The Limits of Neuroplasticity

By Neuroskeptic | November 13, 2010 7:43 pm

Neuroplasticity is in.

Books tell us about The Brain That Changes Itself or advise us on how to Train Your Mind, Change Your Brain.

Now there’s no doubt that the brain is plastic, able to rewire itself in response to damage or training, and that it’s more so than was generally believed, say, 20 years ago. It’s clearly an important and interesting field, but a little caution is warranted. Neuroplasticity can’t fix everything.

If the brain were infinitely plastic, brain damage would be no big deal. You’d get over it pretty quickly, so long as some of your brain was intact and able to rewire itself to compensate. Unfortunately, that’s rarely what happens. Well, unfortunately unless you’re a neurologist; they’d be out of a job if it were otherwise…

Swiss neurologists Bindschaedler et al have provided a nice example of the limits of neuroplasticity in a new paper: Growing up with bilateral hippocampal atrophy: From childhood to teenage.

Patient “VJ” was diagnosed with bilateral atrophy of the hippocampus at age 8. The damage almost certainly dated back a few hours after his birth which followed a normal pregnancy and delivery:

VJ had convulsions, then an episode of apnoea, which required treatment by phenobarbital and mask-assisted ventilation respectively. Hypertonia and hyperreactivity were followed by a period of hypotonia and somnolence during the 8 following days.

He seemed to make a full recovery, and never suffered another seizure. However, as he grew up, his parents noticed that he was forgetful and had difficulty concentrating. At the age of 8, he was referred for an MRI scan, which revealed severe atrophy of the hippocampus, and the related structures the fornix and the mammilary bodies, on both sides of the brain.

The diagnosis was hypoxic-ischaemic encephalopathy: a lack of oxygen. For some reason, the hippocampus is especially vulnerable to this; selective hippocampal damage is also common after carbon monoxide poisoning.

VJ is a bit like a childhood version of the famous adult patient HM, however, there are important differences. Apart from occurring much earlier, obviously, VJ’s damage was less severe. Unlike HM he did not lack the nearby entorhinal cortex or parahippocampal cortex.

The authors followed him up to age 17, and did lots of tests of his cognitive function. The pattern that emerged is that VJ showed a selective impairment of memory for personal events (episodic memory). He learned to read normally, and he scored well on tests of general knowledge. So he had preserved semantic memory, memory for facts. His IQ was normal.

Even his episodic memory impairment was selective, however. He was severely impaired on tests of memory recall, i.e. “What did you do yesterday?”. But on tests of recognition – “Have you ever seen this picture before?” – he did perfectly well.

This fits with lots of previous studies showing that the hippocampus is required for recall while the nearby cortex is more important for recognition. When asked to describe events in his past, he was essentially unable to do so, unless he was provided with “clues” or “reminders” to trigger recognition.

So VJ’s brain couldn’t rewire itself to compensate for the lack of a hippocampus, despite the fact that the damage occurred at birth, and the brain is considered to be at its most plastic during childhood.

This is not all that surprising really. The hippocampus is a unique region, containing specialized circuitry which is just not found anywhere else in the brain. Most of the evidence for large-scale neuroplasticity concerns the cerebral cortex. When part of the cortex is damaged, other cortical areas can sometimes compensate for the loss: but the cortex can’t turn itself a substitute hippocampus.

ResearchBlogging.orgBindschaedler, C., Peter-Favre, C., Maeder, P., Hirsbrunner, T., & Clarke, S. (2010). Growing up with bilateral hippocampal atrophy: From childhood to teenage Cortex DOI: 10.1016/j.cortex.2010.09.005

  • spike

    So 2 points:
    1) Is there a quantitative electrophysiological marker of neuroplasticity? (perhaps a stupid question!)
    2) If the answer to 1 is yes, or can be yes, do you think that the metric would have a uniform distribution over the entire 'cerebral cortex'?

  • petrossa

    Cortical functions are more plastic since they are 'virtual' neural networks being relatively new, Subcortical functions are just hardwired so logically you can't replace them.

