Changes in the Default Mode Network and Functional Connectivity

Changes in the Default Mode Network and
Functional Connectivity in Epilepsy

Ambica M. Tumkur, MD

Emory University School of Medicine, Atlanta, Georgia

Recent studies using novel imaging techniques are providing important new clues into the pathogenesis of epilepsy and the anatomic changes occurring in patients with the disease. Magnetic resonance imaging has provided greater understanding of the pathophysiology of epilepsy, and study of the default mode network has allowed more intense scrutiny of interictal functional disturbances. At the "Hot Topics" symposium held during the 68th Annual Meeting of the American Epilepsy Society, speakers delved into the research investigating the structural changes in cortical volume and white-matter connectivity seen in epilepsy patients and the functional consequences of these anatomic alterations.

Ambica M. Tumkur, MDEpilepsy is a prevalent neurologic disorder affecting 1%–3% of the general population and more than 2 million people.1–3 Epilepsy impedes upon a patient's personal, social, and psychologic well-being and wreaks financial havoc on the healthcare system, with the annual cost of epilepsy care topping $12.5 billion (in 2003 dollars).2

Years of research have heralded significant improvements in epilepsy treatment from both a pharmacologic and surgical perspective. New drug therapies and enhanced surgical and neurostimulatory options such as laser ablation, and responsive neurostimulation provide hope for a better quality of life for epilepsy patients, but they are far from a panacea. Despite these treatment advances, up to 30% of epilepsy patients endure ongoing seizure activity that devastates their daily functioning.

Given the high prevalence of epilepsy and the degree of disability experienced by many patients, organizers of the 68th Annual Meeting of the American Epilepsy Society (AES) meeting in Seattle, Washington, brought together preeminent epilepsy researchers and clinicians to discuss the latest studies, epilepsy models, and treatments available to help this population. Among other subjects, the "Hot Topics" session addressed a common and important question often asked by patients and their families: Can seizures—even infrequent ones—cause permanent brain damage? Jean Gotman, PhD, Professor of Neurology and Neurosurgery, and Andrea Bernasconi, MD, Associate Professor of Neurology at the Montreal Neurological Institute and Hospital in Montreal, Quebec, Canada, presented results derived from newer imaging modalities to examine the effect of epilepsy on the structure and function (connectivity) of the brain. In addition, they discussed the impact of epilepsy and seizures on resting functional networks. In particular, they examined the default mode network and reviewed data on structural changes in cortical volume and white-matter connectivity that have been observed in patients with epilepsy.

Traditionally, the pathophysiology of epilepsy has been linked to an "epileptic network," which has been compared with a forest fire, with seizures starting as a "spark" (ie, a focal epileptic discharge) in one region that quickly engulfs surrounding areas of the brain.

Table 1The introduction of new imaging techniques, such as functional magnetic resonance imaging (fMRI) and diffusion tension imaging (DTi), and improvements in scalp and intracranial electroencephalography (EEG) allow us to follow the epileptic brain better and track the propagation of epileptic activity with improved spatial and temporal resolution. These new techniques have changed the epilepsy community's understanding of the disorder. Gone is the isolated "epileptic network" paradigm of epilepsy pathophysiology. Synchronous use of fMRI and EEG has shown that the impact of focal epileptic discharges/dysfunction is not limited to a localized region. Instead, it involves a vast array of interconnected distant networks related to a multitude of functions that include attention, cognition, and memory.4–6

Examining Functional Networks
The brain is composed of billions of neurons that synapse on each other to create an intricate web/schema of connections. Research performed during the 1990s demonstrated that within this web, there are discrete connections between different regions of the brain called functional networks.

In 1995, Biswal and colleagues7 introduced the concept of resting-state networks. These discrete groups of neurons demonstrate synchronized activity while the patient is at rest but are deactivated during the engagement of a task.8 There are many kinds of resting-state networks, including motor and sensory networks and networks involved in attention, reward, language, and executive control (Table 1).5 Out of these resting-state networks, the default mode network has been identified as the most consistent and prominent (Figure 1).9

Figure 1

FIGURE 1 Composite positron emission tomography scan using population-averaged values from 132 participants to demonstrate areas of the default mode network. The blue areas represent the most active areas during passive tasks (ie, in the resting state). Reproduced, with permission, from Buckner et al.9

