Mirror, Mirror In The Brain: Mirror Neurons, Self-understanding And Autism Research

Some scientists speculate that a mirror system in people forms the basis for social behavior, for our ability to imitate, acquire language, and show empathy and understanding. It also may have played a role in the evolution of speech. Mirror neurons were so named because, by firing both when an animal acts and when it simply watches the same action, they were thought to "mirror" movement, as though the observer itself were acting.

Advances in the past few years have newly defined different types of mirror neurons in monkeys and shown how finely tuned these subsets of mirror neurons can be. New studies also have further characterized abnormal-as well as normal-mirror activity in the brains of children with the social communication disorder known as autism, suggesting new approaches to treatment.

"The tremendous excitement that has been generated in the field by the study of mirror neurons stems from the implications of the findings, which have led to numerous new hypotheses about behavior, human evolution, and neurodevelopmental disorders," says Mahlon DeLong, MD, of Emory University School of Medicine.

Mirror neurons, a class of nerve cells in areas of the brain relaying signals for planning movement and carrying it out, were discovered 11 years ago, an offshoot of studies examining hand and mouth movements in monkeys. Mirror neuron research in the intervening years has expanded into a diverse array of fields. And the implications have been enormous, encompassing evolutionary development, theories of self and mind, and treatments for schizophrenia and stroke.

Findings being presented at Neuroscience 2007 include new research based on work in monkeys, showing that subsets of mirror neurons distinguish between observed actions carried out within hand's reach and those beyond the animal's personal space.

In his study, Peter Thier, PhD, at Tübingen University, first identified a group of mirror neurons by recording single nerve cell activity from electrodes when a monkey gripped different objects and when the monkey watched a person grasp the same objects, both nearby and farther away. About half of the nerve cells that were active when the monkey picked up the objects also sprung into action when it watched a person do so. Thier was assisted by research fellow Antonio Casile and PhD student Vittorio Caggiano, and worked closely with the lab of Giacomo Rizzolatti, MD, at the University of Parma.

They also noticed that some of these confirmed mirror neurons were active only when the monkey was watching activity within its personal space, defined as within reaching distance; others responded only to actions performed in a place outside the monkey's grasp. Thier and colleagues recorded this preferential activity in 22 nerve cells, or together half of the mirror neurons. The other half of the mirror neurons showed activity that did not depend on how close the grasping action was to the monkey.

Although at this stage assigning a functional role is still speculation, Thier suggests this proximity-specific activity in mirror neurons may play an important role when we monitor what goes on around us, or serve as the basis for inferring the intentions of others and for cooperative behavior. "These neurons might encode actions of others that the observers might directly influence, or with which he or she can interact," he says.

Other findings show that mirror neuron activity is instrumental for interpreting the facial expressions and actions of others but may not be sufficient for decoding their thoughts and intentions.

The studies examined changes in certain electroencephalograms (EEG) or brain wave patterns known as mu rhythms, which have a frequency of 8-13 hertz, or oscillations per second. Previous findings based on EEG recordings from the part of the brain that is directly involved in relaying signals for movement and sensing stimuli, known as the sensorimotor cortex, indicate that mu rhythms typically are suppressed by mirror activity in premotor areas of the brain. However, this does not happen in children with autism. As a result, the new work suggests, alternative strategies for reading faces and understanding others develop in the brains of these children.

Pursuing two parallel studies, Jaime Pineda, PhD, at the University of California, San Diego, aimed to contribute evidence supporting one of two theories about the ways we evaluate the actions and intentions of other people-either implicitly or through language-based theoretical concepts.

Using EEG recordings to examine patterns of brain wave activity, Pineda first worked with 23 adults, who were asked to look at photos showing just the eye region of people making various facial expressions. In three separate trials, the subjects were asked to identify either the emotion, race, or gender of the people in the photographs. In a subsequent task, subjects looked at three-panel cartoon strips and were asked to choose a fourth panel that completed the strip-either the conclusion of a series of physical actions or the result of a person interacting with an object. A sequence of a prisoner removing the window of his cell, then looking at his bed, for example, could be followed by a frame of the prisoner asleep, yawning, or using the bedsheet to make a rope. Answering correctly depended on interpreting the cartoon character's intentions appropriately or understanding how physical objects interact.

Pineda repeated the studies with 28 children, 7 to 17 years old, half of whom had autism. The other half were typically developing children.

Recordings from the studies with adults showed a correlation between mu suppression, or mirror neuron activity, and accuracy for both tasks. In fact, the suppression of mu rhythms during the facial expression task also correlated with accuracy in the exercise with the cartoons, suggesting that reading people's expressions and interpreting their intentions may draw from similar activity in the brain.

Recordings from the typically developing children showed similar patterns of suppression during the two tasks, indicating that mirror neuron activity is fully developed by age 7.