    Maybe in another million years most of the higher order functions become hardwired as well. (assuming we last even another millenium)

  • Bradley Voytek

    This is something I've very interested in. I think of it in graph theoretic terms: subcortical regions are more densely connected hubs or nodes, so damage to those regions is more severe.

    Primary cortices are the first major input/outputs pathways for the neocortex, and thus there are fewer recovery/rerouting options.

    I've found that subcortical basal ganglia lesions lead to worse behavioral deficits than larger cortical lesions. I've also found that PFC lesion damage is compensated for rapidly and dynamically by the intact, unlesioned PFC:

    Basal ganglia paper:

    Related blog post:

    PFC compensation paper:

    Related blog post:

  • petrossa

    To me it's more like the oldest parts of the brain are completely hardwired. There can't be much takeover since the 'repairsystem' for retraining a new neural network just doesn't exist.

    In the newer parts things are less tied down, with a lot of idle capacity being used to generate what we tend call 'us'.

    The more of this capacity is used for other applications, the lesser the level of awareness and vice versa.

    The very nature of this 'reprogrammable' network let's it reroute damaged parts to a degree.

  • Neuroskeptic

    spike: 1. Well, there's long-term potentiation which is a measure of how the strength of a synapse changes in response to “learning”. That has been extensively studied. But LTP doesn't form new connections, it only strengthens existing ones. More recently people have discovered that entirely new synapses and new cells are also formed, but this is much harder to measure at the moment.

    2. Probably not. As Bradley says the “primary” sensory and motor cortex is probably the least plastic, because these areas are most directly linked to sensory input and movement output commands. Although even there, there is evidence for plasticity e.g. in terms of “phantom limbs” and stuff. More “abstract” processing areas are likely more plastic.

  • Anonymous

    I gag everytime I see an intro neuropsych text talk about plasticity and conclude with the unwarranted optimism — “If you are going to suffer severe brain damage, do it while you are young (i.e. prior to lateralization).” What hogwash — when you read the actual case studies of these patients who demonstrated plasticity, the “success” is extremely limited, e.g. going from blindness to vaguely being able to discern shadows.

  • petrossa

    How one would classify missing limb syndrome as plasticity is beyond me.

    It's to me quite the opposite. The circuits are still there and functional. Instead of being pliable and realize the limb has gone and dedicate themselves to something else the module just stays active as if nothing happened.

    The only thing that's 'plastic' is signal crossover of the input. Which isn't actually plasticity but just open ended input capturing signals from other parts of the body.

    If you setup a simple transistor switch and touch the base with our finger it'll start conducting. The minute signal from your body's static field is enough.

    Same thing.

  • Neuroskeptic

    petrossa: My understanding was that phantom limbs are caused by plasticity to some extent, the “dead” cortex gets rewired such that it responds to stimuli which activate the nearby intact cortex and this causes phantom sensations…but I may be wrong as it's been a while since I read up on it.

    And yes, even if it is an example of some kind of plasticity, it's also an example of the limits of plasticity because a perfectly plastic brain would just adapt to the damage and, correctly, realize that the amputated limb was gone.

  • petrossa

    From what i understand (admittedly not that much) plasticity of the phantom limb kind is limited to being adjacent to another area which is plastic in itself. The body image on the brain is very plastic for some parts, and less so for others.

    When you lose a finger the other fingers next to it will indeed take over the inactive area increasing the sensitivity/versatility.

    However a lowerlimb for example doesn't really need to be plastic.
    It can't do very much else but supporting the weight.

    So if you lose the limb, the module controlling it isn't that easy to reprogram.

    Plasticity comes at a cost. If your CCU can't count upon activating the same area and getting the same results it'd be all over the place.

    If you lose an eye and suddenly your pinky toe starts putting signals in the visual cortex were the missing eye linked to you have a problem.

    But higher order functions such a rational thought i can imagine to be very plastic indeed. Them being mostly virtual it stands to reason.



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar