Vegetative and minimally conscious patients can learn

THE vegetative and minimally conscious states are examples of what are referred to as disorders of consciousness. Patients in these conditions are more or less oblivious to goings-on in their surroundings - they exhibit few, if any, signs of conscious awareness, and are usually unable to communicate in any way. It is, therefore, extremely difficult to establish what these patients are experiencing, and the consciousness disorders are among the least understood, and most commonly misdiagnosed, conditions in medicine.

Although technologies such as functional neuorimaging have enabled clinicians to gain some insight into these conditions, proper assessment and diagnosis of patients are still major challenges, and there are big ethical questions regarding how they should be treated. However, researchers from the University of Cambridge have made what could be a significant advance.

In a paper published online in Nature Neuroscience, they report that some minimally conscious and vegetative patients appear to be capable of learning simple associations between two stimuli, and that this ability is strongly correlated with subsequent recovery. The findings point to a reliable method for evaluating and diagnosing patients suffering from consciousness disorders, and also may enable clinicians and relatives to make informed decisions about treatment. 

Disorders of consciousness are typically defined by behavioural assessment - patients in the so-called minimally conscious state exhibit intermittent evidence of awareness, while those in the vegetative state exhibit no awareness whatsoever. Most patients are usually unresponsive to external stimuli, but this gives no indication of the state of brain function, or of cognitive processes. It is clear that the behavioural criteria for diagnosis of these disorders are inadequate, and it is estimated that approximately 40% of patients have been misdiagnosed

It is also now clear that at least some patients are aware of what is going on around them, despite showing no outward signs of awareness. Several neuroimaging studies have shown that auditory regions in the brains of vegetative patients become activated in response to speech. And another study provided evidence of specific brain activation patterns in a vegetative patient when she was asked to imagine playing tennis or moving around her house.

In some cases, the prognosis, or outcome, of such patients, is inaccurate too. It was also long assumed that patients who have existed in such conditions for long periods cannot recover. However, it is now clear that is not always the case. In 2007 researchers used an experimental surgical procedure called deep brain stimulation to improve brain function in a patient who had been in a minimally conscious state for more than six years. Almost immediately, the patient opened his eyes, and responded to voices. In the following months, he became capable of speaking, swallowing and reaising a cup to his mouth. Then there is the remarkable case of Terry Wallis, who went into a vegetative state following an accident in 1984, only to emerge from it 19 years, despite doctors' insistence that he would never recover.

In the new study, Tristan Bekinschtein of the MRC Cognition and Brain Sciences Unit at the University of Cambridge and his colleagues used classical conditioning to test whether any traces of conscious processing are preserved in minimally conscious and vegetative patients. This procedure is also referred to as Pavlovian conditioning, after the Russian physiologist Ivan Pavlov, who first showed, over 100 years, ago that dogs quickly learn to associate the sound of a ringing bell with the presentation of food, so that after a number of pairings of the two stimuli, they begin to salivate in anticipation of being fed when presented with the bell alone. 

The researchers used the same procedure with 22 patients, all of whom had been in a minimally conscious or vegetative state, for at least 6 months. In 70 learning trials, the patients were presented with a tone, followed half a second later by a puff of air into the eye, which produces a reflexive blinking response. In 70 subsequent trials, the tone was presented alone. This was also carried out on 16 healthy, conscious controls, and on 12 patients undergoing a standard medical procedure, during which they were rendered unconscious by the anaesthetic propofol. 

If an association between the two stimuli has been learned, the tone should elicit a blinking response when presented by itself during the second block of 70 trials. Learning was assessed using event-related potentials (ERPs), by which electrical activity in the brain associated with specific events is measured with electrodes placed on the scalp. An electromyograph was also used to monitor the electrical activity of the eye muscles. Learning in this case would be associated with an increase in activity in specified electrodes during the anticipatory period - the 500 milliseconds after the tone was presented - and peaking at the time at which the puff of air was expected, as well as electrical activity in the eye muscles, which is correlated with the blink response.

These effects were recorded in the healthy conscious control group and in some of the experimental group of vegetative and minimally conscious patients, but not in the unconscious controls who had been under anaesthesia during the task. Significantly, learning effects were observed in both vegetative and minimally conscious patients. From their results, the researchers could predict, with an accuracy of more than 70%, which of the patients had been diagnosed as being in minimally conscious state and which were in a vegetative state. They could also predict, with even greater accuracy, which had suffered a traumatic brain injury, and which had incurred their consciousness disorder as a result of prolonged oxygen deprivation. Furthermore, it was found that learning was accurate predictor of future recovery - the condition of patients who showed signs of learning subsequently improved much more than those patients in whom the effects were not observed, as measured by the Revised Coma Recovery Scale.  

