Humans are born to a longer period of total dependence than any other animal we know of, and we also know that mistreatment or neglect during this time often leads to social, emotional, cognitive and mental health problems in later life. It’s not hard to imagine how a lack of proper stimulation in our earliest years – everything from rich sensory experiences and language exposure to love and care – might adversely affect our development, but scientists have only recently started to pull back the curtain on the genetic, molecular and cellular mechanisms that might explain how these effects arise in the brain.

Chapter 1: In which we encounter a menagerie

You’ll often hear it said that human beings are “social animals”. What biologists tend to mean by that phrase is behaviour like long-lasting relationships or some kind society, whether that’s the social hierarchy of gorillas or the extreme organisation of bees and ants. But, to an extent, most animals are social. A mother usually bonds with its offspring in any species of bird or mammal you care to mention, and almost all animals indulge in some kind of social behaviour when they mate.

But there is another sense in which most animals seem to be fundamentally social. There is an emerging scientific understanding of the way social experience moulds the biochemistry of the brain and it looks like most species don’t just prefer the company of others – they need it to develop properly. Take that staple of genetics research, drosophila - aka the fruit fly. While they are not as social as primates or bees, they are more social than you might think, and there have been studies showing that social isolation can disrupt their mating behaviour or even reduce their lifespan.

Related effects have been seen in a range of animals, from rats through pigs, right up to our close relative, the rhesus monkey. Harry Harlow’s controversial experiments on rhesus monkeys in the 60s showed that prolonged isolation in the first year of life caused severe psychological disturbance, and a more recent study found that rhesus monkeys raised alone performed worse on memory and learning tasks and had smaller corpus callosums than monkeys raised in a social environment. The corpus callosum is a bundle of neural fibres that connects the two hemispheres of the brain and those of isolated monkeys were smaller due to having significantly less “cross-hemispheric projections”.

Chapter 2: A matter of two types of matter

The corpus callosum is also the largest white matter structure in the brain. The brain’s grey matter is neurons, the brain cells that do the heavy lifting of neural processing. The main function of white matter, composed of glial cells and myelinated axons, is to pass messages between different grey matter areas. Axons are the output cables of neurons, carrying electrical impulses to synapses on other neurons, where they influence whether they fire impulses along their axons, and so on, in the unimaginably complex web of interconnections we call a brain.

Glial cells play a largely supportive role, protecting and nourishing neurons, but a type of glial cell known as an oligodendrocyte is also responsible for producing myelin. This fatty white substance insulates axons with a myelin sheath, speeding up the transmission of electrical impulses by a factor of up to at least 50, compared to unmyelinated axons. The speed of transmission is likely to be crucial since although neurons only have one axon, they receive input at multiple synapses, probably from axons of very different lengths. So if the chance of a neuron firing depends on a number of impulses arriving simultaneously, transmission along those different length axons will need to be precisely timed.

At birth the brain is full of grey matter, but white matter is only common in a few places. As we grow the volume of grey matter tends to increase as we head into puberty, and then starts to decrease. White matter on the other hand, continually increases, and the process of myelination continues into our late 20s. This suggests a growth spurt of neurons in childhood, followed by a period of pruning and consolidation in adolescence continuing into adulthood. A lot of this is probably biologically set in stone, but the fine details are no doubt shaped by our life experiences. White matter development has to follow grey to an extent of course, since you can’t insulate an axon that isn’t there, and myelination presumably continues into adulthood as axons continue to sprout new branches and prune others in response to experience.

Chapter 3: The orphans

White matter is also a suspect in the effects of neglect on children. A study of seven children adopted into US families from Romanian institutions used diffusion tensor imaging (a type of MRI) to examine white matter bundles connecting various brain regions. Compared to a group of normally raised children, the adopted kids had less of this white matter, particularly in something called the uncinate fasciculus, which connects parts of the brain’s temporal lobe, including the amygdala, with parts of the prefrontal cortex. The paper makes no specific claims beyond “a structural change” in this white matter “tract”, but it’s worth noting that the amygdala plays a role in memory and emotional reactions and the prefrontal cortex is important for decision making and social behaviour.

The tragedy of Romanian orphans provided psychologists with a “natural experiment” to study the effects of early social deprivation and the Bucharest Early Intervention Project was the first randomised controlled trial to study the benefits of transfer to foster care for very young institutionalised children. A total of 136 children were assessed with a comprehensive battery of tests before being randomly assigned to either remain in an institution or be placed in foster care, at an average age of just under two.1 The project has generated many publications showing benefits in everything from cognitive development to psychiatric outcomes for the children placed in foster care, with the most benefit seen for the youngest children – especially if they were less than two when they were removed from an institution.

