The engagement of the brain’s default mode network (DMN) may be compromised in attention deficit/hyperactivity disorder (ADHD) and one mechanism by which medication has its beneficial effects may be through modulation of this network. Liddle and colleagues in the current issue of the Journal (Liddle et al., 2011) now show that motivation can modulate the effects of methylphenidate in ADHD via effects on the DMN. Regions, which include the posterior cingulate cortex (PCC), precuneus, medial prefrontal cortex (PFC) and inferior parietal lobes and originally referred to brain areas consistently decreasing in activity (or deactivated) relative to goal-directed task conditions.

Liddle and colleagues (2011) used a Go/No-Go task with the balance between speed and restraint modulated in two conditions. The high incentive condition was associated with gain/loss of 5 points, whereas the low incentive condition was associated with gain/loss of 1 point for correct/failed inhibitions. A single point was awarded for correct Go responses. The authors argue that high incentive and methylphenidate had similar effects in ADHD as evidenced by both performance and brain imaging data. For the latter, a weighted mean deactivation was calculated for the DMN for each trial type. Task-induced deactivations (TID) were significant in the control group, but did not differ between incentive conditions. For patients, TID were only significant in the high incentive condition when off medication. Patients on medication, while not significantly deactivating the DMN, differed from themselves when off medication for the low incentive condition, and did not differ from controls. The authors concluded that abnormal deactivation in ADHD (under the low incentive condition) can be normalised by either higher incentives or medication. This important conclusion provides a brain target for theories positing abnormal motivational or reward functioning in ADHD. However, this conclusion raises a number of questions regarding the origin and generalisability of the findings, such as: What are the endogenous mediators of the observed effects? Are the results specific to deactivation of the DMN? Can the findings be generalised to other treatments for ADHD? These questions will be addressed in turn.

One of the most interesting aspects of the study by Liddle et al. (2011) is the suggested role for dopamine in modulating the DMN particularly because this does not include regions with the most dense dopaminergic projections, such as the striatum. The effects within the DMN may thus be due to a combination of direct effects of dopamine within the regions modulated, and indirect effects putatively mediated via the striatum. The link with TID is evidenced by a negative correlation between dopamine transporter availability in the striatum and TID in the precuneus; higher transporter availability was associated with less deactivation during a visual attention task requiring tracking of moving stimuli (Tomasi et al., 2009).

These findings support the generally accepted conclusion that attentional control processes are modulated by dopaminergic systems. This may seem over-inclusive and more specific roles for dopamine within this processing framework seem likely. However, the increased demand for attentional control in most neuroimaging tasks deactivates the brain’s DMN, in addition to activating cortical brain regions. Higher levels of dopamine transmission in the striatum are thus hypothesised to be associated with augmented TID. While this has not been tested directly, for example with positron emission tomography studies of dopamine release, the findings of Tomasi et al. (2009), Liddle et al. (2011) and others fit with this prediction.

Which aspects of dopamine transmission are important here are not understood, although a separation of more ventral striatal regions (connected with ventromedial PFC and precuneus/PCC) and more dorsal striatal regions (connected with lateral PFC and motor cortex) is predicted. Indeed, (Kelly et al., 2009) provide partial support for a differentiation of cortical effects of dopaminergic agents based on the effects of levodopa on correlations with dorsal and ventral striatal regions.

The use of connectivity analyses provides a qualitatively different perspective to TID analysis. Typically, no active task is used during data acquisition and analysis relies on the so-called ‘intrinsic’ functional connectivity between brain regions. Whether altered striato-cortical connectivity (or correlations), as shown with levodopa, is the basis of the modulation of TID with methylphenidate remains to be determined. This question feeds into the default mode interference hypothesis (Sonuga-Barke & Castellanos, 2007), which argues that the operation of the DMN during task performance is impaired in ADHD, leading to ‘intrusion of spontaneous low frequency brain activity’. What is unknown is how TID relates to low-frequency fluctuations in different incentive contexts and under different levels of dopamine neurotransmission.

