ISC Global Webinar Series. Episode 1 - Professor Michael Chee

The inaugural webinar keynote, lead by world-renowned neurologist Professor Michael Chee, Director of the Centre for Sleep and Cognition, National University of Singapore, Yong Loo Lin School of Medicine.

In this short webinar, Prof. Chee will discuss:

i. The motivation for Need for Sleep studies.

ii. How different sleep restriction schedules affect fundamental cognitive processes in adolescents.

iii. The benefits of a 60-90 minute afternoon nap.

Video Transcript:

Need for Sleep - the case for improving Adolescent Sleep.

International Sleep Charity Global Webinar Series, inaugural keynote.  

[Charlie] Ok, hello and a very warm welcome. We are delighted to invite you to the International Sleep Charity's first Global Webinar. 

My name's Charlie Oulton and I'm one of the founders of the charity and our primary focus is helping people sleep better through facilitating education, scientific research and instigating a platform for International, Global collaboration. You can find out more on our website internationalsleepcharity.org. 

This webinar series is openly accessible over Zoom and each session is about 45 minutes in length. We invite expert clinicians, thought leaders and scientists from around the world to give a brief, sleep themed talk and then conduct a relaxed, intelligent conversation with our guest incorporating questions that you submit in the filtered question and answers, and we will do that at the end of this session.

Now, Professor Michael Chee of the National University of Singapore joins us today and he's a brilliant first guest. He's involved in a really important area of research, he's top of his field and the bit I probably love best of all, he's also good at getting his research implemented for measurable benefit. 

Mike is the director of the Centre of Sleep and Cognition at the NUS Yong Loo Lin School of Medicine. He's a physician and neurologist by training and his present research seeks to reduce the impact of degradable, restricted sleep on cognitive performance, wellbeing and health span. 

I think it's notable that his advocacy for science and its relevance to the common person, and also, his works have appeared in leading international media such as Time Magazine, The Economist and the BBC. He has appeared in scientific documentaries and written commentaries on topics on health science and the importance of sleep and in 2009 he received the National Outstanding Clinician Scientist Award, and this year he has become an inaugural fellow at the organisation of human brain mapping.

So, Mike, thank you so much for joining us from your home in Singapore, where it's late at night.

[Prof. Chee] Thank you Charlie and Connor for the invitation to share this material with you. Shall I get started?

[Charlie] Yes, I think you are the main event Mike so please I'm going to stop sharing and over to you.

[Prof. Chee] Ok. Just give me a second. Alright, so we can see. Right, so I thought I'd share about several studies that I have been associated with now for 6 years which I'm very proud of. It's the Need for Sleep series of studies and I'll show you a link to the microsite at the end of the talk.

Alright, so I think most of us involved in education or being educated, really like to optimise our performance and you know, for those of us in Singapore, in the Far East, sleep is an area that is a tremendous opportunity for us to optimise performance. So, I'm going to share with you a little bit about the motivation for this series of 5 studies conducted so far.

How different sleep schedules affect fundamental cognitive processes in adolescents. 

What the benefits of a long-ish 60-90 minute mid-afternoon nap are, and then just a touch on what potential neural correlates of these naps might be.

03:55 So, the issue of adolescents not getting enough sleep is a global issue, but it's probably more widespread in east Asia where I come from.

And when one talks about recommendations and recommended sleep duration to young people in our part of the world, frequently, when one mentions 8-10 hours, typically we are met with some laughter because more than 85% of our secondary school students sleep less than this recommended amount. You know, this despite the fact that specialists and scientists, neuroscientists have in particular have identified, you know, adolescents as you know, a second window of opportunity for us to influence brain development, because it's a period of vulnerability as well as opportunity, they are you know, and the habits that one puts in place at this time of life can, you know, set us up for a lifetime.

05:13 So, brain imaging has shown that dramatic changes in the maturation in the brain occur in adolescents and here you can see this pretty dramatic thinning of the cerebral cortex, representing the pruning of neurons and remodelling of the brain and alongside we can see these are the respectful compositions of EEG showing the differences in the EEG signatures as the adolescent brain matures.

The key obstacle in, it sounds strange, it's an obstacle, it's something we are very proud of in some ways, to getting more sleep in our part of the world is the obsession on academic success and you can see that Singapore's PISA in 2016 was pretty much top of the board and many of the countries in East Asia are very proud of that position.

However, this extreme focus on academics comes at a cost, and one of the offsets is actually reduction in sleep duration. 

06:36 In this plot that comes from the work of my colleague Josh Gooley, which I'm proud to be part of. You can see that there is a very strong 06:42 doze effect and this is sleep duration and you can see where shorter sleep duration, I hope you can see my cursor, there is an increased risk of self-harm thought, physical signs of anxiety, anxiety itself, feelings of worthlessness. So the mental health impact of sleep restriction is, short sleep is very significant and something which is too big to ignore right, and certainly, one does not like to be near grumpy people!