Raichle and others10 used 38 normal healthy subjects to examine functional changes in the activity of different brain areas during various tasks. Investigators used positron emission tomography (PET; Figure 2)10 and fMRI with blood oxygen level–dependent (BOLD) signaling. PET imaging measures changes in metabolic activity in different brain regions using glucose uptake markers, whereas fMRI with BOLD signaling uses differences in magnetic susceptibility for oxygenated and deoxygenated hemoglobin to measure increased cerebral activity and blood flow during the performance of specific tasks. While patients were engaged in specific goal-directed cognitive tasks, brain imaging demonstrated a predictable increase from baseline in cerebral blood flow in regions correlating with that specific task. Simultaneously, other brain regions exhibited a consistent, predictable decrease from baseline levels in blood flow as patients worked on a task. Thus, whereas an action increased blood flow and "activation" in one region, it resulted in decreased blood flow and "deactivation" in other regions.

Figure 2

FIGURE 2 Positron emission tomography scan images acquired and averaged from nine test subjects showing areas of deactivation during attention-demanding cognitive tasks. Reproduced, with permission, from Raichle et al.10

To ensure that the blood-flow changes seen on BOLD fMRI represented decreased cerebral activity and not just an artifact of the imaging technique, Kunii et al11 subsequently showed that these decreases in cerebral blood flow coincided with diminished fast (γ) activity, as seen on EEG, in those specific brain regions. These findings demonstrated that changes seen on BOLD fMRI were not just an artifact from shunted cerebral blood flow. Instead, they represented a change in neuronal activity, with decreased BOLD signaling corresponding with deactivation or slowing of these regions as seen on EEG.

Identifying the Default Mode Network (DMN)
The most prominent areas of deactivation on BOLD fMRI were in the posterior cingulate cortex, the precuneus, and the medial inferior frontal cortex.10 Deactivation does not depend upon a particular task but is present during any external task requiring attention and cognitive demand. Therefore, during the normal resting state (when the patient is awake/alert but not engaged in an attention-demanding task), neurons in these areas demonstrate continuous baseline tonic activity, thus creating a default level of activity in the brain. However, when the subject is engaged in a cognitively/attention-demanding task, these areas become deactivated as the baseline tonic activity ceases. Later studies would refer to this tonic activity as the DMN.4–6,10

Further research into the DMN conducted by Zhang et al12 expanded the regions to include the medial, lateral, and inferior parietal cortices. The precise function of the DMN is not completely clear, but it may play a role in higher-order processes such as cognition, affective behavior, attention, and consciousness.5 These roles were elucidated through PET and fMRI studies looking at DMN dysfunction in other neuropsychiatric disorders (Alzheimer's disease, schizophrenia, and depression). Patients with these conditions demonstrate decreased connectivity and activation of the DMN regions.9,13 Areas of the DMN (particularly the posterior cingulate cortex and precuneus) have a high baseline metabolic rate, which is consistent with continuous tonic activity and is vulnerable to ischemia and easily affected by vascular injuries.10

The foregoing research demonstrates that the normal brain has an interconnected set of networks designed to perform specific tasks, even when the individual is at rest. These networks have varying levels of baseline activity depending upon the person's degree of mental and physical activity. But how do these resting networks respond when there is aberrant activity (ie, interictal epileptiform discharges or seizures) in the brain? Do these networks change or adapt?

Mapping Discharges in the Epileptic Brain
Fahoum and colleagues4 investigated the impact of epileptic discharges emerging from various cortical regions, including the temporal lobe, supplementary motor area, temporoparietal area, and inferior parietal area, on the DMN. The researchers used BOLD fMRI and single-photon emission computed tomography (SPECT) to spatially assess cerebral function of the DMN. In addition, they used intracranial EEG monitoring, with electrodes implanted within the regions of the DMN, to analyze electrical activity in these regions simultaneously.

In all, five of six patients exhibited a decrease in BOLD fMRI and SPECT signaling, signifying deactivation in the DMN regions during a seizure. The most frequent areas of deactivation were in the precuneus lobule and the inferior parietal lobule. Deactivation also was noted to a lesser extent in the posterior cingulate cortex, the dorsolateral prefrontal cortex, and the medial prefrontal cortex. The intracranial EEG recordings from electrode contacts placed within the DMN also showed lesser power of faster (γ) frequencies and increased δ slowing in the DMN during the course of the patient's seizure. This confirmed that the BOLD fMRI/SPECT signal changes correlated with changes in functional neuronal activity in the brain. Areas outside the DMN did not demonstrate any imaging or EEG changes.4

The Effect of Seizures on Baseline Brain Connectivity
Another major area of research involves the effect of focal-onset seizures (particularly temporal-lobe seizures) on baseline connectivity between different brain regions and on the DMN.1,14,15 Temporal-lobe seizures are of particular interest because they represent the most common type of medically refractory epilepsy.