One study published just last month found differences in the volume of both grey and white matter between children who remained in an institution, those who were placed in foster care and those who had never been institutionalised. Children who had been in an institution had less grey matter, regardless of whether or not they were later placed in foster care. The foster care group however, did not have less white matter than the group who had never been institutionalised, whereas those who remained in an institution did have less. This suggests that placement in foster care at an early enough age might allow white matter development to “catch up” in children moved into better environments. So, since the earlier studies demonstrated a range of beneficial effects of foster care, it doesn’t seem like too much of a leap to suppose that those effects are due to changes in white matter. That’s a long way from scientific proof, but it’s a reasonable working hypothesis.

It’s also worth mentioning that although the orphans adopted by US families mentioned above had differences in white matter which hadn’t been reversed by foster care, the structure involved, the uncinate fasciculus, is the last major white matter tract to mature and the only one that is still developing beyond the age of 30.

Chapter 4: The experiment

But none of this tells us anything about how social deprivation influences white matter, which brings us to the protagonist of this story. Observations such as these led a team at Boston Children's Hospital, led by neurobiologist Gabriel Corfas, to investigate the effects of social isolation on mice bioengineered to develop fluorescent oligodendrocytes – the cells that produce myelin.

Beginning at an age of three weeks, just after weaning, the mice were split into groups: One group was isolated and another was housed normally, with four mice per cage. After four weeks they were tested and the isolated mice performed significantly worse on measures of social interaction and working memory – two behaviours believed to depend on the medial prefrontal cortex (mPFC). Two weeks later, the researchers inspected the oligodendrocytes in the mPFCs of the mice. Sure enough, although the number of cells was the same in both groups, the isolated mice had oligodendrocytes that were stunted, with simpler shapes, fewer branches, and so on, and two genes that produce proteins important for myelination were turned on, or expressed, less often in the isolated mice’s mPFCs. Electron microscopy also revealed significantly thinner myelin sheaths.

Examining normally housed mice showed that the first two weeks of this six week period was when a lot of oligodendrocyte development went on in the mPFC, and analysing mice after two weeks of isolation revealed similar defects to mice isolated for the full term.

So the team ran another experiment. They isolated some three week-old mice for just the first two weeks before returning them to normal housing and took another group from normal housing at five weeks old and isolated them for only the last four weeks. The mice isolated for the last four weeks were indistinguishable from mice housed normally throughout, whereas the mice isolated for the first two weeks showed the same retarded myelination and reduced performance as mice isolated for the full six weeks – even after being returned to normal housing for the last four weeks.

In other words, these first two weeks (the fourth and fifth of a mouse’s life) appears to be what is known as a “critical period” for oligodendrocyte development and myelination in the mPFC and lack of social interaction during this time retards this development, which, in turn, causes problems with memory and social behaviour.

Chapter 5: In which we follow a pathway

But how does this happen? Well, to dig any deeper we need to know about cell signalling pathways. Certain genes code for proteins, which means they produce that protein if expressed. Proteins have a huge number of different functions, but one of them is to transmit a signal from one cell to another. The protein binds to a receptor on the destination cell, and – well, something happens. Again, this can take many forms, but in the context of brain development one likely result is cell differentiation, where a simpler cell changes into a more complex, specialised one, but there are a variety of other things that can happen.

The NRG1 gene codes for a protein essential for brain development called neuregulin-1, which happens to bind to a class of receptors on oligodendrocytes called ErbB. Neuroscientists have known for some time that this signalling pathway is important for oligodendrocyte development, but Klaus-Armin Nave of the Max Planck Institute for Experimental Medicine in Gottingen published a paper in 2006 showing that it also controls the amount of myelin produced by Schwann cells – the peripheral nervous system’s version of oligodendrocytes. Shortly after, another study from Boston Children’s Hospital showed that if specific ErbB receptors in mice are blocked, oligodendrocytes are stunted, myelin is thinner and impulses travel more slowly down axons.

To investigate the influence of this cellular control mechanism on myelination in the mPFC, the researchers genetically engineered mice in which they could eliminate certain ErbB receptors with a drug. They found that if they did this before the two week critical period, mice housed normally had the same reduction in myelination and lower performance as ordinary mice that had been isolated. If they did it after the critical period it had no such effect. So it seems the neuregulin-to-ErbB binding mechanism needs to be working for mice to benefit from normal social interaction during the critical period, but after that it doesn’t matter. Finally, the team compared levels of all the components of this mechanism in mice isolated during this period with levels in regularly housed mice. The isolated mice had less of a specific kind of neuregulin-1, known as type III.