You know, when we first started this work, the literature until 2015 suggested that there were likely effects on sustained attention or vigilance. Possible effects on working memory and executive function and that sleep after learning helps but it's less clear that sleep before learning is useful. And those of you who are interested can look at this systematic review by Eduard de Bruin in about 2017.

Now one of the works that actually spurred us on to perform the investigations we subsequently did was this work by Voderholzer who actually put adolescents through different levels of sleep restriction from 7 down to 5 hours and had participants encode material before this period of sleep restriction and I think they underwent about several days of sleep restriction and there was apparently no effect of when memory was tested at day 8 and day 36 and they argued that this was because of critical stage of sleep, slow wave sleep was relatively well defended even when you went down to 5 hours time in bed at night and this was over several nights of sleep restriction.

So in 2014 we launched the first of the Need for Sleep series of studies. Each of these studies was conducted in a high school dormitory and involved secondary school students between the ages of 15 to 19 and we typically were able to get just shy of 60 participants for each of these studies and they stayed in this dormitory for 15 to 17 days under close 09:20  laboratory conditions.

They all underwent standardised tests 3 times a day from which you get a lot of the data that I'm going to share now. And the participants were typically healthy 15 - 18 years of age and they were high performing students. So, we focussed initially on fundamental cognitive constructs like attention, working memory, executive function, speed of processing because external research has shown that they predict general intelligence which in turn predicts success in a variety of life outcomes. So just to remind you, we tested participants 3 times a day and analysed the results. 

So this is what the study protocol for the first 5 studies looks like. There were 3 baseline nights where the participants had 9 hours time in bed or sleep opportunity and then the group of about 60 participants was split into a control group who slept 9 hours and a manipulation group M who slept 5 hours a night for in this case 7 nights and you can see that the objective evidence that this schedule was maintained and there was a period of recovery sleep.

So, our most robust test of the effects of sleep restriction was something we called the Psychomotor Vigilance Task in which one you know, looks out for the appearance of a running series of digits and they are supposed to respond as quickly as possible. And after this you see your response time. If your response time is twice that of the average, which is about half a second, it is marked as a lapse and you can see that increasing the cumulative increase in number of lapses as successive nights of sleep restriction occurs and then incomplete recovery, during the recovery nights of sleep.

Now what's interesting that in these results shown by 11:36 and others, that while the objective measure of vigilance shows cumulative worsening through successive nights of sleep restriction, subjective sleepiness does worsen but it tends to plateau after a few nights and this is the association between objective and subjective markers of sleepiness that is so important for young people to recognise because it is like boiling a frog, you do it slowly, the frog might not notice but if you just dumped it in hot water it would jump out.

So, in addition to vigilance, speed of processing evaluated by 2 digit mental arithmetic or a symbol digit modalities test where the participant has to decode the symbols and punch in digits as quickly as possible shows in the control group gradual improvement but in the sleep restricted group, an impairment of improvement which doesn't fully catch up during recovery sleep and this works for both tests.

And with working memory and executive function this is that N-Back task where one, a classic task of working memory and we can see a decline [of the test score] after multiple nights of sleep restriction. Finally, very important, mood declines, positive mood is significantly lowered after a few nights of successive sleep restriction.

Now, in the subsequent studies, we shortened the period of manipulation to 5 nights to simulate a school day week. What we found was we replicated the effect of cumulative decline in vigilance incomplete recovery on the simulated weekend but what was interesting is that upon the exposure to a second bout of sleep restriction, the effects were compounded. 

So we liken this repeated exposure on successive nights of sleep restriction like going above the 8,000 metre line in mountaineering which is the death zone and even people who are supremely fit can not survive for extended periods there. So, sleep restriction has the analogy there, you will suffer if you are exposed repeatedly to sleep restriction.

Now, when we do surveys on our secondary school students and not hundreds but thousands of them, we found that amongst the higher performing students, many of them claim that they only sleep 6.5 hours a night on weekday nights. We tested this claim in our 14:43 laboratory study and we found that you can't sustain optimal vigilance with this level of sleep restriction so what's happening is likely these students are falling asleep in class or they're taking naps in the afternoon in order to obtain some semblance of attention in their classes. So you can see that 6.5 hours a night is not much better than 5 hours a night, in fact it is just as bad but you only see the lessening of the duration in the second week of sleep restriction. So 6.5 hours per night isn't enough for optimal vigilance performance but the recommended 8 hours a night will suffice and you can see this is the result of a study that we just completed and published recently. 8 hours a night will give you comparable vigilance performance to the control group who slept 9 hours a night.