Voets et al15 and Haneef et al16 discovered that patients with temporal-lobe epilepsy have less connectivity and activation of the posterior DMN, even at baseline. Whether the etiology behind the decreased connectivity in temporal-lobe seizures is secondary to the degree of hippocampal atrophy (mesial temporal sclerosis) often seen in affected patients is debatable.14 However, even after adjusting for gray-matter variability (as manifested by different degrees of gray-matter thickness and mesial temporal sclerosis), MRI with DTi still showed decreased structural connectivity between the temporal lobes and regions in the DMN (Figure 3).14

Figure 3

FIGURE 3 Magnetic resonance imaging/diffusion tensor imaging tractography used to compare white-matter tracts in a control individual with those in a patient with mesial temporal-lobe epilepsy. Note the diminished connectivity between the posterior cingulate cortex, the precuneus, the medial prefrontal cortex, and the medial temporal lobes. Reproduced, with permission, from Liao et al.14

These changes in white matter may be at the heart of DMN dysfunction—there is significant correlation between these structural changes and functional activation of these networks.14 According to Zhang and others,12 for a neuronal network to function properly, the network activity depends upon the functional integrity of all the nodes (sites) within the network. Even small alterations in the "connections" or "nodes" of these networks may lead to dysfunction of the entire network.12 The degree of structural changes and connectivity problems is directly proportional to the severity and duration of a patient's epilepsy.1,5

The Effect of Generalized Discharges
Much of the research into the effects of epilepsy on the DMN has focused on focal epilepsies (especially temporal-lobe epilepsy). Gotman et al6 looked at the effects of generalized discharges on both thalamocortical structure and the DMN in the brain. In this study, as well as in pediatric research conducted by Moeller et al,17 patients with generalized epilepsies were monitored with fMRI and scalp EEG recordings. During generalized epileptic discharges, there was activation of the bilateral thalamus and, particularly, of the medial thalamic structures. Simultaneously, there was deactivation in the bilateral anterior frontal and parietal regions and in the posterior cingulate gyrus. The regions of deactivation correlated with areas within the DMN.

In follow-up studies, Luo and colleagues13 used BOLD fMRI to demonstrate decreased regional functional connectivity in the frontal, parietal, and temporal lobes in patients with generalized epilepsies, even at baseline and in the absence of any seizures or interictal discharges (Figure 4).

Figure 4

FIGURE 4 Schematic representation of the connectivity between different networks, showing decreased connectivity to the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), and bilateral temporal lobes in patients with generalized epilepsies. Adapted, with permission, from Luo et al.13

Kay et al18 investigated how treatment of epilepsy affected the connectivity in the DMN. They compared the degree of DMN connectivity among patients with generalized epilepsy who responded or did not respond to valproic acid. Treatment-resistant patients had decreased DMN connectivity when compared with responders; thus, treatment resistance was more highly associated with a greater reduction in functional connectivity than with epilepsy alone (Figure 5).18

Figure 5

FIGURE 5 Default mode network (DMN) connectivity was significantly lower in patients with generalized epilepsy whose seizures were uncontrolled (seizures+) or were valproate unresponsive (VPA–). Patients who had well-controlled epilepsy (seizures–) and/or were valproate responsive (VPA+) did not demonstrate a significant difference in DMN connectivity from the control group, as measured by (a) independent component analysis and dual regression or (b) seed-based voxel correction. Adapted, with permission, from Kay et al.18

These data suggest that ongoing seizure activity confers continued damage within the DMN and propagates and worsens connectivity issues. Previously discussed studies by Gotman et al6 and Luo et al13 also demonstrated that the degree of deactivation and loss of connectivity of the DMN seemed to be based on the severity of epilepsy, as reflected by the duration of epilepsy and treatment resistance.18

So why should the epilepsy community care about resting-mode networks, particularly the DMN? Resting-mode networks may play a crucial role in maintaining consciousness, alertness, and general cognitive function. Disruption of the DMN may account, in part, for the consciousness/awareness that many patients experience with focal-onset epilepsies. In generalized epilepsies, suspension/disruption of the DMN may dampen perceptual sensory inputs and decrease the ability of activation from the thalamus and mid-frontal regions. This, in turn, plays a role in the altered consciousness and reduced sensory awareness seen in generalized epilepsies.13

In addition to disruption of the DMN, other resting-mode networks also seem to exhibit dysfunction during epileptic discharges. These networks include the attentional, executive control, and reward networks that affect attention, problem-solving, and mood. All of these problems are well-known comorbidities in epilepsy patients. Disabilities such as memory loss and slowed cognition that once were attributed to structural defects and medication effects actually may be, at least partially, secondary to disruption of these resting-mode networks.

