Almost everyone has had a night when sleep would not come. When the room is dark, the clock is loud, and the brain simply refuses to switch off. What science has learned in the last decade is that those hours of sleep, or the lack of them, are doing far more to your brain than you might think. At a Baycrest Neuro Network Speaker Series evening devoted to sleep, two scientists made the case plainly. Dr. Allison Sekuler, a cognitive neuroscientist and chief scientist at the Baycrest Academy, framed why sleep matters for the aging brain. Dr. Brian Levine, a neuropsychologist and senior scientist at the Rotman Research Institute, presented new research showing how sleep quietly rebuilds memory while you rest.

Why Sleep Belongs in the Dementia Conversation

For the first time in history there is real reason for optimism about dementia. Research over the past five years suggests that lifestyle changes can prevent or significantly delay close to half of dementia cases. The widely cited Lancet Commission list now identifies 14 modifiable risk factors, things people can actually change in everyday life, from blood pressure to physical activity.

Sleep is not yet on that list. But researchers fully expect it to be added when the next report appears, and the reasons are already clear. Poor sleep is bound up with many of the risk factors that are on the list: it is linked to obesity, diabetes, physical inactivity, and high blood pressure. Sleep sits underneath much of the rest of the picture.

The Brain's Overnight Cleaning System

One of the most important reasons sleep matters is that the brain does not simply shut down at night. It goes to work. During deep sleep the brain runs what amounts to a dishwasher cycle. A network known as the glymphatic system, the brain's equivalent of the lymphatic system, becomes active and rinses out the waste that accumulates during waking hours.

That waste includes misfolded proteins, the kind that can build up in the brain and are associated with Alzheimer's disease. When you are not getting enough deep sleep, this clearance process cannot operate as effectively or efficiently. That does not mean a lack of sleep causes Alzheimer's. It means that a lack of deep sleep can increase the risk.

The consequences of poor sleep reach further still. Too little sleep is associated with more inflammation in the brain, with changes in blood flow that can stiffen and restrict arteries, and with altered functioning of the brain's neural networks. And, as Dr. Levine's research shows, it directly affects memory.

A Memory Scientist Turns to Sleep

Dr. Levine's path to studying sleep began at the very start of his career, working as a sleep technician at what was then one of the largest sleep laboratories in the world, in Detroit. There he learned the full range of assessments used to diagnose sleep disorders. He then trained as a clinical psychologist and a neuropsychologist, specializing in the assessment of brain and behaviour, and joined the Rotman Research Institute in 1993.

Sleep and memory then sat in separate worlds for years. But over the past 10 to 15 years, memory scientists have become deeply interested in sleep. It had been known since the late 1980s and 1990s that sleep helps consolidate motor sequences, the kind of learning involved in playing a piano. The newer question was how sleep shapes memory for everyday, real-world experience.

The Baycrest Tour: A Memory Test You Can Walk Through

To answer that, Dr. Levine's lab built a more naturalistic memory test than the usual lists of words or pictures on a computer screen. They called it the Baycrest Tour. Participants put on headphones, much like a museum audio guide, walked through the building, and learned about the works of art on display. The point was movement and a real experience that people might find at least mildly interesting. The study, led by graduate students Nick Diamond and Stephanie Simpson, was published last year in Nature Human Behavior.

Afterward, the team tested two kinds of memory. The first was feature memory, the details of an object, tested with simple true or false statements such as whether a particular sculpture was made of wood. The second was sequence memory, the order of events, such as whether one artwork was encountered before another. Four separate sets of questions were created so participants could be tested at four time points without ever repeating a question: right away, the next day, one week later, and one month later.

Feature memory behaved exactly as decades of research would predict. People forgot the details in a steady, exponential decline, one of the most reliable findings in all of memory science. Sequence memory did the opposite. After a single night of sleep, it actually improved, and it stayed high a month later. It is very rare to give a memory test, with no practice, and watch performance go up over time.

One Night of Sleep, Measured a Year Later

To prove that sleep was responsible, the team ran a clean comparison using a 12-hour delay. Half the participants came in at 9 a.m., took the tour, and were tested at 9 p.m. the same day, staying awake in between. The other half came in at 9 p.m., took the tour, slept, and were tested at 9 a.m. the next morning. Everything else was identical. The only difference was whether the first 12 hours included a night of sleep.

The sleep group showed the sequence-memory boost. The wake group did not. And when the researchers tracked people down again much later, the benefit had held: one night of sleep significantly improved memory for sequences, and that improvement was still measurable a full year afterward, while those who had stayed awake showed no such benefit.

Inside the Sleeping Brain: Spindles and Slow Waves

Why does sleep do this? The answer lies in the structure of the night. Sleep cycles between deep sleep, also called slow-wave sleep, and REM sleep. In deep sleep the whole brain is synchronized and largely tuned out from the environment. REM sleep looks almost like wakefulness on a brain recording, which is why it was once called paradoxical sleep, with a great deal of activity going on. Deep sleep tends to dominate the early part of the night and REM the later part.