Now, it's very difficult in many places to persuade governments or school boards to shift school start times to a later time, and we in Singapore start at 7.30 already so it's pretty bad. It's very difficult to shift the entrenched desire to sleep late to study more and then have to wake up early for school but maybe you can do something with afternoon naps. So when we introduced a 60 minute mid-afternoon nap we found that it had a really dramatic effect on reducing the vigilance decline in our adolescent participants. As you can see here, with the 1 hour nap, there isn't complete normalisation of performance. You see the improvement in sleepiness which is also very strong.

Now, what is interesting is that you can nap without, some people advise against naps in adults because they're concerned it can compromise nocturnal sleep, but it doesn't and it doesn't when sleep is restricted to 5 hours time in bed. It's not even, nocturnal sleep is not even compromised if you have 6.5 hours of sleep at night, as I will show you in a bit.

So what about other combinations of sleep? So in particular, how does an equivalent amount of sleep over 24 hours pan out if you take it in one continuous swoop at night versus a shorter nocturnal sleep and a mid-afternoon nap? So this is a question we posed in the 4th and 5th edition of our Need for Sleep series of studies and here what you see is that 17:45 the study protocol we compared 8 hours at night continuous versus split into 6.5 and then 1.5, or 6.5hours at night compared to 5 hours at night and 1.5hours afternoon nap. 

So this is the summary of a lot of data collected, we can see that naps have a very dramatic effect on reducing vigilance lapses but you know, when sleep, the total sleep time is restricted, they have a dramatic benefit that they don't normalise performance. 8 hours will cut it and what is very interesting is how 8 hours of sleep is distributed doesn't materially affect vigilance. Now, whether it affects other things like metabolic function, we don't know yet, but 8 hours, if you take it continuously or you split it, seems to give the same, pretty good results. 

Now, many students go to school now to just stay awake and hammer, you know, play the game of knock them all out, but to learn about things and in particular they are there to accumulate a fair number of facts to get through school. There is a wealth of research talking about the importance of sleep for memory consolidation and less research on encoding, which we specialise in, but here is a little schema showing schematic, showing how you could test the different aspects of memory processes, so learning something effectively involves properly encoding the material, consolidating it, meaning that you protect it from either decay or interference with similar but actually not the same material, and then retrieval. And these are ways in which you can evaluate encoding after multiple nights of sleep restriction and then you allow for some recovery and then you can test recall. And for consolidation you would learn, allow for either a proper sleep or short sleep and then test recall, and finally for retrieval, you would learn with adequate sleep and then allow adequate sleep consolidation and you test retrieval and then you can test the same thing with shortened sleep.

Now, picture encoding is a task that we used in our experiments and it was inspired by this work by Matt Walker's group showing that there is impaired hippocampal recruitment following a night of total sleep deprivation when participants had to encode pictures. And we found that the cortical representation in higher visual areas is compromised if you had to encode this material when you are sleep deprived.

So we have actually evaluated, and if you look through what we have produced, different tests on memory. But, today in the interest of time, I'll just talk about picture encoding and fact learning. So, with picture encoding, what participants do is they view pictures and they have to make decisions on whether the picture was a building or not a building, incidentally encoding the picture and then they would test it later, whether they recall seeing a particular picture, and the confidence with which they made that judgement.

So, we allowed participants to  undergo a baseline, normal sleep and then the manipulation period and then the test and encoding after different doses of sleep restriction. And, what we found was that, with 9 versus 5, and this is pretty obvious, you're going to be better off with 9 hours than 5 hours. What was surprising is that if you have only 6.5 hours of sleep, the split sleep schedule  gave as much or almost equivalent performance to the 9 hour group. Now, we put this on separate plots because the timing of the encoding and testing was slightly different, but in a subsequent experiment, comparing the total of 6.5 hours and total of 8 hours, you can see that pretty much, you get the same result. So, the split sleep schedule is actually superior for picture encoding.

So, episodic encoding, split sleep schedule looks superior to continuous sleep.

Now fact learning is something all students have to do and maybe are not crazy about as it is really quite demanding - this kind of learning involves a body of facts, multiple facts, involves multiple learning opportunities, because you are reading material, trying to digest it, trying to find structure and a way these facts are organised so you set up what is called a shema and retrieving this information, consolidating it and then retrieving it involves long term memory and this is how we tested it.

James Cousins (scientist name) who was involved created many colourful flashcards contains numerous facts about different sets of creatures so that, frogs, salamanders, you can see from this that there is structure involved in learning multiple facts about different animals and the some of the ways, the way in which you can retain that information is benefited from properly learning the structure of information about one particular animal.

