempt was made to measure clinical response to the acute MPH challenge. However, in view of the highly structured, controlled and relatively novel environment required for neuropsychological testing, it is unlikely that such ratings (even with relatively objective measures such as from an actometer) would accurately capture clinically relevant symptom changes. Also, although participants were randomised following baseline testing, it is possible that the present findings are a consequence of latent differences between the treatment groups in either clinical or neuropsychological characteristics. However, there were no statistically significant differences between the three treatment groups on a wide range of clinical variables: age, BPVS Percentile Rank, the presence of a wide range of comorbid conditions, or on neuropsychological functioning at baseline. Whilst our sample size was modest, it easily exceeds that of previous, comparable studies. The use of a single session between-subjects rather than multiple session within-subject design results in a reduction in power; however, as detailed above, such repeated measures designs also add potential confounding factors. To evaluate the first-ever neuropsychological response to MPH in drug-naïve subjects mandates a between-subjects design. The F-ratios reported above are frequently <1, suggesting that the loss of power of a between-subjects design is unlikely to have been a major issue with respect to the negative findings. It is also worth noting that the consistently negative results across multiple neuropsychological domains reinforce the possibility that MPH may not exert the previously anticipated effects.

One possible explanation for the absence of effects of MPH on executive functioning is that we have only studied the acute response to a first-ever dose and chronic administration may have very different effects. Clinically, the positive impact of MPH on behaviour within approximately an hour of administration has traditionally been interpreted as evidence that its therapeutic effects are immediate and are not mediated by long-term neurobiological adaptation (Solanto, 1998). Perhaps this is incorrect? Previous controlled study designs, utilising these same tasks, have considered the possibility that the ‘enhanced’ performance seen with MPH may be a consequence of neuropsychological rebound secondary to acute medication withdrawal. Definitive prospective studies comparing acute and chronic responses in stimulant-naïve subjects are awaited.

ADHD is associated with a wide range of neuropsychological deficits. While findings from this study do not support the popular hypothesis that MPH enhances performance on executive functioning tasks, they do suggest selective enhancement of several aspects of non-executive cognitive functioning which may reflect increased capacity for self-regulation. As a majority of children with ADHD demonstrate positive clinical responses to MPH, this suggests that clinically important cognitive deficits may not all be ‘executive’ in nature. Importantly, MPH does not impair performance on any of the tasks studied.