These results can be interpreted in two ways. One interpretation is that patients with consciousness disorders still retain conscious processing to some extent. The other is that the classical conditioning demonstrated in the experimental group can occur in the absence of consciousness. The fact that the anaesthetized patients in the second control group did not exhibit associative learning suggests that the first interpretation is the more likely one. Either way, the findings suggest that those in the experimental group who did show signs of learning had properly functioning medial temporal lobe structures, such as the hippocampus and surrounding structures, which are crucial for memory formation and in learning the time interval between two stimuli during classical conditioning.

The findings of this study have important implications for the treatment of patients with disorders of consciousness. First, testing for signs of associative learning could help clinicians to distinguish between the different states, and to assess patients' levels of arousal more reliably. The knowledge that learning effects predict future recovery could also provide a means of accurate prognosis, and help relatives of patients who have been in an a minimally conscious or vegetative state for long periods of time to make informed decisions based on realistic possible outcomes.     


Bekinschtein, T., et al. (2009). Classical conditioning in the vegetative and minimally conscious state. Nat. Neurosci. DOI: 10.1038/nn.2391


More like this

I caught this story (with an accompanying video) over at Mind Hacks and Neuron Culture about this poor woman from the UK who fell unconscious from a viral infection in her brain stem. Using fMRI, a doctor at Cambridge named Adrian Owen showed that she still had residual brain functions such as…
Researchers report in today's issue of Nature that they have improved brain function in a minimally conscious patient by implanting electrodes into his brain. Schiff et al used deep brain stimulation (DBS), an experimental surgical technique that has previously been used to treat Parkinson's…
I talked earlier this year about a patient who recovered from a coma after 20 years. In that post, I discussed how -- with respect to the diagnostic criteria -- the difference between a persistent vegetative state and a minimally conscious state is the difference between someone who has only…
In The New Yorker, Jerome Groopman discusses the work of Adrian Owen, a researcher at Cambridge University's Cognitive and Brain Sciences Unit who has been using functional imaging to assess patients in a vegetative state. Neurologists face major problems in diagnosing the persistent vegetative…

These results can be interpreted in two ways. One interpretation is that patients with consciousness disorders still retain conscious processing to some extent. The other is that the classical conditioning demonstrated in the experimental group can occur in the absence of consciousness.

Poorly understood? Yes.

Among the most commonly diagnosed conditions in medicine? Hardly!

By Epictetus (not verified) on 21 Sep 2009 #permalink

Thank you for this great write-up. What a wonderful discovery, and with such simple methods!

By Jan-Maarten (not verified) on 22 Sep 2009 #permalink

Very interesting stuff from a clinical perspective. However, the claims about consciousness are not justified by the data.

The major problem is that there is good evidence that Pavlovian conditioning doesn't require consciousness. Pavlovian conditioning can be evoked in the body of a spinalized cat(that is, in a preparation where the spinal cord is severed from the brain stem, the body still learns). Less conclusively, invertebrates like snails learn associations via Pavlovian conditioning, and one could argue they are not conscious.

We must applaud that they included any controls, which makes their study better than the one where 'Play tennis' responses were observed.

On the other hand, the results from the anesthesia control are also very hard to interpret. It isn't as if anesthesia abolishes only consciousness: the entire brain is affected by general anesthesia. Even reflexes can be turned off by general anesthesia (indeed, in rats one of the checks that we have them under deeply enough is to pinch their tail to make sure the withdrawal reflex is no longer activated). Clearly, basic reflexes don't require consciousness or even brains (again, we can evoke reflexes just in spinal cord preparations). Hence using anesthesia as a control could be like pulling the plug on a computer to show that computers without a fan can't do computations.

On the third hand, I must admit I don't know what controls would be better. One avenue to explore is that not all general anesthetics are the same. Some seem to turn off consciousness by basically pulling the plug on most of the brain (an imprecise control). Others seem to act on specific nuclei in the brain stem. They used propofol, and my quick PubMed search suggests not much is known about its effects at the biophysical level.

Clearly, the most exciting conclusion is the clinical correlation between learning and recovery of consciousness. As someone noted its simplicity is beautiful. It should be easy to automate with cheap medical supplies. They note in the paper that recovery is also correlated with degree of cortical damage, but that isn't necessarily a cheap or easy thing to ascertain.

I think Terry Shaivo's cortical damage was fairly complete, so even using these results her diagnosis seems accurate.

Durkovic and Damianopoulos (1986) Forward and backward classical conditioning of the flexion reflex in the spinal cat, Journal of Neuroscience, Vol 6, 2921-2925.

