Skipping Sleep, Skipping Marks and Why Sleep Is a Study Strategy, Not a Luxury

The All-Nighter Myth

Have you ever pulled an "all-nighter"? You have a deadline looming, or a test the following day, and you decide that you have just got to stay up and get the work done. According to some informal polls, up to 85% of students (at least at university level) claim to have done just that, and cramming for an exam is the most popular reason for doing this.

It seems so logical on the surface. You have an exam tomorrow, it's only a couple of hours long, so you spend the "wasted" time you would normally be sleeping on revising instead. But the science tells a very different story.

What the Science Actually Shows

In a study conducted in 2006 (Yoo et al., 2007), sleep researcher Matthew Walker and colleagues experimentally tested our ability to learn while sleep-deprived. Students were randomly allocated to either a "sleep" group or a "sleep-deprived" group, and both groups stayed awake all day as normal. When it came time for bed, the sleep group were allowed a full night's sleep, while the deprived group had to stay awake all night. Both groups were then awake all of the following morning, meaning the deprived group had been awake for around 29 hours when they were placed in an MRI scanner and given a list of facts to learn. The sleep group learnt exactly the same list of facts inside the MRI scanner as their brains were imaged.

And then the study got really interesting. Instead of immediately asking the students to recall the facts, they were all sent home to get two full nights of sleep. Why? Walker was looking to see whether learning was impaired, rather than recall. To test this, the effects of tiredness on recall had to be removed, so the participants had to be fresh and awake when it came time to show what they had learnt.

The results were clear: the sleep-deprived group had a 40% deficit in their ability to learn the facts, compared to the group who'd had a full night's sleep. That's a startling difference. Just one night of sleep deprivation meant a 40% drop in the ability to recall information learnt while tired (those bits of important information you're trying to cram into your brain at 2 am, ready for the exam the next day). 40% isn't just the difference between an A and a B, it's the difference between a pass and a fail.

You may be wondering what the point of the MRI scanner was. It tells us not only that the students were far less capable of learning new material, but also why. The hippocampus is the area of the brain responsible for consolidating memory, for turning new information into long-term memories that you can recall later. In other words, it's responsible for learning. In the students who'd had a good night's sleep before learning, their hippocampus was perfectly normal, with lots of learning-related activity. The deprived group, however, had no significant activity at all. None.

It's Not Just About All-Nighters

Now you might think, "Well, that's all fine, but I don't stay up all night, so I'll be okay." Not true. Further studies showed that simply disrupting the depth of an individual's NREM sleep, stopping them from getting good-quality deep sleep, even without actually waking them up, produced similar learning deficits.

So something as simple as a phone by your bed that goes "ping" with a message several times a night is enough to keep you in shallow sleep (still asleep, but not the deep, slow-wave sleep needed for memory consolidation), impacting how much you're able to learn in the process. Without good-quality deep sleep, your brain simply cannot take on new information in a meaningful way. And to perform well in an exam, that is exactly what you need.

Before and After: Why Sleep Matters on Both Sides of Learning

But it isn't just about that "night before the exam" cramming session. Getting good sleep both before and after you're presented with new information is what enables you to learn it.

When we have a good night's sleep before learning, we move all the information the hippocampus took on the previous day into the long-term "secure vault" of the cortex. Because the hippocampus has limited capacity, this frees up space, ready to take on more information. And, as you can probably infer from the above, it is during sleep that this "file transfer" happens, when you download the information of the day and store it safely away.

Think of it like taking photos on your phone with no Wi-Fi. If your phone was lost or stolen, those photos would be gone too. But if you'd got home and the images had synced to the cloud, then even if the phone was lost, the photos would be safe. Sleep is when that connection to the cloud happens, when you "upload" all your memories and store them safely.

And, as mentioned, it isn't just any old sleep but specifically deep NREM sleep, and that matters because sleep isn't the same throughout the night. In the first couple of hours, we get more NREM sleep and less REM sleep, while later in the night we get more REM and progressively less NREM. So if you short-change your sleep by going to bed later, you miss out on the NREM-heavy early part of the night, going straight into the REM-heavy sleep of the early morning.

The Takeaway

All of this evidence tells us the same thing: we don't learn when we're sitting in the classroom or reading through our notes. We learn when we're asleep, in deep NREM sleep. Far from being a "waste of time" that could be "better spent cramming," sleep is essential for your brain's ability to actually learn the information it has encountered, to lock it away safely in long-term memory, ready to be recalled later.

Going to bed late, or hindering your ability to get into deep sleep through noise and distraction in the room (a phone, music, TV, even using screens before bed, which delays how quickly you fall asleep), means that all the work you put in during the day counts for very little.

Reference: Yoo, S. S., Hu, P., Gujar, N., et al. A deficit in the ability to form new human memories without sleep. Nat Neurosci 10, 385-392 (2007). https://doi.org/10.1038/nn1851