Why researchers are using rats to work out whether there's a link between VR and dementia
"We made it very immersive."
Last November I wrote a piece investigating a few examples of video games and related technology that seek to explore, or even treat, neurological conditions. This included virtual reality applications, which prompted a comment from a Eurogamer reader called Pilotmonkey who said that, "I stopped using my PSVR because of reports that it triggered responses akin to dementia in the brain". Pilotmonkey went on to refer to a study that had been conducted in this area. Curious, I spent the next few months looking into this possibility in detail.
I have something of a vested interest. I own a PSVR myself, which I use fairly regularly. I've been playing team shooter Firewall: Zero Hour on and off since its August release, for example, and I still go back to the incomparable experience of Tetris Effect in virtual reality. I've never noticed any ill effects after playing any VR game; I don't even suffer from motion sickness.
Looking into this subject a little, I quickly found myself in complex territory. To begin, we need to take a quick look at the concept of spatial mapping. For many games, the brain needs to operate in virtual reality just as it does in the real world. Take my beloved Firewall, for example. I need to recognise each map as one continuous space, and navigate it as such: to see somewhere in the virtual distance and move there, or move back to a spot I'd previously occupied. Moving in and out of cover, aiming and shooting at enemies, attacking and defending objectives, my brain needs to act as though I'm really there so I can see a space and understand how it works.
The part of the brain responsible for our sense of space is the hippocampus, which is also crucial for memory function. The hippocampus allows us to recognise and therefore navigate space through spatial mapping. To better understand this, I got in contact with Mayank Mehta, a professor in the departments of physics, astronomy, and neurology at UCLA.
When I get hold of Professor Mehta on Skype, I can immediately notice the intense, slightly nervous energy that he gives off. He's a communicator, one of those people who's capable of immediately answering questions with detailed and relatable replies that sound for all the world as though he's carefully prepared them beforehand.
"It's not fully understood, but the basic idea is that when you are in one place in a room, you experience a certain light and a certain sound," says Mehta when I ask about spatial mapping. "And then you go to the next place in the room, and you experience another set of light and sound. And then in your brain there is something called neuroplasticity. The neuroplasticity changes the way the brain is wired when you experience the sound and light in this place, and the way the neuron circuits get modified is that neurons that fire together, wire together.
"Neurons that fire together, wire together."
"The two neural circuits [relevant to experiencing the two different places in the room] get combined together," he continues, "And now, you would call the whole area one space, rather than one set of views here, and another set of views there."
In other words, neuroplasticity (the brain's ability to change how neurons, the cells responsible for brain function, are connected to one another) puts all the pieces together. Imagine that you're putting together an enormously complicated jigsaw puzzle. Perhaps you've managed to form an image of a cloud, and an image of a car. These are two separate and seemingly unrelated images surrounded by hundreds of individual puzzle pieces. Neuroplasticity puts the whole puzzle together for you, so that instead of a cloud to your left and a car to your right, you can now see one complete scene; a sunny sky full of clouds above a busy street full of traffic. You can look from car to car, from cloud to cloud, to car, and to cloud again, and understand that it's one continuous image. This is how the brain works in real life. But it's not, surprisingly, how it appears to work in virtual reality.
To stretch the puzzle analogy a little further, in VR, the brain's neuroplasticity seems to connect only a small number of puzzle pieces, and seemingly chooses them at random to create a theoretically indecipherable image - yet the VR user is still able to function as if the puzzle were complete. To put this into context, we must discuss Mehta's work at UCLA.
Mehta led a team of researchers that studied the brain activity of rats navigating a virtual reality environment. (It seems this may be the study referenced by Pilotmonkey). "We created a virtual reality for rats," Mehta tells me. "They are seated comfortably, they are harnessed, they are totally comfortable. They take naps. As soon as they start running, a little ball underneath their feet starts to move, and they are held gently by the harness, so they don't go anywhere. The movement of the ball creates movement in the virtual reality engine, in the visual scene all around the rat.
"We made it very immersive," he continues. "We put the screen not just in front of him like a TV, but as a screen that goes all around him, and the image comes all the way up to his feet, so he's completely immersed, [better than] typical VR that's available these days. You put on goggles, but you don't see your legs, you don't see your hands. But [the rats] can totally see themselves, they can see their own shadows, so it's a fully immersive and non-invasive virtual reality. So kind of the Rolls Royce of virtual reality compared to what we have for humans."
"So kind of the Rolls Royce of virtual reality compared to what we have for humans."
Mehta's team ensured that the area was clean and the ball that triggered movement was smooth, so that no smells or textures would interfere with the rats' perception of the virtual reality maze. The original intention of the study was to better understand how spatial maps are created in the hippocampus.
The rats took to virtual reality extremely well, reacting as if it were a real-world maze, and performing tasks that they were taught. Everything was going wonderfully. "Then we measured the brain signals, because we thought, 'Well this is a perfectly clean maze, we should get beautiful space mapping'," Mehta tells me. "And instead of that, what we found was very surprising."