Quite right, Eric. The authors actually mention the fact that invertebrates can be classically conditioned, even though they do not possess anything like human consciousness. I didn't go into any detail about whether conditioned responses require conscious awareness (perhaps I should have), because, as you say, the most interesting aspect of the study is the correlation between conditioning and subsequent recovery.

Thanks for the comment.

This study is cool no doubt if for no other reason than it suggests a degree of cortical processing that predicts recovery. I want to find some more support for the "retains some consciousness" proposal but all i recognize is the possibility that the degree of cortical activity for certain patients in vegetative states extends to the capacity to develop nondeclarative memory. My question: what is new here?

Previous studies have shown there can be a significant degree of cortical process for patients in a vegetative state, e.g. (D. K. Meno et al, 1998), a study examining reactions to familiar faces. It is easy to take this study as an indication that objects can be represented even under such debilitative conditions (Damasio, somewhere).

The formation of a memory is more involved, perhaps, than the recognition of a face. Other studies suggest that memories formed from changes in the perceptual and response systems can be developed from subliminal signals or by amnesic patients with damage to hippocampal formations, an eye blink study by Clark and Squire (1998): memories formed without awareness. In other words, awareness is not necessary for nondeclarative memory formation. It is fair to say that a vegetative state is an extreme form on nonawareness.

I somewhat understand a bit of this. I have a 19yr daughter that fell 27feet into concrete and most oh her injury was all in her brain. The doctors of course give no hope at all they say that she now just needs LTCF. As a mother I want her in home care they wont let me. How is a mother to help her daughter that is in this type of state as you describe. What can i do to help my daughter, cause no one is doing nothing no more for her??
But lots of helpful information for me. some parts a bit confusing. must put in plain English for folks like me. Just a working mother going thru this type of Brain Injury

By Emerald Banda (not verified) on 06 Oct 2009 #permalink

Emerald: I'm so sorry to hear about your daughter. That must be the most awful for a parent to hear. I'm not qualified to give you any medical advice, but unfortunately I think there's very little that you can do to help her.

The most important thing for you to understand about this study is that the tests accurately predict whether there will be any improvement at all in a patient's condition at some point in the future. So you might want to consider asking the doctors who are taking care of your daughter if they know about this study, and perhaps to have them carry out the tests on her.

This may not be a direct link to the memory test but I think it still helps connect some of the dots.
I have seen people in many states of consciousness and have been "unconscious" but aware many ,many times myself. People in the room with me saying all manner of things about me ,even telling others not to worry I could not hear them or would not remember anyway. I corrected them and scared quite a few.Any kind of test is believable to me but I am an easy subject to convince.
A few weeks ago my daughter's grand father and my former father in law was in a "vegetative" state in an I.C.U. in hospice after 2 brain surgeries and a stroke after the second surgery left him paralyzed and unresponsive. He was over medicated and suppose to be "vegetative" and not responsive.
I told my daughter that is anyone was going to get him to giggle that it would be a cute grand daughter or a smart Alec former son in law. I whispered a dirty joke in his ear that also included a few words about his daughter and my former wife.
OOPS I set the monitors off and his heart rate went up and what ever the beepers and buzzers were for they worked.
The nurses came and checked and at the same time my daughter was whispering a little lullaby and sweet nothings in his other ear.
We spoke with the the doctor and head nurse and this was the first response from this man in around 5 days.
They laughed and scolded me a little bit.
For what ? My daughter and I stayed there all evening and just the sound of my voice would cause a slight movement on the non-paralyzed side and my daughter singing him little love songs caused his rates to level off.In a "vegetative" man that reacted to nothing else for days.
That was his last night alive and my daughter will always remember grandpa with a smile from her memories.


My 27 year old daughter is in her third week after suffering a TBI of an unexplained cause. She is a Type I brittle diabetic and now is in a nursing home in Nampa, Idaho. Other than a CT scan and EEG (pretty good EEG according to a neurologist), no one has done a brain study of her, no one has done anything to stimulate the brain, and no one in the 12 days at a regionail hospital did anything to monitor brain pressure.

My daughter was cognizant of me from the first day. She is a valuable person, four classes away from her masters degree, and I firmly believe that someday she will return home.

I understand how frustrating Mrs. Banda feels right now and wish I could talk to her face to face. Medicine has become a political ballgame and the patients suffer in the end. Since my daughter was taken to the hospital that did NOT have a brain injury clinic, she could not be transferred to the facilities connected to the hospital that DOES have a brain injury clinic.

By Pamela Dowd (not verified) on 14 Jun 2010 #permalink