These findings provide epileptologists with greater insight into the long-term effects of seizures and ongoing epileptic discharges on brain structure and function. They also reinforce the importance of aggressive seizure management from both a pharmacologic and surgical perspective and of seizure elimination—not just seizure control—to help minimize the deleterious long-term effects of epilepsy.


  1. Bernhardt BC, Hong S, Bernasconi A, Bernasconi N. Imaging structural and functional brain networks in temporal lobe epilepsy. Front Hum Neurosci. 2013;7:1–14.
  2. Chang BS, Lowenstein DH. Epilepsy. N Engl J Med. 2003;349:1257–1266.
  3. Wilner AN, Sharma BK, Thompson A, Soucy A, Krueger A. Diagnoses, procedures, drug utilization comorbidities and cost of health care for people with epilepsy in 2012. Epilepsy Behav. 2014;41:83–90.
  4. Fahoum F, Zelman R, Tyvaert L, Dubeau F, Gotman J. Epileptic discharges affect the default mode network—fMRI and intracerebral EEG evidence. PLoS One. 2013;8:e68038.
  5. Cataldi M, Avoli M, de Villers-Sidani E. Resting state networks in temporal lobe epilepsy. Epilepsia. 2013;54:2048–2059.
  6. Gotman J, Grova C, Bagshaw A, Kobayashi E, Aghakhani Y, Dubeau F. Generalized epileptic discharges show thalamocortical activation and suspension of the default state of the brain. Proc Natl Acad Sci U S A. 2005;102:15236–15240.
  7. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn Reson Med. 1995;34:537–541.
  8. Archer JS, Abbott DF, Waites AB, Jackson GD. fMRI "deactivation" of posterior cingulate during generalized spike and wave. Neuroimage. 2003;20:1915–1922.
  9. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain's default network. Ann N Y Acad Sci. 2008;1124:1–38.
  10. Raichle ME, Macleod AM, Snyder AZ, Powers WJ, Gusnard DA, Shulman GL. A default mode of brain function. Proc Natl Acad Sci U S A. 2001;98:676–682.
  11. Kunii N, Kamada K, Ota T, Kawai K, Saito N. Characteristic profiles of high gamma activity and blood oxygenation level-dependent responses in various language areas. Neuroimage. 2013;65:242–249.
  12. Zhang Z, Lu G, Zhong Y, et al. Altered spontaneous neuronal activity of the default mode network in mesial temporal lobe epilepsy. Brain Res. 2010;1323:152–160.
  13. Luo C, Li Q, Lai Y, et al. Altered functional connectivity in default mode network in absence epilepsy: a resting state fMRI study. Hum Brain Mapp. 2011;32:438–449.
  14. Liao W, Zhan Z, Pan Z, et al. Default mode network abnormalities in mesial temporal lobe epilepsy: a study combining fMRI and DTI. Hum Brain Mapp. 2011;32:883–895.
  15. Voets NL, Beckman CF, Cole DM, Hong S, Bernasconi A, Bernasconi N. Structural substrates for resting network disruption in temporal lobe epilepsy. Brain. 2012;135:2350–2357.
  16. Haneef Z, Lenartowicz A, Yeh HJ, Levin HS, Engel J, Stern JM. Functional connectivity of hippocampal networks in temporal lobe epilepsy. Epilepsia. 2014;55:137–145.
  17. Moeller F, Sieber HR, Wolff S, et al. Changes in activity of striato-thalamo-cortical network precede generalized spike wave discharges. Neuroimage. 2008;39:1839–1849.
  18. Kay BP, DiFrancesco MW, Privitera MD, Gotman J, Holland SK, Szaflarski JP. Reduced default mode network connectivity in treatment resistant idiopathic generalized epilepsy. Epilepsia. 2013;54:461–470.

Dr. Tumkur is a Clinical Neurophysiology Fellow in the Department of Neurology, Emory University School of Medicine, Atlanta, Georgia.

View/download a PDF version of this article