Mehta and his team discovered something completely unexpected; something that triggered his eagerness to see more research done in the area. "We found that 60 per cent of neurons [in the hippocampus] simply shut down in virtual reality, which was a big difference. We didn't expect such a large proportion of neurons to shut down. Then we said, 'Alright, let's look at the remaining 40 per cent of neurons. What do they do?' We found that the activity pattern of the remaining neurons was also substantially altered. As if randomised, as if those neurons had no idea where the rat is in space. So that was very surprising, because he seems to clearly know where he is, he's able to navigate in the virtual maze.
"So we are still following up on those experiments," Mehta tells me. "We are doing new experiments to test how good their memory is, of virtual reality. Because maybe they can see the virtual world, but maybe they're not able to make a mental map of the world, a memory."
This follow-up study remains unpublished at time of writing, so Mehta was unable to discuss it, but his hope is that within a few months, it will have been published, and he will have a few more answers to take him and his team a little closer to an understanding of the situation.
"We are still following up on those experiments."
For now, two questions present themselves for you and I. Firstly, can findings from a rat's brain be considered relevant to a human's brain? A rat's brain is, by Mehta's own admission, much easier to study. And even if the findings can be assumed to apply to humans, why should we care?
The fact that something as fundamentally important to life as the perception of 3D space is involved almost answers the first question by itself. "We believe the way all animals perceive space is identical, and it must be so," says Mehta. "Because if you think that wine is delicious, but your dog doesn't, that's okay. But if your dog didn't agree with where you are, you are going to collide. Lions and zebras are going to collide; perhaps all zebras are going to be eaten up, because lions will catch them too soon, and then lions will die because they'll get too fat. Life on the planet, all animate life on the planet, would come to an end if all the species - doesn't matter how different they are, birds, crocodiles, zebras - didn't agree on exactly what space and time are one hundred percent. We believe this is a fundamental property of all animal cognition, and it is commonly shared. That's why studying the rat's behaviour of perceiving space gets us to understanding how we create abstract ideas."
As for what Mehta's findings could potentially mean, there are precedents for the brain activity he's observed in rats using virtual reality which, considered in isolation, are alarming. "[The rat's] vision is telling him he's moving forward, his claws are saying he's moving forward, but his sense of acceleration is telling him that he's not going anywhere," says Mehta. "And it's that mismatch between different things that causes these neurons to fire abnormally. And we believe this kind of mismatch may be happening under various diseases as well... Like watching an old television or an old movie that's not working perfectly, in the speech, and the sounds, and the light, the scene is slightly off. You can tell that something is wrong. That's exactly what we think is going on in virtual reality."
It's an established fact that neuroplasticity, the brain's ability to rewire its circuits, remains in place throughout our lives. The fear is that there's a possibility that VR use, which appears to trigger abnormal brain function in rats at least, could 'teach' the brain to rewire itself in an undesirable way. This proves nothing in and of itself, however, and Mehta is keen to stress that much more work is needed to study the issue. His is one single study in a complex subject, and he has no interest in scaremongering.
"The long term consequences are really hard to measure in the human brain," he explains. "Because humans age very slowly. They live for around a hundred years, rats live for roughly two or three years. We can't wait for forty years, for teenagers who are today using virtual reality to see what happens to them when they're sixty. Does it cause Alzheimer's? Does it cause some other [disorder]? Or is it therapeutic? Maybe it's good! I don't know! But we need to measure that in rats, and those experiments can be done in a couple of years. So it's still not too late. And that will be very informative, what happens to rats. For humans in the long term. But that still needs to be done."
"The long term consequences are really hard to measure in the human brain, because humans age very slowly."
It's easy to imagine a tabloid selectively quoting Mehta and screaming, "VR causes Alzheimer's!" across its front page. Equally, it's unhelpful to refuse to examine the possibility, if only to disprove it. At the moment, we are at the beginning of a long process of investigation into this entire area; Mehta himself will be the first to tell you that. But before we go any further, let's take a moment to define what dementia - an umbrella term that diseases like Alzheimer's falls under - actually is.
"Dementia is a term given to a number of different symptoms that people might experience," Ed Pinches, Alzheimer Research UK's Science Media Officer, tells me. "The most common thing that people think of is memory loss, which is one symptom of dementia. But there are others. Things like problems with your spatial navigation for example, personality changes, behaviour changes. Dementia is caused by a number of different diseases. These are diseases of the brain, the most common of which is Alzheimer's Disease".
