top of page

Sleep and the Brain: What a Good Night Really Does

Writer: Danny George
Danny George
6 minutes ago
8 min read

the brain sleeping

Think of your brain like a shop that never really closes. All day it takes in orders, files away memories, and sweeps up the mess from thinking, moving, and living.


At night, when the front door locks and the lights go low, the night crew comes in. They file the memories that matter, toss the ones that do not, and take out the trash.


Sleep is that night crew. And like any crew, it does its best work when it gets a full shift, on a regular schedule, in a quiet building.


Last week we talked about food. This week, sleep.


It is the third Sunday in our brain series, and it may be the one most of us can change the fastest.


You cannot eat your way to a better brain in a week.


You can go to bed at a decent time tonight.


Here is what the studies found. Including the one most people would rather skip.


The big number: 30 percent


The study that anchors this whole topic came out in 2021. Severine Sabia and her colleagues followed 7,959 British civil servants in the Whitehall II study for about 25 years. They asked people how long they slept at age 50, then again at 60 and at 70, and watched who developed dementia. In that time, 521 people did.


The people who slept 6 hours or less a night at age 50 had a 22 percent higher risk of dementia than the people who slept 7 hours. At age 60, the gap was 37 percent. And the people who slept 6 hours or less at all three ages, 50, 60, and 70, had a 30 percent higher risk than the people who slept a steady 7 hours.


Seven hours was the sweet spot in this study. Not nine. Not five. Seven.


Think of it like a bank account. A short night here and there is a small withdrawal, and your brain can cover it. Short nights every night for twenty years are a different story.


Now, I want to be gentle with this part, because it matters and I do not want to overpromise you anything. This was an observational study, which means the researchers watched how people already slept rather than handing anyone a sleep schedule and testing it.


And the truth is, people who sleep badly often have other things going on too, stress or pain, low mood or health problems, and no amount of math can fully untangle all of that.


So the number is real, you need more than 6 hours a night and it is worth paying attention to, but it is not proof that short sleep causes dementia. It is a good reason to take sleep seriously, not a reason to be afraid of it.


There is a second honest note. The link was strongest at age 50 and 60, and weaker by age 70. That fits a story where sleep in midlife matters most, and where poor sleep in very old age may partly be an early sign of the disease itself, not just a cause of it. The arrow may point both ways.


The deep sleep connection: the night crew at work


So why would sleep matter to memory at all? A smaller study gives us the clearest picture of the mechanism, and i'll share the boring details because it's worth knowing.


In 2015, Bryce Mander and his team at UC Berkeley studied 26 cognitively normal older adults, average age about 75, none with dementia. They scanned their brains for beta amyloid, the protein that builds up in Alzheimer disease, and then watched their sleep with brain wave recordings and tested their memory before and after a night of sleep.


The people with more beta amyloid in the front of the brain had weaker deep sleep, the slow wave sleep that does the heavy filing work overnight. And weaker deep sleep meant worse overnight memory. People remembered less of what they had learned the day before.


The model that fit the data best was a chain: amyloid to weaker deep sleep to worse memory. Sleep sat in the middle of the chain, not off to the side.


But read the fine print with me. This was 26 people, studied at one point in time. It cannot prove that increased amyloid caused the weak sleep, or that weak sleep caused the memory loss. It shows a pattern that fits, in a small group, at one moment. It is a clue, not a verdict.


As I've mentioned in previous blogs my son Bennett builds a sand castle one bucket at a time. He does not stand back and plan the whole castle. He fills a bucket, flips it over, fills another one, then after a few failed bucket flips and some help from his dad an hour later there is a castle. Deep sleep works the same way, in reverse. Each night of deep sleep is one bucket of filing and cleaning. Miss enough nights and the castle never quite gets built.


My daughter Cecily copies everything her brother does. Sleep is like that too. Your brain copies the pattern you give it. Same bedtime, same wake time, and the night crew learns when to show up. Random hours, and the crew never knows when the shift starts.


The honest plot twist: fixing sleep did not fix memory, at least not in a year


Now the part most newsletters would skip. I am not going to skip it, because your trust is worth more than a clean story.


In 2026, researchers published the SIESTA trial in Alzheimer Disease and Dementia. They took 200 cognitively normal older adults with insomnia symptoms and split them in half. One half got cognitive behavioral therapy for insomnia, called CBT-I, the structured talk therapy that is the first line treatment for chronic insomnia.


The other half got a control. They tested thinking and memory at the start, at 6 weeks, and at 1 year, and measured beta amyloid in a smaller group of 50.


Sleep got better in the therapy group. Insomnia severity fell, and people felt more confident about their sleep.


Memory did not change. Thinking speed did not change. Executive function did not change. Beta amyloid did not change. At one year, the group that slept better looked the same on every brain test as the group that did not.


So did sleep fail? I do not read it that way, and here is why.