Theoretically, there is little reason for the effects of dopaminergic drugs to be limited to TID, given the role of other regions in attentional control functions, such as the anterior cingulate cortex, dorsolateral PFC and posterior parietal cortex. Indeed, all of these regions have variously been shown to be modulated by dopaminergic drugs, including methylphenidate. While Liddle et al. (2011) did not present an analysis of the task activation networks, the study by Tomasi et al. (2009) showed a positive correlation between dopamine transporter availability in the caudate nucleus and activity in multiple regions, including the anterior cingulate cortex and superior parietal lobe, during their visual attention task. Unlike task activation networks, TID are more regionally consistent across studies and Liddle et al. (2011) utilised this fact to justify the use of a predefined network from which the changes in fMRI signal were averaged. Exploiting recent developments in multivariate image analysis, Marquand et al. (in press) also showed augmentation of TID following methylphenidate, but in healthy volunteers performing a rewarded working-memory task. Differences between drug conditions were defined by a coherent pattern of changes rather than a collection of local changes and reduced activity on methylphenidate was seen for TID and the task-induced activation networks.

Similar effects were observed for the noradrenaline reuptake inhibitor atomoxetine, questioning the degree to which purely (subcortical) dopaminergic mechanisms can account for the effects of methylphenidate on TID. For both drugs the effects were most prominent in a high incentive condition, when accurate performance was rewarded. This finding in healthy adults contrasts with Liddle et al. (2011) where TID were augmented in children with ADHD in a low incentive condition. An important question to answer is whether adults with ADHD show more similar effects to children with ADHD or healthy adults. Nonetheless, the similarity in the effects of methylphenidate and atomoxetine on task-induced activation and deactivation networks in healthy adults is striking and leads to the prediction that reduction of task-related activity (reduced activation and augmented deactivation) might represent a general mechanism by which drugs that increase catecholamine transmission can influence cognitive performance (which in turn interacts with the degree of incentive).

Remarkably, modafinil, a wake-promoting agent with efficacy in reducing ADHD symptoms, and probable positive effects on the dopamine and noradrenaline systems, can also augment TID (Minzenberg, Yoon, & Carter, in press) – but limited to the ventromedial PFC. Motivation was not modulated in this healthy volunteer study, although effects of such modulation would now be expected. Unfortunately, as is the case in most areas of neuroimaging, all the studies cited used different tasks, with different imaging protocols on different scanners, in cohorts of different ages, in ADHD and healthy volunteers, and where incentives were applied the methods differed. What is clear is from this convergent literature that a view of ADHD and treatment effects as increased or reduced activity in certain brain regions, or even networks, is no longer sufficient. Incentive context, network coherence and multivariate representations of data all need to be taken into account.

Correspondence to

  1. Top of page
  2. Correspondence to
  3. References

Mitul A. Mehta, Department of Neuroimaging, Centre for Neuroimaging Sciences, Institute of Psychiatry, King’s College London, London SE5 8AF, UK; Email:


  1. Top of page
  2. Correspondence to
  3. References
  • Kelly, C., De Zubicaray, G., Di Martino, A., Copland, D.A., Reiss, P.T., Klein, D.F., Castellanos, F.X., Milham, M.P., & McMahon, K. (2009). L-dopa modulates functional connectivity in striatal cognitive and motor networks: A double-blind placebo-controlled study. Journal of Neuroscience, 29, 73647378.
  • Liddle, E.B., Hollis, C., Batty, M.J., Groom, M.J., Totman, J.J., Liotti, M., Scerif, G., & Liddle, P.F. (2011). Task-related default mode network modulation and inhibitory control in ADHD: Effects of motivation and methylphenidate. Journal of Child Psychology and Psychiatry, 52, 761771.
  • Marquand, A.F., De Simoni, S., O’Daly, O.G., Williams, S.C.R., Mourao-Miranda, J., & Mehta, M.A. (in press). Pattern classification of working memory networks reveals differential effects of methylphenidate, atomoxetine and placebo in healthy volunteers. Neuropsychopharmacology.
  • Minzenberg, M.J., Yoon, J.H., & Carter, C.S. (in press). Modafinil modulation of the default mode network. Psychopharmacology (Berlin), December 14. [Epub ahead of print]
  • Sonuga-Barke, E.J., & Castellanos, F.X. (2007). Spontaneous attentional fluctuations in impaired states and pathological conditions: A neurobiological hypothesis. Neuroscience and Biobehavioral Reviews, 31, 977986.
  • Tomasi, D., Volkow, N.D., Wang, R., Telang, F., Wang, G.J., Chang, L., Ernst, T., & Fowler, J.S. (2009). Dopamine transporters in striatum correlate with deactivation in the default mode network during visuospatial attention. PLoS One, 4, e6102.