To check that all these effects of isolation weren't just part of a decline across the board, the researchers also looked at general physical activity and changes in gene expression in the motor cortex, which controls movement and sits right next to the mPFC in the brain. They found no changes in either of these measures, showing that isolation only affects specific things.

So let’s recap. Mice deprived of company during a sensitive time in their young lives express less of a certain protein in their prefrontal cortex. This interferes with a signalling mechanism involved in oligodendrocyte development and myelination. Reduced myelination alters the speed at which neural impulses travel down axons, which messes with the delicate timing of neural processing. This leads to problems with working memory and social behaviour, which aren’t reversed by reintroducing the mice back into society. And there you have it.

Chapter 6: The lonely brain

Well, actually no. This is the brain we’re talking about and nothing about the brain is ever simple. For a start there are other interpretations of how myelin deficiencies could cause behavioural changes. Corfas himself has shown that blocking ErbB receptors causes changes in dopamine signalling which could offer an alternative explanation for the effects of isolation on behaviour.

Secondly, this isn’t the only mechanism that’s been proposed to account for the effects of experience on myelination. R Douglas Fields and colleagues published a paper last year showing that a neurotransmitter called glutamate, released along axons in response to electrical activity, increases myelination by stimulating production of myelin proteins in oligodendrocytes. This was only done in a culture dish, not in a live animal, but it’s an appealing theory of how experience might shape brain development because electrical activity in axons is neural activity and neural activity is experience, so it’s an intuitive mechanism for the effects of experience on white matter development. This is one of the thorny questions in this area: does experience shape white matter, which then shapes neural activity, or does neural activity shape white matter? The answer is probably both.

Also, nobody is suggesting the effects of social isolation are confined to white matter and myelination. The brain is a staggeringly complex thing and a scenario as broad as social experience, or the lack of it, is likely to have a wide range of consequences. Remember the orphans in chapter 3 who had less white matter than children raised in families? They also had less grey matter. And while this does go through a period of decline as we develop, it’s unlikely their reduced grey matter was an indication of accelerated development as they were only around 9 at the time and so not yet at the age when this usually happens. It also wouldn’t really tally with the mental health problems they invariably suffered from.

As a specific example, there was a paper published this July looking at the effects of social isolation on behavioural performance in rats. It focussed on the barrel cortex, the part of a rat’s brain that receives input from their whiskers, as whiskers are an important social communication channel for rats. The researchers found that early isolation disrupted a signalling pathway involved in forming healthy neural circuitry in the barrel cortex, leading to decreased whisker sensitivity and deficiencies in “whisker-related” behaviour. This had nothing to do with white matter and everything to do with grey matter.


Schizophrenia and mood disorders usually crop up in adolescence and have been linked to disturbances in white matter and myelination, and disruptions of the neuregulin-1-to-ErbB signalling pathway. Corfas has also shown that disruption of oligodendrocyte genes involving our old friend neuregulin, causes schizophrenic-like behaviour in mice. All of which suggests that this latest study may shed more light on new ways to understand and possibly even treat these conditions.

Corfas’ team is also currently looking at ways to target neuregulin-1 and ErbB signalling pathways with drugs that might stimulate myelination. It’s conceivable that this could eventually lead to ways to treat the debilitating effects of early social deprivation, but we will have to proceed very cautiously down this particular path as too much myelination is likely to be as bad as too little.


1 (Before you start worrying about the morality of randomly assigning orphans to foster care, be assured this study went to great lengths to ensure it was a force for good. The study recruited its own foster parents, as “government-sponsored foster care was limited to about one family” and none of the children received less care than would have happened if the study hadn’t been conducted. Also, crucially, following the ethical guideline that “no subject should be randomised to an intervention known to be inferior to the standard of care”, the researchers point out: “at the start of our study there was uncertainty about the relative merits of institutional and foster care in the Romanian child welfare community, with a historical bias in favour of institutional care.” Government officials and child protection professionals were kept informed of the results as the study progressed, with the result that the Romanian government eventually passed laws prohibiting placing infants younger than two in institutions. If you’re still not convinced, read this.)

Images: Neuron with oligodendrocyte and myelin sheath by Andrew c; Oligodendrocyte by Jurjen Broeke.