The memory action happens in the interplay between brain regions. The hippocampus, the brain's memory hub, sits deep inside, while the cortex wraps the outside. The thalamus, in the middle, acts as a relay. Two signals can be measured during sleep: slow waves, and short fast bursts called sleep spindles. Crucially, it is the coupling of slow waves and spindles, spindles riding on top of slow waves, that supports memory. That coupling is the deep memory centres of the brain communicating with the cortex to solidify what was learned.

Working with colleagues who developed sophisticated measures, the lab found that among 90 participants, the more coupled spindles a person had, the better they did on sequence memory. Uncoupled spindles, the ones not riding a slow wave, showed no such relationship. The size of the effect was modest, around 5 percent, but it was statistically reliable, and given how many competing thoughts and events could interfere across a day, seeing the effect at all is striking. In real life, the practical impact is likely larger.

Sleep Engineering: Hacking Memory While You Rest

If specific brain rhythms strengthen memory, could you boost them on purpose? That is the idea behind a technique called targeted memory reactivation, or TMR, pioneered in recent years by Ken Paller at Northwestern University. During learning, a piece of information is paired with a sound, perhaps an environmental noise like a passing train. Later, while the person is in slow-wave sleep, that sound is replayed through a closed-loop EEG system, timed precisely to the slow-wave and spindle couplings. When the timing is right, memory for whatever was paired with the sound is enhanced.

The approach extends into stranger territory. Researchers have used dream engineering with lucid dreamers, people who become aware they are dreaming and can intentionally act within the dream. By training them on a problem while awake and cueing them during sleep, they have improved problem-solving performance carried out inside the dream. Baycrest is not pursuing dream engineering yet, but it is preparing to repeat the Baycrest Tour with paired sounds to see whether TMR can strengthen memory for a naturalistic experience.

From the Lab to Trauma Recovery

The final thread of the research connects sleep, memory, and psychological trauma. Dr. Levine's lab studies post-traumatic stress disorder, including work with paramedics. PTSD is a disabling condition that involves both memory and sleep, and the long-term goal is to translate the basic science of sleep and memory into treatment.

The link runs through sequence memory itself. Part of what sequence memory does is tell you where things sit in time and space. When someone with PTSD has an intrusive memory or flashback, it can feel as if the traumatic event is happening right now. Recovery involves learning that the event happened then, somewhere else, and is not happening in the present. Achieving that psychological distance is, in essence, a sequence-memory problem.

What the Science Means for Your Sleep

The practical message of the evening was reassuring: don't panic, because most sleep problems can be improved. The first step, if you suspect something like restless legs or sleep apnea, is to see a doctor, since many conditions are treatable in relatively simple ways. Treating sleep apnea matters because it is itself a risk factor for stroke and other conditions, and the standard first-line treatment, a CPAP machine, is extremely effective.

For the more common problem of a mind that will not stop running, the leading evidence-based approach is cognitive behavioural therapy for insomnia, known as CBT-I.

A useful free resource, developed at Dalhousie University, is the MySleepWell program, which offers strong CBT-I tools and shows that you do not necessarily need one-on-one therapy to benefit.

Pills, Wine, and Wearables

Two of the most common shortcuts people reach for can backfire. Taking a sleeping pill for one or two nights to get over a rough patch is generally fine, but habitual use is one of the worst things you can do for sleep. Over the long run it can be extremely disruptive and can actually cause insomnia, and unwinding a dependence is very difficult. A nightly glass of wine works the same way: it may feel like it helps, but it fragments sleep and trains the body to struggle without it.

Wearable trackers such as smart rings and watches are a mixed bag. They are fairly accurate at estimating how much you slept, which is genuinely useful because people are notoriously poor at judging their own sleep. They are far less reliable at sleep staging, the breakdown of REM and deep sleep, because they infer it from movement, heart rate, and temperature rather than measuring brain signals directly, and their algorithms are proprietary. For some people a tracker provides a helpful reality check. For others it fuels anxiety, and the better move is to take it off.

Sleep Across the Lifespan

How much sleep you need is individual, with no hard cutoff. The clearest warning sign is consistency: regularly getting less than about four hours is harmful, and too much sleep can also be unhelpful. Healthy adults typically spend roughly 10 to 15 percent of the night in deep sleep. Because deep sleep clusters early in the night and REM later, waking too early tends to cost you REM, which matters too; the cycling between stages may itself be important for memory.

Sleep also changes across life. Infants and school-aged children get far more deep sleep than adults. With age, the need for sleep declines somewhat and the proportion of deep sleep falls, a normal part of aging. Hormonal transitions matter as well: many people develop sleep problems during perimenopause and menopause, driven by shifting rhythms and changes in body-temperature regulation, and guidance on hormone therapy has changed considerably over the past decade, so it is worth discussing with a physician.

Perhaps the most important reframing of the evening was about timing across the years. Alzheimer's disease does not begin at 75. Sleep habits at 35, 45, and 55 all influence brain health later in life. The takeaway is not to wait until old age to start caring about sleep, but to build good habits early, alongside attention to stress, nutrition, and exercise. A good night's sleep, it turns out, is one of the most accessible tools available for protecting the brain over a lifetime.