In contrast, recordings from the children with autism showed that mu rhythms were enhanced during both tasks. Enhancement is an indication that the mirror neuron system is disengaged. However, because the children still were able to perform the task, Pineda says, "we propose that children with autism develop alternative, non-mirror neuron-based coping strategies for understanding facial expressions and interpreting others' mental states." He suggests that "these compensatory strategies involve inhibition of residual mirror neuron functioning."

These results could be applied to the development of treatments for autism. Pineda and his group have been using neurofeedback training to successfully renormalize functioning in this system. That is, they see mu suppression that is more characteristic of the typically developing brain following such training. "Our findings are consistent with the idea that mirror neurons are not absent in autism," Pineda says, "but rather are abnormally responsive to stimuli and abnormally integrated into wider social-cognitive brain circuits.

"This idea implies that a retraining of mirror neurons to respond appropriately to stimuli and integrate normally into wider circuits may reduce the social symptoms of autism."

Advances in recording brain activity also have made possible findings showing that mirror systems are active even when we are not observing an action with an eye to repeating it.

Suresh Muthukumaraswamy, PhD, at Cardiff University, found that the mirror system is activated when we watch specific actions, even when we are concentrating on a separate task.

The results are based on previous research showing that motor systems in the brain are activated when a person observes an action being performed and on interpretations suggesting that we understand and learn to imitate the actions of others through these brain mechanisms.

Working with 13 adults with an average age of 29, Muthukumaraswamy compared brain activity recorded via magnetoencephalography (MEG). This monitoring technique measures the weak magnetic fields emitted by nerve cells, and, recording from 275 locations, Muthukumaraswamy was able to monitor changes in activity every 600th of a second.

"Although MEG has been in existence for more than 20 years, recent advances in hardware, computing technology, and the algorithms used to analyze the data allow much more detailed analysis of brain function than was previously possible," he says.

Brain activity was recorded as the subjects passively watched a sequence of finger movements, watched the movements knowing they would be asked to repeat them, added up the number of fingers moved as they watched, and performed the sequence of movements themselves.

Results from these recordings showed similar activity when the subjects performed the movement sequence and when they watched someone else do it. In addition, Muthukumaraswamy noted increased activity in areas of the brain regulating motor activity when subjects observed the movements knowing they would later do them, and when they added up the number of fingers used, compared with passive watching.

"These data suggest that activity of human mirror neuron systems is generally increased by attention relative to passive observation, even if that attention is not directed toward a specific motor activity," says Muthukumaraswamy. "Our results suggest that the mirror system remains active regardless of any concurrent task and hence is probably an automatic system.

"A good scientific understanding of the properties of the mirror system in normal humans is important," he adds, "because this may help to understand clinical disorders such as autism where the mirror system may not be functioning normally."

Other findings based on EEG recordings provide the first evidence of normal mirror activity in children with autism: People familiar to children with autism may activate mirror areas of the brain in normal patterns when unfamiliar people do not.

Previous research has shown that mu rhythms are suppressed when a subject identifies with an active person being observed. Based on this work, Lindsay Oberman, PhD, at the University of California, San Diego, examined the role of two separate factors in the mirror system response of children with autism.

Six videos were shown to a group of 26 boys, 8 to 12 years old; half had autism. Three videos showed images representing varying degrees of social interaction: two bouncing balls (the baseline measurement), three people tossing a ball to themselves, and three people throwing the ball to each other and off the screen to the viewer. The other set of videos showed people with varying degrees of familiarity to the subjects: strangers opening and closing their hand, family members making the same hand movement, and the subjects themselves doing the same.

EEG recordings from 13 electrodes in a cap showed that mu activity was suppressed most when subjects watched videos of themselves, indicating the greatest mirror neuron activity. For both groups, the measurements showed a slightly lower level of suppression when subjects watched familiar people in the video and the least when watching strangers. This indicates that normal mirror neuron activity was evoked when children with autism watched family members, but not strangers.

"Thus, to say that the mirror neuron system is nonfunctional may only be partially correct," says Oberman. "Perhaps individuals with autism have fewer mirror neurons and/or less functional mirror neurons that require a greater degree of activation than a typical child's system in order to respond."

The mirror neuron system may react to stimuli that the observer sees as "like me." If this is the case, suggests Oberman, "perhaps typical individuals apply this identification to all people (both familiar and unfamiliar), resulting in activation of these areas in response to the observed stimuli, while individuals on the autism spectrum only consider familiar individuals (including themselves) as 'like me,' " she says.

This evidence for normal mirror neuron activity in autistic children may indicate that mirror system dysfunction in these cases reflects an impairment in identifying with and assigning personal significance to unfamiliar people and things, Oberman suggests. Whether deficits in relating to unfamiliar people that are characteristic of autism are the cause or the result of a dysfunctional mirror neuron system is unclear.