24:21 We had two sets of creatures in here (frog and newt), and then, they were learned either in the morning following sleep restriction or in the afternoon after a nap or without a nap. And then participants were tested at a neutral time. So here we looked at the interaction between nap opportunity, time of day learning and total sleep duration over 24h. And what we found was that independent of whether the participants slept 6.5 or 8h in total students assigned the sleep split group performed better than those in the continuous sleep schedule. And what is interesting is that the benefit of learning in the afternoon, following a nap, did not come at a cost of compromising morning learning. So, it thus seems that the split sleep schedule is better for leaning in the afternoon following a nap. 

25:41 So how is this realised, what is the magic behind this? I am gonna go to that in a second.

But first I am gonna share another study we did on napping, which is comparing the relative benefit of napping vs cramming. If you had time in the afternoon, would you rather nap or get another bout of learning. So J Cousins again, introduced this experiment, where participants had 4 bouts of leaning, but during the experimental manipulation period they could either nap for an hour, continue learning or just chill or take a break, and then they had these biology flash cards to pick up. And what James found was that at 30 min after the last learning opportunity, participants who nap or cram both did better in retrieving the factual information than those who stayed awake and watched a video. But at 7 days, whether there was overall certain memory, the participants who napped were the ones who showed better memory scores compared to those who crammed. So if you have the time, it is probably better to take a nap than take a break.

27:13 So how do naps work to benefit memory, what is the secret sauce? Now during sleep there are two things that happen that may benefit memory. Firstly, all the synapses and connections that formed as a result of your learning activities, if they are allowed to continue, could compromise further learning. So downscaling, the pruning of less salient neural connections, synapses, occurs during sleep. And the other thing is that sleep oscillations like slow waves provide a milieu for systems consolidation and memory. 

And what we found is that the nap benefit of encoding, what you can see here, is associated with increased hippocampal activation following a nap, and this increases hippocampal activation compared to the baseline period, or the first encoding before the nap. And just to remind you, the naps also provide a second opportunity to reduce sleep pressure. How you can tell this, is that the slow wave activity in the nights following the split sleep schedule is lower, than the initial slow wave activity suggesting that there is an induction in the homeostat of the sleep pressure if you provide an opportunity for the afternoon nap this may help restore the encoding

29: 06 Slide title: At 6.5 Nocturnal sleep opportunity everyone can nap, and everyday too. So who gains from afternoon napping? Because not everyone naps habitually. What we found is that when participants are given an 8h of sleep opportunity, all participants could sleep in the afternoon, so even with slightly reduced nocturnal sleep the typical adolescent is able to nap. 

To summarise what we have learned: 

• Naps can offset the effects of multi-night sleep restriction on vigilance and memory encoding.  

• Naps have a restorative effect perhaps on hippocampal function 

• Most adolescents can nap if allowed to 

• This ability to nap in the afternoon does not compromise the nocturnal sleep (No compromise in nocturnal sleep with 90 min mid afternoon nap)

• If available nocturnal sleep time is curtailed, split sleep schedule incorporating naps yield favorable cognitive outcomes, I think herein lies our contribution to the science of learning. 

Questions: 

Is the circadian time of the nap important? Generally speaking, we are talking about studies that were performed with the students sleeping a reasonable sleep schedule. However, in modern times, the undergraduates who do online learning now sleep funny sleep schedules. With a  caveat that we are talking about a sleep schedule with a 11 pm onset and waking up early in the morning. A mid afternoon nap is what we think is suitable, at the moment we are actually looking at different sleep timings. Some students say then can only get back home at 6 in the evening, and that is when they crash and then they can sleep for 2 to 3 h which is going to interfere with nocturnal sleep if you sleep that long and close to the sleep forbidden zone. Mid afternoon naps are what we think works, as you get a fair amount of slow wave sleep, you get less REM, and we did not have a problem with sleep inertia (?)

What have you done in Singapore about starting the schools at a later time? We ran a study on a group of high school girls. In this girl’s high school the starting time was moved from 7:30 to 8:15, 45 min, time in bed was prolonged for 22 min, the actual gain in sleep time was 10 min, but subjectively the girls felt better and they kept that schedule now, it has been 5 years. However it has been very difficult to get schools to participate. There is evidence starting school later is beneficial, the challenge is social and political. 

He is currently looking at power naps. The reason they looked at longer naps was to follow up other studies, e.g. at Harvard, they wanted it to be longer in order to include some REM sleep, so that in this period of sleeping you can get a full cycle, but for our students, with valuable sow wave sleep, which restores memory encoding activity, young brains hold to slow wave sleep like sponges, and they are very good at napping, the fall asleep immediately. 

What about the effects of food on the study? His colleague measured it, has to write them up. They had to keep buying snacks for them during sleep restriction. 

Work coming up: Nap titrations, different durations and different combinations with nocturnal sleep and nap, also sleep regularity. Also metabolic consequences, the split sleep was good for learning but it was not good when given a bout of glucose. We have to look at sleep as a holistic effect on sleep.