First, one year is short. The Whitehall study watched people for 25 years. If short sleep raises risk over decades, one year of better sleep in healthy older adults may simply be too short to move a memory score.


Second, these were cognitively normal people with no memory problems at the start. There was not much room to improve, and a year is not long enough to see a decline that better sleep might have prevented.


Third, better sleep is still worth having on its own. It helps mood, energy, balance, and the next day, even if it does not show up on a memory test in twelve months.


Here is what I take from it, and I want to be clear this is my read, not the trial's conclusion. Sleep is probably one leg of the table, just like food was last week. It matters, and it is worth protecting, but sleep alone, for one year, in healthy people, is not a magic switch for memory.


The good news: sleep therapy really does work for sleep


If the SIESTA trial makes you feel hopeless about getting improved sleep know this. The Cognitive Behavioral Therapy (CBT) itself works, and it works without a pill.


In 2025, Lee Ritterband and colleagues at the University of Virginia published a trial of SHUTi OASIS, an online CBT-I program built specifically for older adults.


They enrolled 311 adults aged 55 to 95 from 38 states, average age about 66. One group used the online program alone for 9 weeks. A second group used it with a little human support when they fell behind. A third group got online sleep education only.


Insomnia severity fell a lot in both program groups, and the benefit held at 6 months and at 12 months. At one year, 40 percent of the program alone group and 51 percent of the program plus support group were in remission, meaning they no longer met the cutoff for clinically significant insomnia. In the education only group, it was 16 percent.


That is a real, practical result. No drug. No side effects. Nine weeks of learning a new way to sleep, with the benefit still there a year later.


I train for a half marathon, and I track my own sleep while I train because it is the most fixable lever I have. A 6 hour night and an 8 and a half hour night feel like different bodies the next morning. Your brain is no different. The night you sleep well, the next day is easier to think through, easier to move through, easier to enjoy.


So what do you actually do tonight?


No pills. No gadgets to buy. No perfect bedroom to build first. Just a routine your brain can count on, the same way it counts on food at regular meals.


Start with these five, and start with the first two if five feels like too much.


  1. Same wake time every day, even on weekends. Your wake time sets your sleep time. Pick one you can keep and protect it.

  2. Same bedtime, within half an hour, seven nights a week. Regular hours teach the night crew when the shift starts.

  3. Light in the morning, dark at night. Get outside or near a bright window in the first hour after waking. Dim the lights in the last hour before bed.

  4. Move during the day. A walk, a chair workout, gardening, anything that tires the body a little. A tired body sleeps deeper. Do the harder movement earlier in the day when you can.

  5. A wind down, not a scroll down. In the last 30 minutes before bed, put the phone in another room, lower the lights, and do something quiet. Reading, stretching, a warm drink. If your mind races, get up, sit in a dim room, and come back when you feel sleepy. The bed is for sleep, not for worrying.

Two warnings that matter for this audience. First, sleeping pills and sedatives raise fall risk and confusion risk in older adults, and they do not give you the deep sleep your brain needs.


If you take one, do not stop it on your own. Talk to your doctor about a plan, and ask whether CBT-I is an option for you.


Second, loud snoring, gasping, or waking up unrefreshed most mornings is worth a conversation with your doctor. Sleep apnea is common, treatable, and it steals deep sleep night after night.


The bottom line


Sleeping 6 hours or less in your 50s and 60s was linked to a 30 percent higher dementia risk over 25 years in almost 8,000 people, with 7 hours as the sweet spot. Deep sleep is where the filing and cleaning happen, and in a small Berkeley study weaker deep sleep meant worse overnight memory. But a careful trial that fixed sleep for a year did not move memory scores, so sleep is a leg of the table, not the whole table.


That is still good news. Food, movement, sleep, brain training, and people, that is the table. Last week was food. This week is sleep. Next week is strength. Each leg holds a little of the weight, and together they hold a lot.


Tonight, give the night crew a full shift. Same time, same routine, lights low, phone away. Your future self, the one still sharp, still independent, still remembering where she put her keys, will thank you.


If you want help building a sleep routine around your body, your schedule, and the way you actually live, that is what I do. Fill out my contact form below (click the APPLY NOW buttton). I've included a free download for you to help you work on improving your sleep.



Danny



If you or a family member can benefit from my services I'd love to help. Apply to work with me one-on-one below.





Sources: Sabia and colleagues, Nature Communications, 2021. Mander and colleagues, Nature Neuroscience, 2015. Ritterband and colleagues, npj Digital Medicine, 2025. SIESTA trial, Alzheimer Disease and Dementia, 2026.

This post is educational, not medical advice. Talk to your doctor before changing your sleep habits, especially if you take medication for sleep, blood pressure, blood sugar, or mood, if you snore loudly or gasp at night, or if you have fallen asleep during the day.

 
 
 

Comments


bottom of page