Without in-depth knowledge of Mehta's study, Pinches was unable to comment on it. He spoke with me at length about the roots of dementia, however, including an important biological factor. "There are proteins which build up inside and outside the brain's nerve cells," he tells me. "These nerve cells are really important for carrying messages, they're helping us so we can remember things. And the way that they send messages to each other is the way our memories are encoded. Something about these [proteins] seems to be toxic to the cells. Amyloid [outside the nerve cells] normally comes first, and tau [inside the nerve cells] comes later. Then you'd have nerve cells which die, and therefore they're unable to send messages to each other. Your brain physically shrinks, so you have less nerve cells than before, therefore less capacity to do things".
These well-understood biological factors don't appear to support the potential link between neuroplasticity and dementia. When we look closer at the progression of Alzheimer's disease, though, there is one tenuous similarity with Mehta's hypothesis; the importance of a certain area of the brain.
"The area of the brain that's affected first is the bit that deals with your short-term memory recall, which is the hippocampus," says Pinches. "Which is why one of the first symptoms of Alzheimer's disease is becoming forgetful. There are also other things that occur during the disease, so it's not just these two proteins. There's a lot of research now thinking that the immune system of the brain has a really big part to play in the progression of the disease. There are brain cells that help with the immune response, and also there are support cells in the brain that help these nerve cells do their job properly, cells called microglia and astrocytes".
However, the hippocampus is not always the part of the brain affected. "Frontotemporal dementia, or FTD, is caused by diseases that occur in the frontal lobe of the brain. We also have variants of Alzheimer's disease that can affect the sides or the back of the brain as well. So the hippocampus, yes, is very important, but it depends which disease you have". Continuing to elaborate on the complexity of Alzheimer's disease, Pinches tells me: "There are a number of different causes, it's a complex mix of things. Age is one of the biggest risk factors. As you get older, you're more likely to get the disease. That doesn't mean that dementia is an inevitable part of ageing. There are genetic causes as well. In very rare cases of Alzheimer's disease, you can have a faulty version of a gene which causes you to have the disease. But there are also other genes which are called risk genes, which by having a copy of the gene, doesn't mean you'll have the disease, but it does increase your likelihood of getting it. Then we also know that environment has a massive part to play, around 30 per cent of our risk of developing Alzheimer's disease is due to modifiable risk factors. These are things like levels of education, wealth."
Education is needed. "77 per cent of people know that they can reduce their risk of cardiovascular disease, but only 34 per cent of people know that they can reduce their risk of dementia," Pinches tells me. "Which is quite a shocking statistic, really. What we say is: having a healthy heart is also having a healthy brain. All those good health messages that we talk about, like good blood pressure, keeping your cholesterol in check, not smoking, drinking within the recommended guidelines... all are true for dementia as well."
"77 per cent of people know that they can reduce their risk of cardiovascular disease, but only 34 per cent of people know that they can reduce their risk of dementia."
So, while the jury is currently very much out on any potential link between VR use and dementia, maintaining a healthy lifestyle has certainly been established as a good way to reduce your risk of developing any such disease. VR is actually being used to help power research into the link between spatial navigation and dementia, and if this could open up a new avenue of diagnosis.
"Spatial navigation is thought to be one of the first things that might go wrong in Alzheimer's disease, we're not entirely certain yet, but this is something that researchers are looking into," Pinches explains. "Those proteins are building up in the brain decades before any symptoms start to show, so we're wanting to be able to diagnose and detect these diseases earlier. One way of doing that may be to look at people's spatial navigation levels, seeing how well they are navigating, and seeing if that is an indicator of someone developing the disease. So Alzheimer's Research UK teamed up with Deutsche Telekom, and we created an app called Sea Hero Quest. This is an app that's also available on VR, Samsung, Oculus VR. By downloading and playing the game, you're contributing to the research yourself. So when you download the game you're the captain of a ship. You have to remember how you got to different places, and shoot flares back to your starting position [...] by downloading and playing the game, it's helping scientists in the lab gain a lot of hours of research that simply couldn't be done in a lab setting".
Sea Hero Quest has been played by roughly 2.5 million people at time of writing, and the project has already produced some interesting results. (According to the ARUK website, the app produced "over 9,400 years worth of equivalent lab-based research".) In order to avoid the results being skewed by those most familiar with video games, results from tutorial levels were compared to the main body of results. So far, people living in North America, Nordic countries, Australia, and New Zealand have been found to have the best spatial navigational abilities, with men generally performing better than women. Predictably, this difference is less pronounced in those countries with greater gender equality. Equally predictably, perhaps, it was those countries with the greatest GDP that performed best. Overall, spatial navigation performance consistently declined with age.
Amongst all the uncertainties, one thing is clear; much more research is necessary to fully understand the causes and progression of Alzheimer's and associated diseases. There is nothing to prove that VR use can cause dementia, so don't throw your headsets in the bin just yet. Nonetheless, we can not say with confidence that the possibility has yet been disproven, either.
If our industry is to continue to grow and mature, we must embrace awkward discussions and uncomfortable possibilities; not hide ourselves away in... well, in an alternate reality. Whenever any such concerns are found, we should all encourage investigation just as strongly as we condemn lazy scaremongering.